Pentameric circular iron(II) double helicates and a molecular pentafoil knot

Article abstract of DOI:10.1021/ja303355v

We report on the synthesis of 11 pentameric cyclic helicates formed by imine condensation of alkyl monoamines with a common bis(formylpyridine) bipyridyl-derived building block and iron(II) and chloride ions. The cyclic double-stranded helicates were characterized by NMR spectroscopy, mass spectrometry, and in the case of a 2,4-dimethoxybenzylamine-derived pentameric cyclic helicate, X-ray crystallography. The factors influencing the assembly process (reactant stoichiometry, concentration, solvent, nature and amount of anion) were studied in detail: the role of chloride in the assembly process appears not to be limited to that of a simple template, and larger circular helicates observed with related tris(bipyridine) ligands with different iron salts are not produced with the imine ligands. Using certain chiral amines, pentameric cyclic helices of single handedness could be isolated and the stereochemistry of the helix determined by circular dichroism. By employing a particular diamine, a closed-loop molecular pentafoil knot was prepared. The pentafoil knot was characterized by NMR spectroscopy, mass spectrometry, and X-ray crystallography, confirming the topology and providing insights into the reasons for its formation.

Full text of DOI:10.1021/ja303355v

Article
pubs.acs.org/JACS
Pentameric Circular Iron(II) Double Helicates and a Molecular
Pentafoil Knot
Jean-Franco̧
is Ayme,† Jonathon E. Beves,† David A. Leigh,*,†,‡ Roy T. McBurney,† Kari Rissanen,¶
and David Schultz†
†School of Chemistry, University of Edinburgh, The King’s Buildings, West Mains Road, Edinburgh EH9 3JJ, United Kingdom
‡School of Chemistry, University of Manchester, Oxford Road, Manchester M13 9PL, United Kingdom
¶Department of Chemistry, Nanoscience Center, University of Jyvaskyla, P.O. Box 35, 40014 JYU, Finland
̈
̈
S
* Supporting Information
ABSTRACT: We report on the synthesis of 11 pentameric cyclic helicates formed
by imine condensation of alkyl monoamines with a common bis(formylpyridine)-
bipyridyl-derived building block and iron(II) and chloride ions. The cyclic double-
stranded helicates were characterized by NMR spectroscopy, mass spectrometry,
and in the case of a 2,4-dimethoxybenzylamine-derived pentameric cyclic helicate,
X-ray crystallography. The factors influencing the assembly process (reactant
stoichiometry, concentration, solvent, nature and amount of anion) were studied in
detail: the role of chloride in the assembly process appears not to be limited to that
of a simple template, and larger circular helicates observed with related
tris(bipyridine) ligands with different iron salts are not produced with the imine
ligands. Using certain chiral amines, pentameric cyclic helices of single handedness could be isolated and the stereochemistry of
the helix determined by circular dichroism. By employing a particular diamine, a closed-loop molecular pentafoil knot was
prepared. The pentafoil knot was characterized by NMR spectroscopy, mass spectrometry, and X-ray crystallography, confirming
the topology and providing insights into the reasons for its formation.
prepared from monoamines, and the use of this framework to
form the molecular pentafoil knot. The cyclic helicates are
characterized in solution and the solid state, and the factors
affecting the assembly process investigated. Both similarities
and differences are found between the synthesis of these
circular helicates featuring imine bonds and analogous circular
helicates derived from tris(bipyridine) ligands.
1. INTRODUCTION
The interweaving of molecular strands in DNA,1 proteins,2 and
natural3 and synthetic polymers4 significantly affects their
mechanical, physical, and chemical properties. The presence of
knots in proteins can increase stability and improve function5
and has provided insights into the mechanisms of protein
folding.2a,3 Knots have been tied in biopolymers with optical
tweezers6 and can also be formed from surfactant nanotubes7
and chiral nematic colloids.8 For synthetic chemists the key
challenge in the synthesis of entwined and interlocked
molecular structures is the generation and linking (with the
correct connectivity) of crossing points.9 Template and self-
assembly methods9−14including the use of metal ions,9,10
π−π interactions,11 hydrogen bonding,12 hydrophobic inter-
actions,13 and anions14have proven powerful tools for this
task. Sauvage pioneered the use of linear metal helicates15 to
access mechanically interlocked molecules, entwining ligands
about one, two, or three tetrahedral copper(I) centers to
generate the required crossing points for [2]catenanes,16 trefoil
knots17 and a Solomon link,18 respectively. Despite these and
other metal template syntheses of trefoil knots19 and strategies
based on hydrogen bonding20 and π−π stacking,21 higher-order
non-DNA molecular knots remained elusive until the recently
described22 synthesis of a molecular pentafoil knot, a closed-
loop pentameric cyclic Fe(II) double helicate. Here we report
on the chemistry behind that synthesis, including the assembly
of 11 ‘open’ pentameric cyclic iron(II) double helicates
2. CYCLIC HELICATES AND THEIR POTENTIAL FOR
MOLECULAR KNOT SYNTHESIS
Despite the success in using metal helicates to form interlocked
molecules with up to four crossing points,16−18 attempts to
generate interlocked structures from longer linear helicates
have, to date, proved unsuccessful.23 This is probably due to the
large separation between the end groups of each of the ligand
strands disfavoring their reaction to form structures with the
required connectivity. Cyclic systems might offer a way of
overcoming this problem, since the reactive end-groups can be
brought close together in such architectures. The first circular
helicates24 were discovered by Lehn and co-workers,24a,b,d who
noted their potential for forming complex topologies.24a
However, multiple coupling reactions would be required to
form entwined closed loops, which are inherently more
demanding than single ring-closing reactions. The coupling
Received: April 7, 2012
Published: May 3, 2012
© 2012 American Chemical Society
9488
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
reactions would need to be functional group specific, high
yielding, and ideally, reversible to allow the potential for error
correction of wrongly connected strands during the reaction.
The conditions (ethylene glycol, 170 °C24a,b,d) used to form
circular helicates with Lehn’s original tris(bipyridine) ligands
are not compatible with most types of reactive functional
groups.
In recent years, great strides have been made in dynamic
covalent chemistry.25 Reversible imine-bond formation has
proven highly effective for linking multiple building blocks in
the construction of catenanes,26 rotaxanes,27 cages,28 Solomon
links,29 and Borromean rings.30 We decided to investigate this
method for the formation of cyclic helicates using a ligand
inspired by Lehn’s original three-bipyridine-strand design,24a
replacing the two terminal bipyridine units with 2-formylpyr-
idine groups. Reaction of dialdehyde 1 with amines (to form
imine groups) would generate a tris(bidentate) ligand strand,
which we hoped could be assembled into cyclic double helicate
structures with iron(II) ions.
Figure 1. ESI-MS of cyclic helicate [3aCl](PF6)9 showing signals
corresponding to sequential loss of PF6 anions. Calculated peaks (m/
z): 558.2 [M-7PF6]7+, 675.4 [M-6PF6]6+, 839.5 [M-5PF6]5+, 1085.8
[M-4PF6]4+.
to sequential loss of PF6 anions from a pentameric species
containing five ligands, five iron(II) cations, and one chloride
anion. No chloride-free species were observed, confirming the
binding of a single chloride ion that could not be removed by
repeated anion exchange with KPF6.
