Synthesis of a four-component [3]catenane using three distinct noncovalent interactions

Article abstract of DOI:10.1039/c2ob26587g

A multicomponent assembly is described resulting in [2] and [3]catenanes using three flexible components and three distinct noncovalent interactions. Despite the possibility of competing side-products, only the desired assemblies are generated and characterized spectroscopically. The Royal Society of Chemistry.

Full text of DOI:10.1039/c2ob26587g

Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: DOI: 10.1039/c2ob26587g
View Article Online
View Journal
www.rsc.org/obc
COMMUNICATION
Synthesis of a four-component [3]catenane using three distinct noncovalent
interactions†‡
Miguel Á. Alemán García and Nick Bampos*
Received 10th August 2012, Accepted 17th September 2012
DOI: 10.1039/c2ob26587g
A multicomponent assembly is described resulting in [2] and
[3]catenanes using three flexible components and three
distinct noncovalent interactions. Despite the possibility of
competing side-products, only the desired assemblies are
generated and characterized spectroscopically.
Nature exploits many inter(intra)molecular interactions in para-
llel to create functional superstructures such as DNA, ribosomes
and Ca2+ ATPases.1 One of the major challenges in supramole-
cular chemistry is to harness several orthogonal weak inter-
actions to generate structures that display comparable complexity
and functionality to those found in biological systems. However,
as the diversity of building blocks and noncovalent interactions
Fig. 1 Schematic representation (MM+ modeling) of the [2] and [3]
involved in the assembly process increases, it becomes exponen-
catenane (1 and 2 respectively) in which each of the components is
tially more difficult to exercise control over the final structure.
Therefore, most of the systems reported to date employ one or
two distinct noncovalent interactions.2
colored for clarity (3 in red, 4 in blue and 5 in purple).
Interlocked molecular complexes such as rotaxanes and cate-
nanes have been the focus of major synthetic efforts as they offer
The components of our assembly exhibit significant flexibility
which comes with a ‘cost’ of organizational energy, but
which also allows sampling of the various possible confor-
mations that ultimately lead to the desired structure. This flexi-
the opportunity of being incorporated into molecular devices or
smart materials.3 While coordination bonds,4 hydrogen bonds5
or π–π stacking6 interactions have been elegantly used in the
bility may very well explain why more than one crown ether
construction of catenanes and rotaxanes, there are no comparable
threaded onto the heteroleptic macrocycle described below. It is
examples to those outlined in this report in which three (or
also noteworthy that due to the cooperative effect of the pyri-
more) noncovalent interactions are combined to assemble
dine–metal porphyrin interaction, it was possible to direct the
formation of a system that would otherwise be difficult to
mechanically interlocked molecules.
Our interest has been directed towards exploring heteroleptic,
assemble.
multicomponent architectures in which a number of noncovalent
interactions are exploited to generate a heteroleptic assembly, in
Fig. 1) is the structure of the metallomacrocyclic ring, which
the absence of competing side-products. As part of this study,
we report the synthesis and structural characterization of a new
A distinctive feature of the [n]catenanes (where n = 2 or 3,
consists of two components connected via two distinct noncova-
lent interactions: Ru–pyridine and heteroleptic Fe(II) (or Zn(II))
class of multicomponent [2] and [3]catenanes (1 and 2 respecti-
terpyridine. This is to the best of our knowledge the first
vely) utilizing up to four building blocks and three orthogonal
example of clean formation of heteroleptic terpyridine com-
plexes using Fe(II) or Zn(II)7 – in both cases the desired complex
interactions (Fig. 1).
was the exclusive product.
