Head-to-tail and heteroleptic pentanuclear circular helicates

Article abstract of DOI:10.1002/anie.201003342

Everybody form a circle: Careful design of ligand strands and their reaction with Cu2+ ions leads to the formation of the title helicates. Incorporation of differing numbers of N-donor units within a ligand strand containing a phenyl spacer results in a p

Full text of DOI:10.1002/anie.201003342

Angewandte
Chemie
DOI: 10.1002/anie.201003342
Supramolecular Chemistry
Head-To-Tail and Heteroleptic Pentanuclear Circular Helicates
Kirsty E. Allen, Robert A. Faulkner, Lindsay P. Harding, Craig R. Rice,* T. Riis-Johannessen,
Melanie L. Voss, and Martina Whitehead
The rational design of polynuclear helicates is one of the
major achievements of metallosupramolecular chemistry.[1]
These linear structures form by self-assembly and consist of
two or more multidentate ligand strands that are helically
wrapped about a central array of metal cations.[1] Not only can
polynuclear double-, triple-, and quadruple-stranded heli-
cates now be made in a predictable fashion,[1f] they can also be
programmed to express certain structural features of higher-
order complexity. This goal may be achieved by elaborating
on the basic design principles that govern helicate formation
itself (i.e., careful consideration of ligand topology and metal
Figure 1. Formation of a double-stranded a) HT helicate, b) HH heli-
cate, and c) heteroleptic helicate: black, grey, and open circles repre-
sent five-, four-, and six-coordinate metals, respectively.
stereoelectronic preference) and, amongst others, can entail:
1) directional control over ligand alignment, 2) selective
incorporation of different metal cations, and 3) selective
incorporation of different ligand strands, within the helical
array.[1]
The first two of these scenarios are intimately linked. They
can be illustrated by considering a ditopic Cs-symmetric N-
donor ligand with one bidentate and one tridentate donor
site.[2] When combined with equal amounts of a five-
coordinate metal cation, a double-stranded dinuclear C2-
symmetric head-to-tail (HT) helicate results, wherein each
metal is coordinated by a “2+3” donor set comprising the
tridentate head of one ligand and the bidentate tail of the
other (Figure 1a). Alternatively, combination of the same
ligand with half an equivalent each of a four- and six-
coordinate cation gives a hetero-bimetallic head-to-head
(HH) helicate. In this case, one metal is octahedrally (3+3)
coordinated by the two tridentate heads whilst the other is
tetrahedrally (2+2) coordinated by the two bidentate tails
(Figure 1b).[2] Importantly, the two ligands are now co-
aligned in a parallel fashion. The third scenario can likewise
be realized by combining a five-coordinate cation with half an
equivalent each of a C2v-symmetric bis(bidentate) and
bis(tridentate) ligand. The resulting structure is a C2-sym-
metric heteroleptic (mixed ligand) helicate, wherein both
cations have the required bi- and tridentate (2+3) donor sets
(Figure 1c).[3]
More subtle effects, such as interligand interactions[4] and
cation or binding-site size mismatch,[5] can also be used to
fine-tune structural complexity in linear helicates. However,
neither these effects nor the more intuitive principles
described above have ever been applied to helicates that
have a higher nuclearity, that is, circular helicates. Herein we
describe how some of the basic rules that govern linear
helicate assembly can also be applied to generate the first
heteroleptic and head-to-tail circular helicates.