3. RESULTS AND DISCUSSION
Dialdehyde 1 was synthesized from commercially available 5-
bromo-2-iodopyridine and 5-bromo-2-formylpyridine via a
series of Sonogashira coupling reactions and deprotection
steps in 19% overall yield, with six steps in the longest linear
sequence (for details, see the Supporting Information [SI]).
Reaction of 1 with 2.2 equiv of 4-methoxylbenzylamine (2a)
in DMSO-d6 and 1.1 equiv of anhydrous iron(II) chloride
(Scheme 1) immediately gave an intensely colored purple
1
The H NMR spectrum of the isolated product contained
only one set of signals for each type of building block (no end-
groups, Figure 2e), consistent with a symmetrical cyclic
structure. Signals for the py-CH2 and N−CH2 methylene
groups appear as diastereotopic pairs which, in combination
with the ESI-MS data, confirmed the product to be a chiral
(racemic) cyclic helicate [3aCl]9+ (Scheme 1). The HA3 signal
(see Scheme 1 for numbering scheme) is shifted to very low
field (9.87 ppm), consistent with strong hydrogen bonding of
these protons to the central chloride ion. Having established
that the reaction of these building blocks afforded pentameric
cyclic helicates, we sought to probe the structural tolerance of
the reaction and investigate factors that control the assembly
process.
Scheme 1. Synthesis of Pentameric Cyclic Helicates [3a−
a
k]Cl(PF6)9
4. STRUCTURAL REQUIREMENTS FOR PENTAMERIC
CYCLIC HELICATE ASSEMBLY
The pentameric cyclic helicate assembly process shown in
Scheme 1 proved to be very sensitive to the nature of the
monoamine building block (Table 1). The effect of having
different substituents on benzylamine was profound: electron-
withdrawing groups [4-bromo-2l (Table 1, entry 12); 3-bromo-
2m (Table 1, entry 13); 4-trifluoromethyl-2n; (Table 1, entry
14)] gave exclusively polymeric or oligomeric products (e.g.,
Figure 2a). In contrast, the use of electron-rich benzylamine
derivatives [4-methoxy-2a (Table 1, entry 1); 2,4-dimethoxy-2b
(Table 1, entry 2)] formed the cyclic helicates in high yields
with few byproducts (d and e of Figure 2). Benzylamine itself
gave a modest yield (30%) of cyclic helicate (Table 1, entry 4)
accompanied by significant amounts of oligomeric byproducts
in the crude reaction mixture.
The steric bulk about the amine functional group also
significantly influences the outcome of the reaction (Table 1).
3-Phenylpropylamine (2f, Table 1, entry 6, 55%) and 4-
phenylbutylamine (2g, Table 1, entry 7, 52%) resulted in higher
cyclic helicate yields than the more hindered 2-phenylethyl-
amine (2e, Table 1, entry 5, 43%). The use of anilines (2o,
Table 1, entry 15) or sterically hindered amines (2p and 2q,
Table 1, entries 16 and 17) initially gave intensely purple-
colored solutions apparently consisting of polymeric species
a
Reaction conditions: 1. Dialdehyde 1/FeCl2/amine 2a−q (1:1.1:2.2),
DMSO-d6, 60 °C, 1−2 days. 2. Aqueous KPF6.
solution, typical of low-spin iron(II) tris(diimine) complexes.31
The H NMR spectrum (DMSO-d6) of the initial reaction
1
mixture contained only broad signals, indicative of the
formation of poorly defined oligomeric and polymeric species.
However, on heating at 60 °C, the spectrum gradually
simplified until after 15 h a single major species was present
in solution. Anion exchange was performed using excess
aqueous potassium hexafluorophosphate (KPF6) to give a fine
purple suspension, which was collected, washed, and dissolved
in acetonitrile, to give [3aCl](PF6)9 in 54% isolated yield.
Electrospray mass spectrometry (ESI-MS, Figure 1) showed 3+,
4+, 5+, 6+, and 7+ signals with isotope patterns corresponding
9489
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
Figure 2. 1H NMR (CD3CN, 500 MHz, 298 K) of the products isolated from the reaction of dialdehyde 1 and various amines (2l, 2f, 2e, 2b, and
1
2a) in the presence of FeCl2 (Scheme 1). (a) Amine used 4-bromobenzylamine (2l; Table 1, entry 12). The broad H NMR spectrum is
characteristic of polymers. (b) [3fCl](PF6)9. (c) [3eCl](PF6)9. (d) [3bCl](PF6)9. (e) [3aCl](PF6)9. All samples were colored purple of similar
intensities. The assignments correspond to the labeling shown in Scheme 1.
1
(broad H NMR spectra). However, upon being heated, pale-
orange high-spin iron(II) complexes were formed, with no ESI-
MS evidence for the formation of short linear or cyclic
helicates. Sterically unhindered aliphatic primary amines (e.g.,
hexylamine 2h and dodecylamine 2i, Table 1, entries 8 and 9)
gave the highest yields of cyclic helicates (63% for [3hCl]-
(PF6)9) and the fewest byproducts.
of FeCl2 (with respect to 1) giving 84, 85, and 91% yields of
[3hCl](PF6)9, respectively (Figure 3).
5.2. Influence of Anions. The substitution of FeCl2 for
other iron(II) salts [Fe(SO4)2·7H2O, Fe(CH3CO2)2, Fe-
(BF4)2·6H2O, Fe(ClO4)2·6H2O, Fe(ClO4)2·xH2O] in the
reactions shown in Scheme 1 invariably gave complex mixtures
with no ESI-MS evidence for the formation of circular helicates
of any size, in contrast to the cyclic hexameric helicate reported
by Lehn using his related tris(bipyridine) ligand with
Fe(BF4)2·6H2O or FeBr2.24b However, when 4-methylbenzyl-
amine (2c) was employed with iron(II) tetrafluoroborate
[Fe(BF4)2·6H2O] or iron(II) perchlorate [Fe(ClO4)2·6H2O] a
discrete low-molecular weight product was formed in under
10% yield; no related products were observed with hexylamine
(2h) or other aliphatic amines. ESI-MS evidence supported the
assignment of this minor product as the linear trinuclear triple-
stranded helicate24a [Fe3L3]6+ (see SI, page S22).
5. PROBING THE REACTION CONDITIONS OF
PENTAMERIC CYCLIC HELICATE ASSEMBLY
Various factors (reagent stoichiometry, influence of various
anions, the role of chloride, the effect of solvent, concentration,
and preforming the diimine ligand) influencing the outcome of
the pentameric cyclic helicate-forming reaction shown in
Scheme 1 were investigated, mainly through a series of studies
using hexylamine (2h) as the amine (Scheme 2).
5.3. Role of Chloride Ions. The role of chloride in the
pentameric cyclic helicate forming reaction shown in Scheme 2
was investigated using varying amounts of Fe(BF4)2·6H2O and
FeCl2 in order to vary the amount of chloride present in the
reaction while maintaining a constant iron(II) concentration
(Figure 4, blue data points). The yield of the isolated cyclic
pentameric helicate [3hCl]9+ increased steeply with the amount
of chloride, up to 0.4 chloride ions per iron(II) (i.e.,
approximately two chloride ions per cyclic helicate, 44%
yield), and then increased less sharply as the amount of
chloride increased further up to 63% with 2.0 Cl− per Fe(II).