The components of the [n]catenanes were synthesised accord-
ing to Scheme 1 (see ESI†), whereby the naphthalenediimide
Department of Chemistry, Lensfield Road, University of Cambridge,
(NDI) thread 3 was synthesized from the commercially available
1,4,5,8-naphthalenetetracarboxylic dianhydride8 in an overall
yield of 13%. The crown ether macrocycle 4 was prepared using
the methodology previously described by Stoddart et al.,9 while
the porphyrin–terpyridyl component of the assembly 5 was
obtained in 40% yield from porphyrin 810 (either as the
Cambridge CB2 1EW, UK. E-mail: nb10013@cam.ac.uk;
Tel: +44 (0)1223 762970
†Electronic supplementary information (ESI) available: Synthetic and
experimental procedures are described in addition to spectroscopic data.
See DOI: 10.1039/c2ob26587g
‡This manuscript is dedicated to Prof. Maxwell J. Gunter in recognition
of his guidance, support and friendship.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem.

View Article Online
ruthenium or zinc metallated analogues) and compound 7 in the
presence of K2CO3.
catenane 1. Since the formation of 1 was not quantitative when
one equivalent of crown ether 4 was used, higher concentrations
of 4 were required to drive the equilibrium towards the exclusive
formation of the [2]catenane. Spectroscopic characterization
indicated formation of the unexpected [3]catenane 2 in addition
to the predicted [2]catenane 1 (as the major product). To prepare
the [3]catenane (2) a one-pot stepwise synthesis was adopted
(Scheme 2a) in which the first step involved the addition of a
saturated solution of 4 to a solution containing the NDI 3 in
CD2Cl2. A characteristic pale red color was observed (Fig. S1†)
indicating that the charge transfer complex had formed (which is
in equilibrium with the free components at room temperature).11
The solution was then cooled to 195 K to shift the equilibrium
in favor of the donor–acceptor complex, and one equivalent of
ruthenium terpyridyl Ru-5 was added to bind and block the
pyridyl end of the NDI 3. Finally one equivalent of Fe(BF4)2 (in
CD3OD) was added to coordinate the two terpyridyl ends of the
heteroleptic complex. The [2]catenane 1 was generated in sol-
ution using the same sequence of additions, except that in the
first step one equivalent of the crown ether 4 was added.
The initial target structure ([2]catenane, 1) was assembled
using a “capping” approach (Scheme 2) in which the donor–
acceptor interaction between the electron deficient NDI (3) and
the electron rich crown ether (4) was enhanced by lowering the
temperature.11 Once the pseudo rotaxane was formed at 195 K,
the two ends were capped with a single molecular component
Ru-5 possessing two distinct recognition units. At one end, the
ruthenium porphyrin interacts with the pyridyl group of the
pseudo rotaxane, while addition of Fe(II) salt connects the two
terpyridyl ends of the assembly to form the heteroleptic [2]
The porphyrin–pyridine interaction was crucial in the first step
of the heteroleptic assembly, as confirmed by the addition of one
equivalent of Fe(BF4)2 to a one-to-one mixture of the NDI (3)
and Ru-5 (Scheme 2b). In the absence of any cooperativity one
might expect a statistical 1 : 2 : 1 mixture (3·3, 3·5/5·3, 5·5) of Fe
1
(II)-terpyridyl complexes, however H NMR and HR-MS proved
that the heteroleptic complex 6 was the exclusive product. The
heteroleptic dimer was also observed when the same experiment
was carried out using the weaker coordinating zinc metallated
analogue of 5.
A control experiment was undertaken in order to establish
whether it was possible to observe threading of crown ether 4
after the formation of the heteroleptic macrocycle 6. When a
saturated solution of the crown (4) was added to the pre-
assembled complex, there was no indication of threading of the
crown over a period of six months (Scheme 2b). This finding is
rather surprising since previously reported ruthenium(II) por-
phyrin assemblies were shown to be in thermodynamic
Scheme 1 Synthesis of the components 3 and 5: (a) 2-(2-(2-(2-amino-
ethoxy)ethoxy)ethoxy)ethanol, DMF, μw 5 min, 89%; (b) HOBt, DCC,
nicotinic acid/THF–CH2Cl2 1 : 1, 12 h, 34%; (c) Ag2O, TsCl/CH2Cl2,
RT, 70%; (d) K2CO3/HO-terpy, acetone, 70 °C, 61%; (e) DMF, 70 °C,
40%. Chemical structure and schematic representation of 4 is also
shown.