If the linear double-stranded helicates discussed above are
denoted as [M2(L)2], then circular helicates are cyclic
oligomers that have the general formula [Mn(L)n] (n > 2)
and retain the “over-and-under” ligand motif requisite of
helical chirality.[6] Circular helicates can arise from intermo-
lecular templating effects or intramolecular interactions that
destabilize their more entropically favored linear counter-
parts.[6] In a recent report, we showed that bis(tridentate)
ligand L1 (Scheme 1) forms pentanuclear circular helicates
with small transition-metal dications.[7] Formation of the
observed structure was driven by repulsion between the
central phenyl groups of L1, which are brought too close to
one another in the putative dimer. Since this appeared to be a
[*] K. E. Allen, R. A. Faulkner, Dr. L. P. Harding, Dr. C. R. Rice, M. L. Voss,
M. Whitehead
Department of Chemical and Biological Sciences
University of Huddersfield
Huddersfield HD1 3DH (UK)
Fax : (+44)148-447-2182
E-mail: c.r.rice@hud.ac.uk
Dr. T. Riis-Johannessen
EPFL SB ISIC LCS, BCH 3307 (Bꢀtiment de chimie UNIL)
1015 Lausanne (Switzerland)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003342.
Scheme 1. Multidentate ligands L1–L3.
Angew. Chem. Int. Ed. 2010, 49, 6655 –6658
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6655

Communications
robust approach for generating circular helicates, we decided
quantitatively retained in solution. Nonetheless, we have
conducted tandem MS experiments to ascertain whether the
lower nuclearity species observed in the ESI-MS spectrum of
[Cu5(L2)5](ClO4)10 are present in solution or merely products
of gas-phase fragmentation in the ion source. The results
supported the latter scenario: selective collision-induced
fragmentation of the ion at m/z 1745 gave rise to peaks at
m/z 901, 1376, and 2114, thus indicating clear correlation
between the pentanuclear parent ion and {[Cu2(L2)](ClO4)3}+,
{[Cu2(L2)2](ClO4)3}+, and {[Cu3(L2)3](ClO4)5}+, respectively.
The peak at m/z 1114 that arises from {[Cu(L2)2](ClO4)}+ was
not observed, though it is possible that this species is a
fragment ion from further down the fragmentation cascade.
We next focused on forming the heteroleptic circular
helicate (Figure 1c). To this end, we chose C2v-symmetric
bis(tridentate) and bis(bidentate) ligands L1 and L3, respec-
tively. Arrangement of these ligands in a pentameric circular
array afforded any number of different structures, in which
the occurrence and disposition of four-, five- and six-
coordinate binding sites depends on the ratio and relative
orientation of the ligand strands. The use of mixtures of metal
ions with different stereoelectronic preference therefore
represents one approach for selecting a desired heteroleptic
species (i.e., a hybrid of the two cases outlined in Fig-
ure 1b,c). However, preliminary attempts were unsuccessful
and so we again opted for the use of CuII ions, in the hope that
its (relatively) diverse coordination tendencies would permit
the formation of a homo-pentametallic structure that features
both five- and six-coordinate binding sites.
to explore whether L1, and its variants L2 and L3 (Scheme 1),
could be used to further diversify structural complexity in
circular helicates.
We first addressed the issue of ligand directionality
(Figure 1a). Reasoning that a five-coordinate ion would
combine with a ditopic “3+2” dentate ligand to give a HT
motif regardless of overall structural topology, we prepared
Cs-symmetric L2 by standard transformations (see the Sup-
porting Information). As in other reports,[3] we chose CuII as
the metal dication since, amongst its possible coordination
geometries, five-coordinate CuII ions are frequently observed.
Reaction of L2 with an equimolar amount of Cu-
(ClO4)2·6H2O in MeCN gave a light green solution from
which a crystalline solid was deposited in high yield (ca. 80%)
after slow diffusion of chloroform. The ESI-MS spectrum of
this material showed a number of low-nuclearity fragments,
for example, {[Cu2(L2)](ClO4)3}+ (m/z 901, 3%), {[Cu(L2)2]-
(ClO4)}+ (m/z 1114, 50%), {[Cu2(L2)2](ClO4)3}+ (m/z 1376,
100%) and {[Cu3(L2)3](ClO4)5}+ (m/z 2114, 5%), but also a
peak at m/z 1745 corresponding to {[Cu5(L2)5](ClO4)8}2+, with
the correct isotope pattern for a dicationic species (see
Supporting Information). X-ray structural analysis confirmed
the formation of a pentanuclear circular helicate [Cu5(L1)5]-
(ClO4)10 (Figure 2).[8] In the solid state, the ligands adopt the
The diagram in Figure 3 represents one such possibility;
CuII, L1, and L3 appear in a 5:3:2 ratio, respectively, and one of
the metals occupies an octahedral binding site whilst the
remaining four are five-coordinate. ESI-MS studies indicate
Figure 2. Left: crystal structure of the HT-[Cu5(L2)5]10+ circular helicate.