Similar results were obtained using Fe(BF4)2·6H2O as the
sole iron(II) source and adding tetrabutylammonium chloride32
(Figure 4, red data points, and SI, Figure S15). Under these
conditions chloride/iron(II) ratios greater than 2:1 could be
5.1. Reactant Stoichiometry. Reactant stoichiometry
proved to be extremely important, with a ratio of 1:1.1:2.2
dialdehyde 1/FeCl2/2h giving 63% of cyclic helicate [3h]Cl-
(PF6)9 (Table 1, entry 8). The use of less than 2 equiv of amine
2h (with respect to 1) gave mixtures of products, including
[Fe(1)3]6+ and related species (see SI, Figure S10).
Substoichiometric amounts of FeCl2 (with respect to 1) gave
low yields of circular helicate and complex mixtures of products
(SI, Figure S9). Although employing an excess of amine also
resulted in a substantial decrease in yield (SI, Figure S10), the
lack of other high-molecular weight byproducts meant that the
cyclic helicate products were most easily isolated using these
conditions. Somewhat surprisingly, the use of excess FeCl2 gave
the highest yields of cyclic helicate with 1.5, 2.2, and 4.4 equiv
9490
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
Table 1. Yields of Pentameric Cyclic Helicates from the
Reaction Shown in Scheme 1 with Monoamines 2a−q
Scheme 2. Synthesis of Pentameric Cyclic Helicate
a
[3h]Cl(PF6)9 from Dialdehyde 1 and Hexylamine (2h) or
a
Preformed Diimine Ligand 4
a
Reaction conditions: 1. Dialdehyde 1/FeCl2/2h (1:1.1:2.2) or 4/
FeCl2 (1:1.1), DMSO-d6, 60 °C, 1 day. 2. Aqueous KPF6.
a
Reaction conditions: 1.1:1.1:2.2 of 1/FeCl2/amine, DMSO-d6, 60 °C,
b
1 d (except entry 10, 2 d). 2. Aqueous KPF6. Reaction performed in
1:1 CDCl3/CD3CN due to the poor solubility of 2i in DMSO. Yield
of major diastereoisomer. Intensely colored purple solutions with
only broad H NMR signals, likely polymeric materials. Initially gave
an intensely colored purple solution, which became pale yellow upon
c
d
e
1
Figure 3. Influence of FeCl2/dialdehyde 1 ratio [2.2 equiv of
hexylamine (2h)] on the yield of [3hCl](PF6)9. Error bars 5%.
f
heating overnight. Pale-yellow/orange solutions, assumed to be high-
spin iron(II) complexes, with no ESI-MS evidence for cyclic helicate
formation.
chloride ions play a more complicated role in the assembly
process. It seems likely that they aid the rearrangement of
oligomers in the reaction mixture until the stable chloride-
binding pentameric cyclic helicate is formed.34
obtained, which resulted in slight decreases in yield (e.g., 4:1
Cl/Fe gave 60% yield of [3hCl]9+). Interestingly, the use of
higher ratios with respect to 1 of both Fe(II) and Cl− resulted
in significant increases in yield. For example, where 4.4 equiv of
Fe(BF4)2·6H2O was used (with respect to 1) the yield of
[3hCl]9+ increased linearly with chloride concentration, with
90% yield being obtained where 4 equiv of chloride (with
respect to dialdehyde 1) was added (Figure 4, black data
points). A similar yield was obtained where 4.4 equiv of FeCl2
was used (see section 5.1, Reactant Stoichiometry).
5.4. Solvent and Concentration. DMSO proved to be the
solvent of choice for the pentameric cyclic helicate-forming
reactions for most amines, with methanol, ethanol, nitro-
methane, acetonitrile, diglyme (MeOCH2CH2OMe), and
mixtures of these solvents with halogenated solvents either
giving precipitates or pale-yellow solutions indicative of high-
spin iron(II) complexes. Dodecylamine (2i) proved an
exception,35 with helicate [3iCl]9+ being formed in good
yield in acetonitrile−chloroform (1:1) solution (Table 1, entry
9), presumably due to the solubilizing effect of the long alkyl
chains. Generally, heating the reaction mixtures at higher
temperatures (e.g., 100 °C in DMSO) resulted in discoloration
over several hours, indicating decomposition of the putative
Fe(II)−diimine complexes.36 Reaction concentration is also
important,24b with relatively high initial concentrations of
The pentameric cyclic helicates bind Cl− ions extremely
strongly (Ka > 1010 M−1) in a 1:1 complex.22a The finding that
one equivalent of chloride is not generally sufficient to
effectively template the formation of the circular pentameric
helicates at millimolar concentrations appears to rule out a
simple thermodynamic template effect33,9 and indicates that the
9491
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
formation was more rapid (Figure 5, pink data points),
although some imine hydrolysis was also observed. A
comparison of the rate of formation of [3hCl]9+ when starting
from 4 or 1/2h confirms that metal−ligand exchange is likely
the rate-limiting factor in the rearrangement of rapidly formed
simple complexes (e.g., Fe3(4)3 and related species), oligomers,
and polymers into the pentameric cyclic helicate.
6. X-RAY CRYSTAL STRUCTURE OF PENTAMERIC
CYCLIC HELICATE [3BCL](PF6)9
Purple crystals of pentameric circular helicate [3bCl](PF6)9
suitable for single-crystal X-ray diffraction were grown by slow
diffusion of diethyl ether vapor into a solution of the complex
in nitromethane−acetonitrile. The structure crystallized in the
P21/c space group with the asymmetric unit containing a single
pentameric cyclic helicate (Figure 6).37 The five iron(II)
Figure 4. Effect of chloride ions on the yield of helicate [3hCl]9+. The
reactions were carried out using mixtures of either Fe(BF4)2·6H2O and
FeCl2 (blue data points), or Fe(BF4)2·6H2O and tetrabutylammonium
chloride (red data points = 1.1 equiv of Fe(II); black data points = 4.4
total equiv of Fe(II)), with Bu4NBF4 added to maintain a constant
Bu4N+ concentration. In all cases the ratio of dialdehyde (1)/
hexylamine (2h) was 1:1.1. The highest yield, 90%, was obtained using
4.4 equiv of Fe(BF4)2·6H2O (with respect to dialdehyde 1) with 4
equiv of Bu4NCl (i.e., 0.9 equiv of Cl− per Fe(II); black line). Error
bars 5%. See SI, Figure S14.
reactants (∼5 mM) delivering the highest yields of pentameric
cyclic helicate products. Very dilute reaction conditions (<0.01
mM) gave low yields of circular helicates.
5.5. Rate of Pentameric Circular Helicate Formation.
The rate of formation of pentameric cyclic helicate [3hCl]9+
(Scheme 2) was followed by 1H NMR spectroscopy (Figure 5,
Figure 6. X-ray crystal structure of [3bCl](PF6)9·xsolvent, from a
single crystal obtained by vapor diffusion of diethyl ether into a
acetonitrile−nitromethane solution.37 Anions, solvent molecules, and
hydrogen atoms are omitted for clarity. Nitrogen atoms are shown in
dark blue, oxygen atoms red, chlorine atom green, and carbon atoms
gray; the carbon framework of a single ligand is colored light blue.
Fe(II) centers shown at 50% van der Waals radius, Cl− shown at
100%. Cl−···HC distances (clockwise from top-left blue bipyridine
unit): 2.75, 2.68, 2.69, 2.81, 2.78, 2.68, 2.71, 2.71, 2.73, 2.78 Å. C−H−
Cl angles (deg) 176.3, 177.5, 175.5, 176.7, 178.4, 175.8, 178.1, 179.4,
174.0, 176.8. For packing diagrams, see SI, Figures S24 and S25.