Scheme 2 (a) Graphical representation of the one-pot assembly of the [2] and [3]catenanes, 1 and 2 respectively. In this sequence of additions the
crown 4 is added to the NDI 3 in the first step, and depending on the stoichiometry of the components, it is possible to bias the equilibrium (at low
temperature) towards the [2] and [3]pseudorotaxane. Addition of 5 leads to coordination of the pyridyl end of 3 to the metalloporphyrin, after which
Fe(BF4)2 (or Zn(TfO)2) coordinates the terpyridyl ends of the complex, leading to 1 and 2. (b) Pre-formation of the heteroleptic complex 6 from a
1 : 1 mixture of 3 and Ru-5 suggests that cooperative effects of the assembly do not allow any of the crown 4 to thread onto the metallomacrocyclic
ring at room temperature.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 3 HR-MS of the [3]catenane solution described in Fig. 2e.
At the final point of the stepwise construction of the assembly,
two sets of peaks for the crown ether were identified at room
temperature (Fig. 2e). One set corresponded to the free crown 4
(an excess was added in the first step) at 6.65, 7.23 and
7.71 ppm, which did not change significantly throughout the
experiment, while a second set of peaks at 6.00, 6.54 and
6.65 ppm were coupled to each other, according to a 1H–1H
COSY experiment, and attributed to those participating in the
donor–acceptor interaction. A third set overlapping with those of
the free crown were assigned to the fraction of crown trapped on
the [3]catenane that did not directly contribute to the donor–
acceptor complex. Accordingly to MM+ molecular modeling,
the metallomacrocycle is flexible and large enough to accommo-
date two bound crown ethers. Significantly, pre-forming the
heteroleptic complex (6) and then adding the crown (4) resulted
in a 1H NMR spectrum with no evidence of bound crown reson-
ances (Fig. 2f).
1
Fig. 2 Partial H NMR spectra (400 MHz, CD2Cl2) of: (a) crown 4 at
298 K; (b) NDI 3 at 298 K; (c) NDI 3 in a saturated solution of crown 4
at 298 K; (d) same sample as in spectrum c, at 195 K; (e) NDI 3 in a
saturated solution of crown 4, with one equivalent of Ru-5 and one
equivalent of Fe(BF4)2 (added as a CD3OD solution) at room tempera-
ture; (f) pre-formation of 6 before the addition of crown 4 (at 298 K).
equilibrium.12 The absence of such an equilibrium indicates that
the heteroleptic terpyridine complex and the Ru-5–pyridine
bond act as a “molecular clamp”.
The stepwise construction of the [3]catenane was monitored
and characterized by H NMR spectroscopy (Fig. 2). The free
1
crown 4 (Fig. 2a) and NDI 3 (Fig. 2b) are shown for comparison,
and their relative chemical shifts do not undergo significant
temperature dependent changes. Addition of 4 to a solution of
NDI 3 resulted in the appearance of diagnostic resonances
indicative of the formation of the donor–acceptor complex at
room temperature (Fig. 2c). For example, the resonance of the
NDI unit broadened and shifted from 8.62 ppm to 8.50 ppm.