Right: Schematic representation of HT-[Cu5(L2)5]10+ showing head-to-
tail ligand arrangement (black circles represent five-coordinate CuII
ions).
anticipated “3+2” binding mode, wherein the bidentate and
tridentate N-donor domains span two different CuII centers.
The ligands are furthermore arranged such that a given metal
is coordinated by the bidentate domain of one ligand and the
tridentate domain of the next. The resulting structure has
approximate C5 symmetry, with all CuII centers displaying
Figure 3. Left: Crystal structure of the heteroleptic [Cu5(L1)3(L3)2]10+
circular helicate. Right: Schematic diagram of [Cu5(L1)3(L3)2]10+ (black
and open circles represent five- and six-coordinate CuII ions, respec-
tively).
À
distorted square-based pyramidal geometries (Cu N bond
lengths: 1.936(9)–2.327(9) ꢀ). By analogy with the HT linear
helicate, the complex cation [Cu5(L2)5]10+ may thus be
considered a HT circular helicate.
The distribution of species in the ESI-MS spectrum of
[Cu5(L2)5](ClO4)10 is strikingly similar to that observed for the
previously reported zinc-based circular helicate [Zn5(L1)5]-
(OTf)10,[7] the solid-state structure of which was shown to be
that the reaction of Cu2+ with L1 and L3 in the ideal ratio (i.e.,
5:3:2) indeed gives the desired pentanuclear heteroleptic
complex [Cu5(L1)3(L3)2]10+, but other L1-rich species such as
[Cu5(L1)5]10+, [Cu5(L1)4(L3)]10+, and related fragments are
more abundant (see the Supporting Information). Attempts
at crystallizing the putative heteroleptic complex
6656 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 6655 –6658

Angewandte
Chemie
[Cu5(L1)3(L3)2]10+ were likewise unsuccessful. This result is
perhaps not surprising since, like its previously reported ZnII
analogue [Zn5(L1)5]10+, the homoleptic circular helicate based
on L1 would furnish all five CuII centers with a more favorable
six-coordinate donor set.[9] The relative L1/L3 ratio was
therefore reduced in order to drive the equilibrium towards
the desired heteroleptic species.[10] Combination of CuII, L1,
and L3 in MeCN in a 2:1:1 ratio, respectively, gave a solution
for which ESI-MS now showed virtually no traces of
[Cu5(L1)5]10+ and [Cu5(L1)4(L3)]10+. Peaks for heteroleptic
ions {[Cu(L3)(L2)](OTf)}+ (m/z 1164, 40%), {[Cu2(L3)(L2)]-
(OTf)3}+ (1525, 10%), {[Cu5(L3)3(L2)](OTf)8}2+ (1782, 5%)
and in particular {[Cu5(L3)3(L2)2](OTf)8}2+ (1982, 10%) were,
however, retained.
Since the components were combined in nonstoichiomet-
ric quantities, the solution clearly contains a mixture of
interconverting species. We are not able to estimate the extent
to which the target heteroleptic pentanuclear complex
dominates in solution as the ESI-MS response factors for
the ions observed in the gas phase are unknown. Nonetheless,
attempts at crystallization were successful. When ethyl
acetate was diffused into a MeNO2 solution containing
Cu(OTf)2, L1, and L3 in a 2:1:1 ratio, respectively, a
homogenous crystalline material was deposited in high yield
after several days.[11]
An X-ray analysis confirmed the formation of the target
heteroleptic circular helicate [Cu5(L3)3(L2)2]10+ (Figure 3).[12]
In the solid state, the complex features a cyclic array of five
CuII ions, bound by three strands of L1 and two strands of L3.