Figure 5. Rate of formation of pentameric cyclic helicate [3hCl]9+
(Scheme 2) from dialdehyde 1 and hexylamine (2h) (pink squares) or
from preformed ligand 4 (red squares). The rate of formation of
pentafoil knot [7Cl]9+ from dialdehyde 1 and diamine 6 (Scheme 3) is
also shown (blue diamonds). The reactions were maintained at 60 °C
1
and H NMR spectra collected (600 MHz, DMSO-d6) every 30 min.
Relative peak integral of imine peak shown, normalized to the integral
of the final time-point in each case. See SI, Figures S7 and S8.
centers lie in a near-perfect plane (maximum displacement of a
Fe(II) from the least-squares plane is 0.091(1) Å). The
coordination geometry of the iron(II) ions shows Fe−N and
N−Fe−N bond lengths and angles at the limits of normal
ranges and, as for Lehn’s pentameric cyclic helicate,24a the
overall geometries of the metal centers are among the most
distorted for [Fe(pyridine)6] structures in the Cambridge
Structural Database (CSD) (see SI). The chloride ion is in
close contact with the 10 HA3 protons (Cl−···HA3C distances
2.679(1)−2.814(1) Å) and displaced from the least-squares
plane defined by the five Fe(II) centers by 1.475(1) Å,
indicating that the symmetry of the solution 1H NMR spectrum
must result from oscillation of the anion between the two sides
red data points). Under the standard reaction conditions (1/
FeCl2/2h (1:1.1:2.2), DMSO-d6, 60 °C) the reaction was
essentially complete after 24 h (after which time there was little
1
change in the H NMR spectrum). In order to simplify the
study of the assembly process, ligand 4 was prepared to both
eliminate initial imine formation from the reaction kinetics and
to allow a strict 1:1 ratio of amine/aldehyde to be employed
(see SI, Figure S7).
When employing 4 instead of 1 and 2h in the reaction shown
in Scheme 2, the rate of pentameric cyclic helicate ([3hCl]9+)
9492
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
Figure 7. 1H NMR (500 MHz, CD3CN, 298 K) of pentameric cyclic helicates derived from (a) hexylamine ([3hCl](PF6)9), (b) diastereoselective
helicate [3jCl](PF6)9, (c) mixture of diastereomers formed with (R)-amino-2,3-propandiol (2k) (This spectrum is after partial workup and has been
enriched in the minor diastereomer. The crude reaction mixture shows a ∼1:2 ratio of diastereomers.), and (d) isolated major diastereoisomer
([3kCl](PF6)9) from (R)-amino-2,3-propandiol (2k). The assignments correspond to the labeling shown in Scheme 1.
of the circular helicate that is fast on the NMR time scale. The
phenyl rings adopt a range of orientations, engaging in weak
π−π stacking interactions to form zigzag 2D sheets of helicates
in the ab-plane (see SI), with weaker interactions between these
sheets.
11, and Figure 7c). The major diastereomer was isolated in 46%
yield by recrystallization from acetonitrile−water (the minor
diastereoisomer has significantly higher solubility).
Finally, in contrast to the unsuccessful reactions with other
α-substituted primary amines (Table 1, entries 16 and 17, and
SI, Figure S5), enantiopure (R)- or (S)-2-amino-1-propanol
(2j) generated pentameric cyclic helicate [3jCl]9+ with
complete diastereoselectivity (the enantiomers of the amine-
producing pentameric cyclic helicates of opposite helix
7. CONTROLING THE HELIX STEREOCHEMISTRY OF
PENTAMERIC CYCLIC HELICATES
1
We next focused on attempting to form pentameric cyclic
helicates of single handedness. Many examples of stereo-
chemical control of linear helicate formation have been
reported,38 and one example of a single-stranded circular
helicate,38d usually by employing chiral auxiliaries covalently
attached to the ends of the ligands or by incorporating chiral
structures such as BINOL (1,1′-bi-2-naphthol) or trans-1,2-
diaminocyclohexane into the center regions of the ligands. The
aldehyde groups at either end of building block 1 enables the
straightforward preparation of enantiopure ligands by con-
densation reactions with appropriate chiral amines.38m
Initially, we investigated chiral derivatives closely related to
those of monoamines found to form pentameric cyclic helicates
in reasonable yields, i.e. amines 2p and 2q (Table 1, entries 16
and 17). Disappointingly, however, both amines generated only
pale yellow reaction mixtures (indicative of high-spin iron(II)
complexes) and no discrete complexes could be isolated
following anion exchange. This also proved to be the case for a
range of other chiral α-substituted amines (see SI, Figure S5).
We speculated that the methyl group adjacent to the amine
might be too sterically demanding to allow helicate formation
(see section 4, Structural Requirements for Pentameric Cyclic
Helicate Assembly) and tried a primary amine with a chiral
hydroxyl group two atoms away from the amine center ((R)-
2k). This formed both diastereoisomers of pentameric cyclic
helicate [3kCl]9+ with ∼1:2 diastereoselectivity (Table 1, entry
stereochemistry), in isolated yields of 32%. The H NMR
spectra showed a large and unexpected peak shift for the HA6
signal, shifted downfield by 0.63 ppm relative to the equivalent
signal of helicate [3hCl](PF6)9. A ROESY NMR spectrum was
consistent with the presence of a CH···O hydrogen bond
between the OH group and HA6 of the bipyridine unit (SI,
Figure S18) which likely plays a significant stabilizing role in
formation of the pentameric cyclic helicate from this hindered
amine.
The helix stereochemistry of the pentameric cyclic helicates
formed from the chiral amines was investigated by circular
dichroism (CD). The UV−visible absorption spectrum of
helicate [3jCl](PF6)9 is shown in Figure 8 and is representative
of helicates [3a−kCl](PF6)9. An intense and broad MLCT
transition is centered at 566 nm, with a shoulder at higher
energy (centered at 520 nm)39 and a strong π−π* transition at
310 nm, typical of low-spin [Fe(diimine)3]2+ complexes. The
CD spectra of (R)-[3jCl](PF6)9 and (S)-[3jCl](PF6)9 are
shown in Figure 8 and are qualitatively similar to those of
related systems.40 The interpretation of CD spectra of
octahedral coordination complexes has been shown to be a
reliable means of assigning absolute stereochemistry of
M(N∧N)3 (N∧N = 2,2′-bipyridine, 1,10-phenanthroline)-type
complexes using exciton theory.41 Using this method, the CD
spectra of [(R)-3jCl]9+ and [(S)-3jCl]9+ were assigned as
ΔΔΔΔΔ-[3jCl]9+ and ΛΛΛΛΛ-[3jCl]9+, respectively, in
9493
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
Figure 8. UV−vis absorption (black, left axis) and circular dichroism
of pentameric cyclic helicates [(R)-3jCl](PF6)9 and [(S)-3jCl](PF6)9
in CH3CN (0.12 mM).