Likewise the resonances of the crown 4 were recorded as three
broad peaks at 7.65, 7.10 and 6.44 ppm as a result of fast
exchange on the NMR timescale. At low temperature (Fig. 2d)
any exchange broadening is suppressed as the equilibrium slows
down sufficiently on the NMR timescale to allow identification
of sharp peaks attributed to the bound crown and NDI, with the
excess crown peaks (4) matching those in Fig. 2a. Addition of
one equivalent of Ru-5 led to an ‘open’ heteroleptic assembly
by virtue of the ruthenium porphyrin bound pyridyl resonances
of 3 (Fig. S3†). As the pyridyl protons Hc and Hd experience
the diamagnetic ring current of the porphyrin ring, a significant
change in their chemical shift was observed for Hc (Δδ
−7.11 ppm) and for Hd (Δδ −7.10 ppm), while Ha and Hb
experience less dramatic movement of peaks (Δδ 1.58 and Δδ
−2.05 ppm respectively), as they are further from the distance
dependent effects of the porphyrin ring current. When one
equivalent of Fe(BF4)2 was then added at 195 K, the cyclic
assembly was finally ‘locked’.
When the heteroleptic supramolecular metallomacrocycle is
formed (in 1, 2 or 6), all signals corresponding to the pyridyl
protons split into sets of signals, which by 1H NMR spec-
troscopy suggests that a number of different conformations of
the assembly exist in solution (Fig. S5 and S6†).
Formation of the [2] and [3]catenanes 1 and 2 was also inves-
tigated by DOSY NMR spectroscopy (Fig. S7†) in which the
constituent components were identified with distinct diffusion
coefficients indicative of their relative sizes (in this case 1 and 2
have similar hydrodynamic radii, but the free crown 4 can be
clearly distinguished from 1 and 2). Further evidence for the for-
mation of the heteroleptic assemblies was obtained from HR-MS
(Fig. 3, Fig. S8–S12†), where the heteroleptic complex 6 as well
as the [2] and [3]catenanes (1 and 2) were identified. In contrast,
when a mixture of the heteroleptic macrocycle 6 and the crown
ether 4 (Scheme 2b) was analyzed by High Resolution Mass
Spectrometry (HR-MS), only 6 and 4 were observed, confirming
the structural integrity of 2 resulting from the stepwise assembly
(Scheme 2a).
Threading of crown ether macrocycles into polyethylene
glycol chains (PEG’s) has been reported in the literature.13 The
idea of introducing tetra ethylene glycol chains into our
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem.

View Article Online
H. F. Noller, Science, 1999, 285, 2095; (c) J. D. Watson and
F. H. C. Crick, Nature, 1953, 171, 737.
molecular building blocks was to investigate what effect the
addition of alkali cations, such as Na+ or Li+, will have on the
conformation of the final structures. Furthermore, it has been
reported that addition of Na/Li salts can enhance the donor–
acceptor interaction between crown ether and NDI molecules –
an effect referred to by Sanders et al. as ‘cation reinforcement’.14
However, due to the limited solubility of complexes 1 and 2
after the addition of Na/Li salts it was not possible to analyze
the effects of alkali cations on the assemblies, but we are explor-
ing ways to overcome these limitations to assemble more elabo-
rate heteroleptic multi-component assemblies.