Since the structure has an odd number of ligands, one of the
CuII centers is octahedrally coordinated by two tridentate
thiazole–pyridyl–pyridyl domains from L1. The remaining
four CuII centers are bound by one tridentate domain from L1
and one bidentate domain from L3 to result in five-coordinate
donor sets. As for [Cu5(L2)5]10+, the ligands bridge adjacent
metals in an “over and under” conformation, thus giving the
complex a circular helicate topology with approximate
C2 symmetry.
thermodynamic self-assembly,[13] but rather a dynamic com-
binatorial selection process,[14] the outcome of which is highly
sensitive to the ratio of the two ligands. This factor differ-
entiates the circular complex somewhat from the reported
examples of its linear counterpart,[3] although it also opens up
interesting avenues for the design of complex new architec-
tures by targeted or template-based syntheses.
Received: June 2, 2010
Published online: July 29, 2010
Keywords: copper · helical structures · heteroleptic complexes ·
.
N,S ligands · supramolecular chemistry
[1] a) J.-M. Lehn in Supramolecular Chemistry, VCH, Weinheim,
1995; b) J. W. Steed, J. L. Atwood in Supramolecular Chemistry,
Wiley, Chichester, 2000; c) M. J. Hannon, L. J. Childs, Supramol.
Chem. 2004, 167; d) M. Albrecht, Chem. Rev. 2001, 101, 3457;
e) M. Albrecht, Chem. Soc. Rev. 1998, 27, 281; f) C. Piguet, G.
Bernardinelli, G. Hopfgartner, Chem. Rev. 1997, 97, 2005;
g) E. C. Constable in Comprehensive Supramolecular Chemistry,
Vol. 9 (Ed.: J.-P. Sauvage), Elsevier, Oxford, 1996, p. 213; h) J.-
M. Senegas, S. Koeller, C. Piguet, Chem. Commun. 2005, 2235 –
2237; i) R. Krꢁmer, J.-M. Lehn, A. Marquis-Rigault, Proc. Natl.
Acad. Sci. USA 1993, 90, 5394; j) G. I. Pascu, A. C. G. Hotze, C.
Sanchez-Cano, B. M. Kariuki, M. J. Hannon, Angew. Chem.
2007, 119, 4452; Angew. Chem. Int. Ed. 2007, 46, 4374; k) N.
Andrꢂ, R. Scopelliti, G. Hopfgartner, C. Piguet, J.-C. G. Bꢃnzli,
Chem. Commun. 2002, 214; l) M. Albrecht, M. Schneider, H.
Rꢄttele, Angew. Chem. 1999, 111, 512; Angew. Chem. Int. Ed.
1999, 38, 557.
[2] A pentadentate C2v-symmetric ligand that is able to partition
into one tridentate and one bidentate donor unit is also a
possibility. See for example: a) M. Barley, E. C. Constable, S.
Corr, R. C. S. McQueen, J. C. Nutkins, M. W. Ward, M. G. B.
Drew, J. Chem. Soc. Dalton Trans. 1988, 2655; b) C. R. Rice, C. J.
Baylies, H. J. Clayton, J. C. Jeffery, R. L. Paul, M. D. Ward,
Inorg. Chim. Acta 2003, 351, 207.
[3] For related examples involving tris(tridentate) and tris(biden-
tate ligands) see: a) V. C. Smith, J. M. Lehn, Chem. Commun.
1996, 2733; b) B. Hasenknopf, J.-M. Lehn, G. Baum, D. Fenske,
Proc. Natl. Acad. Sci. USA 1996, 93, 1397.