1
Figure 9. H NMR spectra (DMSO-d6, 500 MHz, 298 K) of the
formation of pentafoil knot [7Cl]9+ (Scheme 3). Reaction mixture
after (e) 5 min, (d) 2 h, (c) 10 h, (b) 26 h, and (a) 48 h at 60 °C. The
assignments correspond to the labeling shown in Scheme 3.
agreement with related assignments (SI, pages S38−S40).42
The isolated major diastereomer [(R)-3kCl]9+ has a CD
spectrum similar to that of [(S)-3jCl]9+, corresponding to
ΛΛΛΛΛ-[3kCl]9+, and in agreement with NMR ROESY data
(SI, Figure S18).
mixture slowly rearranged over two days to form a single major
species (Figure 9a). Aspects of the 1H NMR spectrum are very
similar to that of the hexylamine derivative [3hCl](PF6)9,
suggesting that these two compounds have closely related
structures. After workup, the molecular pentafoil knot [7Cl]-
(PF6)9 was isolated in 44% yield and its structure confirmed by
ESI-MS (see SI, page S21).22
8. ASSEMBLY OF A MOLECULAR PENTAFOIL KNOT
Having established that nonhindered, aliphatic monoamines
formed pentameric circular helicates in the highest yields with
the fewest byproducts, we reasoned that the use of alkyl
diamines could allow the formation of a molecular pentafoil
knot (Scheme 3).22 Somewhat unexpectedly, the use of C6−C12
Some aspects of the assembly of the open cyclic helicates
[3a−kCl]9+ translate poorly to the pentafoil knot [7Cl]9+
assembly process. First, the rate of formation of the knot is
significantly slower than that of the open cyclic helicates
(Figure 5), indicating that interconversions and rearrangements
of the intermediates are slower with the diamine. Second, the
stoichiometry of reagents proved to be more critical for the
pentafoil knot than for the open circular helicate systems. The
use of increasing amounts of FeCl2 relative to dialdehyde 1
resulted in lower yields (e.g., the use of 2 equiv of FeCl2 with
respect to dialdehyde 1 resulted in a 40% drop in yield relative
to when 1.1 equiv of FeCl2 was used; SI, Figure S12). This is in
contrast with the findings for the open systems where increased
ratios of FeCl2 substantially increased yields (see section 5.1.
Reactant Stoichiometry). The use of excess diamine in the
knot-forming reaction resulted in even more severe reductions
in yield (e.g., 2.0 equiv of diamine 6 resulted in a 90% fall in
relative yield, SI, Figure S11). Finally, the reaction also proved
very sensitive to the ratio of FeCl2/amine employed, with a 1:1
ratio giving the highest yields (SI, Figure S13). As with the
open systems, the formation of pentafoil knot [7Cl]9+ was
sensitive to the amount of chloride present, with the yield
increasing to a maximum with a ratio of 2:1 chloride/iron(II)
(SI, Figure S15). The requirement for more than one Cl− ion
per molecular pentafoil knot despite very strong knot/chloride
1:1 binding22 suggests that the role of Cl− in the assembly
process is more complicated than just as a simple template.
Higher amounts of chloride ion in the reaction did not increase
the yield of pentafoil knot further (SI, Figures S15 and S16).
Purple crystals of pentafoil knot [7Cl](PF6)9·x(solvent) were
grown by slow diffusion of diethyl ether vapor into an
acetonitrile−toluene (3:2) solution of the complex and the X-
ray structure determined using diffraction data collected on
beamline I19 at the Diamond Light Source (U.K.).22 The solid
state structure (Figure 10) confirmed the topology of the
pentafoil knot. As for the open cyclic helicate [3bCl](PF6)9
Scheme 3. Synthesis of Molecular Pentafoil Knot
a
[7Cl](PF6)9.
a
Reaction conditions: 1. Dialdehyde 1/FeCl2/diamine 6 (1:1.1:2.2),
DMSO-d6, 60 °C, 2 days. 2. Aqueous KPF6. 44% yield [7Cl](PF6)9.
The use of diamines 5a−e instead of 6 did not generate the
corresponding pentafoil knots.
diaminoalkanes (5a−e) gave only intractable material, with no
evidence for the formation of the desired knots. Similarly,
diamines based on bridged 4-(alkoxy)benzylamines gave only
oligomeric/polymeric products (see SI, Figure S5). When
employing a glycol linker, 2,2′-(ethylenedioxy)bis(ethylamine)
1
(6), the reaction mixture also initially gave a featureless H
NMR spectrum (Figure 9e). However, on heating at 60 °C, the
9494
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
Figure 10. X-ray crystal structure of pentafoil knot [7Cl](PF6)9, from
a single crystal obtained by vapor diffusion of diethyl ether into a
acetonitrile−toluene solution.22 Hexafluorophosphate anions, solvent
molecules, and hydrogen atoms are omitted for clarity. Nitrogen atoms
are shown in dark blue, oxygen atoms red, chlorine atom green, and
carbon atoms gray (the carbon framework originating from a single
building block of 1 is colored light blue). Fe(II) centers shown at 50%
van der Waals radius, Cl− shown at 100%. Cl···HC distances
(clockwise from top left blue bipyridine unit): 2.71, 2.75, 2.76, 2.70,
2.76, 2.70, 2.69, 2.71, 2.71, 2.69. C−H−Cl− angles (deg): 172, 179,
170, 176, 172, 177, 176, 176, 170, 178. For an ORTEP plot, see the SI,
Figure S26.
Figure 11. Crystal packing diagrams for the X-ray structure of
[7Cl](PF6)9·xsolvent viewed down (a) the crystallographic c-axis and
(b) the crystallographic a-axis. Close contacts between glycol chains
dominate the intermolecular contacts, forming 2D sheets of the
molecular knots that are separated by disordered solvent and anions
(omitted for clarity).
The reagent stoichiometry that corresponds to the
composition of the pentameric circular helicate products
[5:5:10:1 dialdehyde/Fe(II)/amine/Cl−] does not give the
highest yield. Rather the use of significant excesses of iron(II)
and chloride ions give the highest (virtually quantitative) yields
of open circular helicates, indicating that these ions likely play
more important roles in the assembly process than that of a
simple thermodynamic template.
Due to the symmetry of the cyclic helicate structure, the
effect of individually rather weak interactions on the assembly
process is magnified, and can significantly alter the product
distribution in the reaction mixture. In some cases this can be
beneficial (for example, to form cyclic helicates of specific
handedness with particular chiral amines) and in others
detrimental (for example, sterically hindered amines favoring
polymer formation over circular helicates).
(Figure 6), the structure contains a single helicate in the
asymmetric unit and crystallizes in the P21/c space group. The
planar geometry of the iron centers and the displacement of the
chloride ion from this plane are similar to that in [3bCl]9+.
Eight of the nine PF6 anions are located in the difference map,
and the structure also contains disordered solvent molecules.
The structure is generally well-ordered (see SI, Figure S26, for
ORTEP plot), with the exception of the glycol linkers which
show large thermal displacement parameters. The glycol chains
have O−C−C−O torsion angles ranging from 57(3)° to 77(2)
°. It seems likely that the gauche effect43 (the preference for
O−C−C−O chains to adopt torsion angles of 60° rather than
the 180° angle of all-carbon chains) stabilizes this low-energy
turn which is likely responsible for diamine 6 forming the
closed-loop pentafoil knot, whereas other diaminoalkanes of
similar length (5a−e) do not (Scheme 3).44 The molecules are
packed in 2D zigzag arrays, with close contacts between the
glycol linkers (CH···HC contacts as short as 2.22 Å, Figure 11).
The subtle effects of structure on the assembly reaction is
also illustrated by the fact that circular helices other than the
pentamer are not observed in the reaction of 1 with amines and
different iron(II) salts. This contrasts with the formation of
hexameric circular helicates with Lehn’s tris(bipyridine) ligands
and Fe(BF4)2·6H2O or FeBr2.24b
9. CONCLUSIONS
Pentameric cyclic double helicates can be efficiently assembled
in a one-pot reaction from iron(II) cations, dialdehyde 1, and a
range of monoamines (2a−k) in the presence of chloride
anions. The assembly process is very sensitive to the structure
of the amine, reactant stoichiometry (dialdehyde/Fe(II)/
amine:Cl−) and reaction conditions (solvent, concentration).