2 (a) E. R. Kay, D. A. Leigh and F. Zerbetto, Angew. Chem., Int. Ed., 2007,
46, 72; (b) A. Coskun, M. Banaszak, R. D. Astumian, J. F. Stoddart and
B. A. Grzybowski, Chem. Soc. Rev., 2012, 41, 19; (c) R. S. Forgan,
J. P. Sauvage and J. F. Stoddart, Chem. Rev., 2011, 111, 5434;
(d) J.-F. Ayme, J. E. Beves, D. A. Leigh, R. T. McBurney, K. Rissanen
and D. Schultz, Nat. Chem., 2012, 4, 15; (e) V. E. Campbell,
X. D. Hatten, N. Delsuc, B. Kauffmann, I. Huc and J. R. Nitschke, Nat.
Chem., 2010, 2, 684; (f) K. Mahata, M. L. Saha and M. Schmittel,
J. Am. Chem. Soc., 2010, 132, 15933; (g) N. Christinat, R. Scopelliti and
K. Severin, Angew. Chem., Int. Ed., 2008, 47, 1848; (h) C. D. Pentecost,
K. S. Chichak, A. J. Peters, G. W. V. Cave, S. J. Cantrill and
J. F. Stoddart, Angew. Chem., Int. Ed., 2007, 46, 218.
3 M. Schmittel and K. Mahata, Angew. Chem., Int. Ed., 2008, 47, 5284.
4 (a) J. E. Beves, B. A. Blight, C. J. Campbell, D. A. Leigh and
R. T. McBurney, Angew. Chem., Int. Ed. Engl., 2011, 50, 9260;
(b) D. J. Mercer and S. J. Loeb, Dalton Trans., 2011, 40, 6385.
5 (a) C. A. Hunter, J. Am. Chem. Soc., 1992, 114, 5303;
(b) A. G. Johnston, D. A. Leigh, L. Nezhat, J. P. Smart and
M. D. Deegan, Angew. Chem., Int. Ed., 1995, 34, 1212;
(c) A. G. Johnston, D. A. Leigh, A. Murphy, J. P. Smart and
M. D. Deegan, J. Am. Chem. Soc., 1996, 118, 10662.
6 (a) S. J. Loeb, Chem. Soc. Rev., 2007, 36, 226; (b) D. J. Mercer and
S. J. Loeb, Dalton Trans., 2011, 40, 6385; (c) J. F. Stoddart and
H. R. Tseng, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 4797.
7 (a) C. Mugemana, P. Guillet, S. Hoeppener, U. S. Schubert, C. A. Fustin
and J. F. Gohy, Chem. Commun., 2010, 46, 1296; (b) E. C. Constable,
Chem. Soc. Rev., 2007, 36, 246.
In summary, this report describes a one-pot sequential syn-
thesis, using three building blocks and three distinct noncovalent
interactions, to construct heteroleptic interlocked supramolecular
metallomacrocycles 1 and 2 as the major or exclusive products.
Current investigations are directed towards understanding the
length dependence of the ethylene glycol chains in the formation
of the assemblies, and incorporating different metalloporphyrins
in order to tune one of the three interactions. We predict that the
use of multiple orthogonal interactions in the construction of
multicomponent assemblies, such as those described in the
present work, will give access to structures of much higher com-
plexity and pave the way towards functional systems.
8 K. M. Mullen, K. D. Johnstone, M. Webb, N. Bampos, J. K. M. Sanders
and M. J. Gunter, Org. Biomol. Chem., 2008, 6, 278.
9 C. J. Bruns, S. Basu and J. F. Stoddart, Tetrahedron Lett., 2010, 51,
983.
10 K. M. Mullen, K. D. Johnstone, D. Nath, N. Bampos, J. K. M. Sanders
and M. J. Gunter, Org. Biomol. Chem., 2009, 7, 293.
11 I. R. Fernando, S. G. Bairu, G. Ramakrishna and G. Mezei,
New. J. Chem., 2010, 34, 2097.
Acknowledgements
We are grateful to the Consejo Nacional de Ciencia y Tecnologia
(CONACYT) Mexico for funding a studentship (MAAG).
12 M. J. Gunter, N. Bampos, K. D. Johnstone and J. K. M. Sanders, New
J. Chem., 2001, 25, 166–173.
13 Y. F. Ng, J. C. Meillon, T. Ryan, A. P. Dominey, A. P. Davis and
J. K. M. Sanders, Angew. Chem., Int. Ed., 2001, 40, 1757.
References
14 S. I. Pascu, T. Jarrosson, C. Naumann, S. Otto, G. Kaiser and
J. K. M. Sanders, New J. Chem., 2005, 29, 80.
1 (a) E. Gouaux and R. MacKinnon, Science, 2005, 310, 1461;
(b) J. H. Cate, M. M. Yusupov, G. Z. Yusupova, T. N. Earnest and
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2012