In summary, we have established that some of the basic
algorithms for programming structural complexity in linear
helicates can also be applied to related cyclic complexes. The
pentanuclear circular helicates HT-[Cu5(L2)5]10+ and
[Cu5(L1)3(L3)2]10+ are analogous to their linear counterparts
shown in Figure 1a,c by virtue of both their specific structural
features and the design principles employed in their synthesis.
The formation of these head-to-tail (in the case of L2) and
heteroleptic (in the case of L1 and L3) helicates is a result of
two main factors. Firstly, the phenylene spacer units which
connect the various N-donor chelates and which prevent the
ligands from forming linear double-stranded assemblies (with
small metal cations). Secondly, the stereoelectronic prefer-
ences of CuII ions, which are sufficiently versatile to enable
both 1) all-five- or all-six-coordinate sites, or 2) a mixture of
five- and six-coordinate sites to occur in the same polynuclear
assembly.
[4] a) T. K. Ronson, H. Adams, T. Riis-Johannessen, J. C. Jeffery,
M. D. Ward, New J. Chem. 2006, 30, 26; b) G. Baum, E. C.
Constable, D. Fenske, C. E. Housecroft, T. Kulke, Chem. Eur. J.
1999, 5, 1862; c) M. Greenwald, D. Wessley, E. Katz, I. Willner,
Y. Cohen, J. Org. Chem. 2000, 65, 1050; d) M. Shaul, Y. Cohen, J.
Org. Chem. 1999, 64, 9358; e) C. R. Rice, C. J. Baylies, L. P.
Harding, J. C. Jeffery, R. L. Paul, M. D. Ward, J. Chem. Soc.
Dalton Trans. 2001, 3039.
[5] a) T. B. Jensen, R. Scopelliti, J.-C. G. Bꢃnzli, Inorg. Chem. 2006,
45, 7806; b) T. B. Jensen, R. Scopelliti, J.-C. G. Bꢃnzli, Chem.
Eur. J. 2007, 13, 8404.
[6] a) “Supramolecular templating in the formation of helicates”:
M. Albrecht in Topics in Current Chemistry, Vol. 248 (Eds.: C. A.
Schalley, F. Vꢄgtle, K.-H. Dꢄtz), Springer, Heidelberg, 2004,
p. 105; b) B. Hasenknopf, J.-M. Lehn, N. Boumediene, E. Leize,
A. Van Dorsselaer, Angew. Chem. 1998, 110, 3458; Angew.
Chem. Int. Ed. 1998, 37, 3265; c) B. Hasenknopf, J.-M. Lehn, N.
Boumediene, A. Dupont-Gervais, A. Van Dorsselaer, B. Knei-
sel, D. Fenske, J. Am. Chem. Soc. 1997, 119, 10956; d) S. P.
Argent, H. Adams, T. Riis-Johannessen, J. C. Jeffery, L. P.
Harding, O. Mamula, M. D. Ward, Inorg. Chem. 2006, 45,
3905; e) J. Hamblin, F. Tuna, S. Bunce, L. J. Childs, A. Jackson,
W. Errington, N. W. Alcock, H. Nierengarten, A. V. Dorsselaer,
It is not entirely clear to what extent the structures persist
in solution since the systems were not amenable to NMR
investigation. Isolation of the heteroleptic complex
[Cu5(L1)3(L3)2]10+, in particular, is likely not a result of strict
Angew. Chem. Int. Ed. 2010, 49, 6655 –6658
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6657

Communications
E. Leize-Wagner, M. J. Hannon, Chem. Eur. J. 2007, 13, 9286;
the homoleptic pentanuclear helicate [Cu5(L1)5]10+ (see the
Supporting Information).
f) Y. Pang, S. Cui, B. Li, J. Zhang, Y. Wang, H. Zhang, Inorg.
Chem. 2008, 47, 10317; g) L. J. Childs, M. Pascu, A. J. Clarke,
N. W. Alcock, M. J. Hannon, Chem. Eur. J. 2004, 10, 4291; h) L. J.
Childs, N. W. Alcock, M. J. Hannon, Angew. Chem. 2002, 114,
4418; Angew. Chem. Int. Ed. 2002, 41, 4244; i) G. Baum, E. C.