Some of the ways that these factors influence the assembly
process are somewhat unexpected:
By using a diamine that can form a low-energy turn due to
the gauche effect on its glycol linkers the assembly system was
successfully used to form a molecular pentafoil knot.22 The
reagent stoichiometry effects observed for open circular helicate
formation do not all translate to the pentafoil knot. For
example, the use of an increased amount of iron(II) with
respect to 1 lowered the yield of the knot, the opposite of that
found for the open circular helicates.
9495
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
Dietrich-Buchecker, C.; Hamann, C.; Jouvenot, D.; Kern, J.-M.;
Mobian, P.; Sauvage, J.-P. In Comprehensive Coordination Chemistry II;
McCleverty, J. A., Meyer, T. J., Eds.; Elsevier: Amsterdam, 2004; Vol.
7, pp 303−326.
(11) (a) Loeb, S. J. Chem. Soc. Rev. 2007, 36, 226−235. (b) Griffiths,
K. E.; Stoddart, J. F. Pure Appl. Chem. 2008, 80, 485−506.
(12) (a) Cantrill, S. J.; Pease, A. R.; Stoddart, J. F. J. Chem. Soc.,
Dalton Trans. 2000, 3715−3734. (b) Kay, E. R.; Leigh, D. A. Top.
Curr. Chem. 2005, 262, 133−177.
(13) (a) Fujita, M. Acc. Chem. Res. 1999, 32, 53−61. (b) Frampton,
M. J.; Anderson, H. L. Angew. Chem., Int. Ed. 2007, 46, 1028−1064.
(14) Mullen, K. M.; Beer, P. D. Chem. Soc. Rev. 2009, 38, 1701−1713.
(15) Albrecht, M. Top. Curr. Chem. 2004, 248, 105−139.
(16) Dietrich-Buchecker, C. O.; Sauvage, J.-P.; Kintzinger, J.-P.
Tetrahedron Lett. 1983, 24, 5095−5098.
(17) (a) Dietrich-Buchecker, C. O.; Sauvage, J.-P. Angew. Chem., Int.
Ed. Engl. 1989, 28, 189−192. (b) Rapenne, G.; Dietrich-Buchecker, C.;
Sauvage, J.-P. J. Am. Chem. Soc. 1996, 118, 10932−10933. (c) Meyer,
M.; Albrecht-Gary, A.-M.; Dietrich-Buchecker, C. O.; Sauvage, J.-P. J.
Am. Chem. Soc. 1997, 119, 4599−4607.
(18) Nierengarten, J. F.; Dietrich-Buchecker, C. O.; Sauvage, J. P. J.
Am. Chem. Soc. 1994, 116, 375−376.
The use of cyclic helicates as scaffolds for the assembly of
interlocked molecules builds upon, and complements, the linear
helicate strategy introduced by Sauvage in the 1980s.10 We
anticipate that the information gleaned from the investigation
of the assembly processes presented here will be useful for the
rational synthesis of higher-order topologically complex
molecular architectures.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures and spectral data for all compounds
and the details of the X-ray analyses of [3bCl](PF6)9·xsolvent
and [7Cl](PF6)9·xsolvent including CIF files. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
David.Leigh@manchester.ac.uk
■
Notes
The authors declare no competing financial interest.
(19) (a) Rapenne, G.; Dietrich-Buchecker, C.; Sauvage, J.-P. J. Am.
Chem. Soc. 1999, 121, 994−1001. (b) Adams, H.; Ashworth, E.;
Breault, G. A.; Guo, J.; Hunter, C. A.; Mayers, P. C. Nature 2001, 411,
763−763. (c) Guo, J.; Mayers, P. C.; Breault, G. A.; Hunter, C. A.
Nature Chem. 2010, 2, 218−222. (d) Barran, P. E.; Cole, H. L.;
Goldup, S. M.; Leigh, D. A.; McGonigal, P. R.; Symes, M. D.; Wu, J.;
Zengerle, M. Angew. Chem., Int. Ed. 2011, 50, 12280−12284.
ACKNOWLEDGMENTS
■
We thank the Diamond Light Source (U.K.) for synchrotron
beamtime on I19 (XR029), the Engineering and Physical
Sciences Research Council (EPSRC) National Crystallography
Service for data collection, and the EPSRC National Mass
Spectrometry Service Centre (Swansea, U.K.) for high-
resolution mass spectrometry. J.E.B. and D.S. are Swiss
National Science Foundation postdoctoral fellows. This
research was funded by the EPSRC and the Academy of
Finland (K.R., Projects 212588 and 218325). We thank Robert
W. McGregor (www.mcgregorfineart.com) for the images of
the X-ray crystal structures.
(20) (a) Safarowsky, O.; Nieger, M.; Frohlich, R.; Vogtle, F. Angew.
̈
̈
Chem., Int. Ed. 2000, 39, 1616−1618. (b) Vogtle, F.; Hunten, A.;
̈
Vogel, E.; Buschbeck, S.; Safarowsky, O.; Recker, J.; Parham, A.-H.;
Knott, M.; Muller, W. M.; Muller, U.; Okamoto, Y.; Kubota, T.;
Lindner, W.; Francotte, E.; Grimme, S. Angew. Chem., Int. Ed. 2001,
40, 2468−2471. (c) Lukin, O.; Kubota, T.; Okamoto, Y.; Kaufmann,
A.; Vogtle, F. Chem.Eur. J. 2004, 10, 2804−2810. (d) Lukin, O.;
̈
Vogtle, F. Angew. Chem., Int. Ed. 2005, 44, 1456−1477. (e) Feigel, M.;
̈
Ladberg, R.; Engels; Herbst-Irmer, R.; Frohlich, R. Angew. Chem., Int.
̈
REFERENCES
Ed. 2006, 45, 5698−5702. (f) Passaniti, P.; Ceroni, P.; Balzani, V.;
■
Lukin, O.; Yoneva, A.; Vogtle, F. Chem.Eur. J. 2006, 12, 5685−5690.
̈
(1) (a) Wasserman, S. A.; Cozzarelli, N. R. Science 1986, 232, 951−
960. (b) Du, S. M.; Seeman, N. C. J. Am. Chem. Soc. 1992, 114, 9652−
9655. (c) Du, S. M.; Stollar, B. D.; Seeman, N. C. J. Am. Chem. Soc.
1995, 117, 1194−1200.
(2) (a) Taylor, W. R. Nature 2000, 406, 916−919. (b) King, N. P.;
Jacobitz, A. W.; Sawaya, M. R.; Goldschmidt, L.; Yeates, T. O. Proc.
Natl. Acad. Sci. U.S.A. 2010, 107, 20732−20737.
(g) Bruggemann, J.; Bitter, S.; Muller, S.; Muller, W. M.; Muller, U.;
̈
̈
̈
̈
Maier, N. M.; Lindner, W.; Vogtle, F. Angew. Chem., Int. Ed. 2007, 46,
254−259.
̈
(21) Ashton, P. R.; Matthews, O. A.; Menzer, S.; Raymo, F. M.;
Spencer, N.; Stoddart, J. F.; Williams, D. J. Liebigs Ann./Recl. 1997,
2485−2494.
(22) (a) Ayme, J.-A.; Beves, J. E.; Leigh, D. A.; McBurney, R. T.;
Rissanen, K.; Schultz, D. Nature Chem. 2012, 4, 15−20. (b) Hardie, M.
J. Nature Chem. 2012, 4, 7−8.
(3) (a) Andersson, F. I.; Pina, D. G.; Mallam, A. L.; Blaser, G.;
Jackson, S. E. FEBS 2009, 276, 2625−2635. (b) Mallam, A. L.; Rogers,
J. M.; Jackson, S. E. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 8189−8194.
(c) Meluzzi, D.; Smith, D. E.; Arya, G. Annu. Rev. Biophys. 2010, 39,
349−366.
(4) (a) Saitta, A. M.; Soper, P. D.; Wasserman, E.; Klein, M. L.
Nature 1999, 399, 46−48. (b) Virnau, P.; Kantor, Y.; Kardar, M. J. Am.
Chem. Soc. 2005, 127, 15102−15106.
(5) Belmonte, A. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 17243−
17244.
(23) Dietrich-Buchecker, C.; Colasson, B.; Jouvenot, D.; Sauvage, J.-
P. Chem.Eur. J. 2005, 11, 4374−4386.
(24) (a) Hasenknopf, B.; Lehn, J.-M.; Kneisel, B. O.; Baum, G.;
Fenske, D. Angew. Chem., Int. Ed. Engl. 1996, 35, 1838−1840.
(b) Hasenknopf, B.; Lehn, J.-M.; Boumediene, N.; Dupont-Gervais, A.;
Van Dorsselaer, A.; Kneisel, B.; Fenske, D. J. Am. Chem. Soc. 1997,
119, 10956−10962. (c) Provent, C.; Hewage, S.; Brand, G.;
̀
Bernardinelli, G.; Charbonniere, L. J.; Williams, A. F. Angew. Chem.,
(6) Arai, Y.; Yasuda, R.; Akashi, K.-I.; Harada, Y.; Miyata, H.;
Kinosita, K., Jr.; Itoh, H. Nature 1999, 399, 446−448.
(7) Lobovkina, T.; Dommersnes, P.; Joanny, J.-F.; Bassereau, P.;
Karlsson, M.; Orwar, O. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 7949−
7953.
Int. Ed. Engl. 1997, 36, 1287−1289. (d) Hasenknopf, B.; Lehn, J.-M.;
Boumediene, N.; Leize, E.; Van Dorsselaer, A. Angew. Chem., Int. Ed.
1998, 37, 3265−3268. (e) Provent, C.; Rivara-Minten, E.; Hewage, S.;
Brunner, G.; Williams, A. F. Chem.Eur. J. 1999, 5, 3487−3494.
(f) Childs, L. J.; Alcock, N. W.; Hannon, M. J. Angew. Chem., Int. Ed.
2002, 41, 4244−4247. (g) Tuna, F.; Hamblin, J.; Jackson, A.;
Clarkson, G.; Alcock, N. W.; Hannon, M. J. Dalton Trans. 2003,
2141−2148. (h) Childs, L. J.; Pascu, M.; Clarke, A. J.; Alcock, N. W.;
Hannon, M. J. Chem.Eur. J. 2004, 10, 4291−4300. (i) Senegas, J.-
M.; Koeller, S.; Bernardinelli, G.; Piguet, C. Chem. Commun. 2005,
2235−2237. (j) Hamblin, J.; Tuna, F.; Bunce, S.; Childs, L. J.; Jackson,
A.; Errington, W.; Alcock, N. W.; Nierengarten, H.; Van Dorsselaer,
̌
̌
(8) Tkalec, U.; Ravnik, M.; Copar, S.; Zumer, S.; Muse
2011, 333, 62−65.
̌ ̌
vic, I. Science
(9) Beves, J. E.; Blight, B. A.; Campbell, C. J.; Leigh, D. A.;
McBurney, R. T. Angew. Chem., Int. Ed. 2011, 50, 9260−9327.
(10) (a) Chambron, J.-C.; Dietrich-Buchecker, C.; Sauvage, J.-P. In
Comprehensive Supramolecular Chemistry; Sauvage, J.-P., Hosseini, M.
W., Eds.; Elsevier: Oxford, 1996; Vol. 9, pp 43−83; (b) Hubin, T. J.;
Busch, D. H. Coord. Chem. Rev. 2000, 200, 5−52. (c) Collin, J.-P.;
9496
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497

Journal of the American Chemical Society
Article
A.; Leize-Wagner, E.; Hannon, M. J. Chem.Eur. J. 2007, 13, 9286−
9296. (k) Pang, Y.; Cui, S.; Li, B.; Zhang, J.; Wang, Y.; Zhang, H. Inorg.
Chem. 2008, 47, 10317−10324. (l) Allen, K. E.; Faulkner, R. A.;
Harding, L. P.; Rice, C. R.; Riis-Johannessen, T.; Voss, M. L.;
Whitehead, M. Angew. Chem., Int. Ed. 2010, 49, 6655−6658. (m) Bain,
L.; Bullock, S.; Harding, L.; Riis-Johannessen, T.; Midgley, G.; Rice
Craig, R.; Whitehead, M. Chem. Commun. 2010, 46, 3496−3498.
2004, 126, 10389−10396. (h) Wood, T. E.; Dalgleish, N. D.; Power,
E. D.; Thompson, A.; Chen, X.; Okamoto, Y. J. Am. Chem. Soc. 2005,
127, 5740−5741. (i) Kreickmann, T.; Diedrich, C.; Pape, T.; Huynh,
H. V.; Grimme, S.; Hahn, F. E. J. Am. Chem. Soc. 2006, 128, 11808−
11819. (j) Bunzen, J.; Bruhn, T.; Bringmann, G.; Lutzen, A. J. Am.
̈
Chem. Soc. 2009, 131, 3621−3630. (k) Stomeo, F.; Lincheneau, C.;
Leonard, J. P.; O’Brien, J. E.; Peacock, R. D.; McCoy, C. P.;
Gunnlaugsson, T. J. Am. Chem. Soc. 2009, 131, 9636−9637.
(l) Amendola, V.; Boiocchi, M.; Brega, V.; Fabbrizzi, L.; Mosca, L.
Inorg. Chem. 2010, 49, 997−1007. (m) Howson, S. E.; Bolhuis, A.;
Brabec, V.; Clarkson, G. J.; Malina, J.; Rodger, A.; Scott, P. Nature
Chem. 2012, 4, 31−36.
(39) (a) This is typical for Fe(diimine)3 complexes. It is known that
the isomeric forms of these complexes have practically identical
absorption spectra [ Krumholz, P. Inorg. Chem., 1965, 609 ], and that
the energy of the MLCT is strongly influenced by the number of
coordinated aromatic rings [ (b) Busch, D. H.; Bailar, J. C., Jr. J. Am.
Chem. Soc. 1956, 78, 1137−1142 ]. For example, MLCT in water:
Fe(Me-NCH−CHN-Me)3 568 nm,39b [Fe(py-CHNH−
CH3)3]I2 551 nm,39b and [Fe(py-CHNH−CH3)3](ClO4)2 551
nm.39a As the number of coordinated aromatic rings increases, the
energy of the MLCT increases, with [Fe(bpy)3]2+ and [Fe(phen)3]2+
exhibiting λmax at 490 and 510 nm, respectively. Combined, these data
surprisingly suggest that the the ligands in the present helicates behave
less like aromatic ligands such as bipyridine and more like nonaromatic
diimines.