Constable, D. Fenske, C. E. Housecroft, T. Kulke, Chem.
Commun. 1999, 195.
[10] E. Coronado, J. R. Galan-Mascaros, P. Gaviꢆa, C. Marti-
Gastaldo, F. M. Romero, S. Tatay, Inorg. Chem. 2008, 47, 5197.
[11] Cu5(L1)3(L3)2](OTf)10 can be obtained in good yield (ca. 75%)
from either MeNO2 or MeCN by diffusion of a variety of
nonpolar solvents. Several batches of crystals were produced
from both solvents and either a full data set or unit cell was
collected. All data gave the same cell parameters as those
observed for the structure reported.
[7] L. Bain, S. Bullock, L. Harding, T. Riis-Johannessen, G. Midgley,
C. R. Rice, M. Whitehead, Chem. Commun. 2010, 46, 3496.
[12] Crystal data for [Cu5(L1)3(L3)2](OTf)10: C150H88Cu5F30N26O30S20,
[8] Crystal
data
for
[Cu5(L2)5]-
ꢀ
0.08 ꢅ 0.06 ꢅ 0.04 mm,
21.0573(13),
92.2160(10),
Mr = 4263.36, Triclinic,
b = 21.0816(13),
b = 100.0820(10),
P1, a =
(ClO4)10·2CHCl3·0.47MeCN·0.53H2O:
c = 22.4884(14) ꢀ,
g = 109.2860(10)8,
a =
V=
C137.93H88.39Cl16Cu5N25.47O41.53S10, 0.28 ꢅ 0.20 ꢅ 0.20 mm, Mr =
ꢀ
3972.46, Triclinic, P1, a = 17.2058(8), b = 20.4431(10), c =
9227.2(10) ꢀ3, Z = 2, Z’ = 1, 1c = 1.534 MgmÀ3, F(000) = 4290,
25.9847(12) ꢀ,
a = 87.9110(10),
b = 85.8010(10),
g =
m(MoKa) = 0.896 mmÀ1
68246 measured reflections, 18187 unique reflections (Rint
,
T= 100(2) K, l(MoKa) = 0.71073 ꢀ.
81.7440(10)8, V= 9017.7(7) ꢀ3, Z = 2, Z’ = 1, 1c = 1.463 MgmÀ3
,
=
F(000) = 4007, m(MoKa) = 1.010 mmÀ1, T= 100(2) K, l(MoKa) =
0.71073 ꢀ. 68110 measured reflections, 17769 unique reflections
(Rint = 0.0526). The structure was refined on F2 to Rw = 0.0844,
R = 0.2369 (12837 reflections with I > 2s(I)) and GOF = 1.110
on F2 for 1978 refined parameters and 1256 restraints. Largest
difference peak and hole 1.813 and À1.463 eꢀÀ3. CCDC 778980
contains the supplementary crystallographic data for this com-
pound. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.
uk/data_request/cif.
0.0506). The structure was refined on F2 to Rw = 0.0919, R =
0.2619 (11972 reflections with I > 2s(I)) and GOF = 1.108 on F2
for 2282 refined parameters and 1113 restraints. Largest differ-
ence peak and hole 1.121 and À1.370 eꢀÀ3. CCDC 778979
contains the supplementary crystallographic data for this com-
pound. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.
uk/data_request/cif.
[13] C. Piguet, M. Borkovec, J. Hamacek, K. Zeckert, Coord. Chem.
Rev. 2005, 249, 705.
[14] Dynamic Combinatorial Chemistry, (Eds.: J. N. H. Reek, S.
Otto), Wiley-VCH, Weinheim, 2010.
[9] The ESI-MS spectrum for a solution containing only CuII and L1
in equimolar amounts shows numerous peaks corresponding to
6658 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 6655 –6658