(n) Tanh Jeazet, H. B.; Gloe, K.; Doert, T.; Kataeva, O. N.; Jager, A.;
̈
Geipel, G.; Bernhard, G.; Buchner, B.; Gloe, K. Chem. Commun. 2010,
̈
46, 2373−2375. (o) Giles, I. D.; Chifotides, H. T.; Shatruk, M.;
Dunbar, K. R. Chem. Commun. 2011, 47, 12604−12606.
(25) (a) Corbett, P. T.; Leclaire, J.; Vial, L.; West, K. R.; Wietor, J.-L.;
Sanders, J. K. M.; Otto, S. Chem. Rev. 2006, 106, 3652−3711.
(b) Lehn, J.-M. Chem. Soc. Rev. 2007, 36, 151−160.
(26) (a) Leigh, D. A.; Lusby, P. J.; Teat, S. J.; Wilson, A. J.; Wong, J.
K. Y. Angew. Chem., Int. Ed. 2001, 40, 1538−1543. (b) Price, J. R.;
Clegg, J. K.; Fenton, R. R.; Lindoy, L. F.; McMurtrie, J. C.; Meehan, G.
V.; Parkin, A.; Perkins, D.; Turner, P. Aust. J. Chem. 2009, 62, 1014−
1019.
(27) Hogg, L.; Leigh, D. A.; Lusby, P. J.; Morelli, A.; Parsons, S.;
Wong, J. K. Y. Angew. Chem., Int. Ed. 2004, 43, 1218−1221.
(28) Hutin, M.; Frantz, R.; Nitschke, J. R. Chem.Eur. J. 2006, 12,
4077−4082.
(29) Pentecost, C. D.; Chichak, K. S.; Peters, A. J.; Cave, G. W. V.;
Cantrill, S. J.; Stoddart, J. F. Angew. Chem., Int. Ed. 2007, 46, 218−222.
(30) Chichak, K. S.; Cantrill, S. J.; Pease, A. R.; Chiu, S.-H.; Cave, G.
W. V.; Atwood, J. L.; Stoddart, J. F. Science 2004, 304, 1308−1312.
(31) Twigg, M. V.; Burgess, J. In Comprehensive Coordination
Chemistry II; McCleverty, J. A., Meyer, T. J., Eds.; Elsevier:
Amsterdam, 2004; Vol. 5, pp 403−553.
(40) (a) Murner, H.; von Zelewsky, A.; Hopfgartner, G. Inorg. Chim.
̈
Acta 1998, 271, 36−39. (b) Kiehne, U.; Lutzen, A. Org. Lett. 2007, 9,
̈
5333−5336. (c) Howson, S. E.; Allan, L. E. N.; Chmel, N. P.;
Clarkson, G. J.; van Gorkum, R.; Scott, P. Chem. Commun. 2009,
1727−1729.
(32) Bu4NBF4 was also added to maintain a constant Bu4N+
concentration.
(41) (a) Bosnich, B. Acc. Chem. Res. 1969, 2, 266−273. (b) Mason, S.
F. Pure Appl. Chem. 1970, 24, 335−359. (c) Ziegler, M.; von Zelewsky,
A. Coord. Chem. Rev. 1998, 177, 257−300.
(33) Thompson, M. C.; Busch, D. H. J. Am. Chem. Soc. 1964, 86,
213−217.
(42) This is in agreement with both previous simple pyridyl−imine
(34) Chloride ions have been previously shown to catalyze the rate of
ligand exchange of [Fe(phen)3]2+ complexes in DMSO, see Farrinton,
D. J.; Jones, J. O.; Twigg, M. V. Inorg. Chim. Acta 1977, 25, L75−L76.
(35) This is likely primarily a solubility issue: in order for the kinetic
distribution of oligomers which are initially formed to rearrange into
the thermodynamically preferred product, these must all remain in
solution at all times. Dodecylamine (which is not soluble in DMSO)
affords the circular helicate increased solubility in chloroform (by
comparison with circular helicates [3a−3g]10+) and allows the
rearrangement process to proceed readily.
complexes40c and a triple-stranded dinuclear Fe(II)-containing
helicate. (a) Murner, H.; von Zelewsky, A.; Hopfgartner, G. Inorg.
̈
Chim. Acta 1998, 271, 36−39. Additionally, the CD band for the
MLCT transition of (S)-[3jCl](PF6)9 also corresponds to that of Λ-
[Fe(bpy)3]: (b) Mason, S. F.; Peart, B. J. J. Chem. Soc., Dalton Trans.
1973, 949−955 (ref 41c) and Λ-[Ru(phen)3] (ref 41a).
(43) Wolfe, S. Acc. Chem. Res. 1972, 5, 102−111.
(44) We reasoned that a more rigid and preorganized diamine based
on the same glycol linker, but with an aromatic ring at the center in
place of an ethyl spacer, may generate the corresponding pentafoil
knot in higher yield. However, we found that diamine 2,2′-(1,2-
bisoxyphenylene)diethanamine gave exclusively polymeric material,
with no evidence for pentamer formation.
(36) The isolated pentameric circular helicates were found to be
stable in MeCN at 60 °C for several weeks, with no evidence for
decomposition. However, the addition of a monoamine to isolated
samples of the circular helicates resulted only in the decomposition of
the helicate, with little evidence for amine exchange. Attempts were
made to reduce the imine groups of the molecular pentafoil knot [7Cl]
(PF6)9 using standard procedures (NaBH4), but these resulted in
decomplexation, and no organic knot could be isolated from the
reaction mixture.
(37) Crystal data for [3bCl](PF6)9·xsolvent: purple blocks, 0.11 mm
× 0.15 mm × 0.32 mm, M = 5358.62, C240H234ClF45N37O20P7.5Fe5,
monoclinic, space group P21/c, a = 34.023(2) Å, b = 23.714(2) Å, c =
36.537(2) Å, β = 94.386(2)°, V = 29393(4) Å3, Z = 4, Dc = 1.211 g/
cm3, 3211 parameters, 151 restraints, R = 0.1388 [Io > 2σ(Io)], wR =
0.3829 (all reflections).
(38) (a) Zarges, W.; Hall, J.; Lehn, J. M.; Bolm, C. Helv. Chim. Acta
1991, 74, 1843−1852. (b) Constable, E. C.; Kulke, T.; Baum, G.;
Fenske, D. Chem. Commun. 1997, 2043−2044. (c) Constable, E. C.;
Kulke, T.; Baum, G.; Fenske, D. Inorg. Chem. Commun. 1998, 1, 80−
82. (d) Mamula, O.; von Zelewsky, A.; Bernardinelli, G. Angew. Chem.,
Int. Ed. 1998, 37, 290−293. (e) Baum, G.; Constable, E. C.; Fenske,
D.; Housecroft, C. E.; Kulke, T. Chem.Eur. J. 1999, 5, 1862−1873.
(f) Lutzen, A.; Hapke, M.; Griep-Raming, J.; Haase, D.; Saak, W.
Angew. Chem., Int. Ed. 2002, 41, 2086−2089. (g) Orita, A.; Nakano, T.;
An, D. L.; Tanikawa, K.; Wakamatsu, K.; Otera, J. J. Am. Chem. Soc.
9497
dx.doi.org/10.1021/ja303355v | J. Am. Chem. Soc. 2012, 134, 9488−9497