Hairpin helicates: A missing link between double-helicates and trefoil knots

Article abstract of DOI:10.1039/b500209e

We have prepared a novel ligand in which two qpy metal-binding domains are linked by a polyoxyethylene spacer. The new ligand reacts with copper(II) salts to form dinuclear double helicates with a hairpin structure. The hairpin complex and a model complex

Full text of DOI:10.1039/b500209e

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Hairpin helicates: a missing link between double-helicates and trefoil
knots
Edwin C. Constable,* Egbert Figgemeier, Inger Annette Hougen, Catherine E. Housecroft,*
Markus Neuburger, Silvia Schaffner and Louise A. Whall
Department of Chemistry, University of Basel, Spitalstrasse 51, CH 4056, Basel, Switzerland.
E-mail: edwin.constable@unibas.ch; Fax: +41 61 267 1018; Tel: +41 61 267 1001
Received 6th January 2005, Accepted 8th February 2005
First published as an Advance Article on the web 22nd February 2005
We have prepared a novel ligand in which two qpy metal-binding domains are linked by a polyoxyethylene spacer.
The new ligand reacts with copper(II) salts to form dinuclear double helicates with a hairpin structure. The hairpin
complex and a model complex have been structurally characterised. The hairpin structure is formed regioselectively
as the head-to-tail (HT) stereoisomer. Detailed electrochemical and spectroelectrochemical studies of the complexes
are reported.
Introduction
The metal-directed assembly of topologically novel supra-
molecules is a topic that continues to attract synthetic and
theoretical interest.1–5 One of the earliest successes in this area
was the construction of helical and multiple helical complexes,
and synthetic studies of helicates resulted in a broad under-
standing of the factors that influence metal-directed assembly
processes. The vast majority of double helicates have been
prepared by self-assembly reactions of two ligand strands with
an appropriate number of commensurate metal centres.4,5
A
number of ligands have been reported which contain multiple
and spatially separated metal-binding domains, but these have
generally been used for the preparation of higher-double-
helicates.6 Rigid7 or highly preorganised8,9 scaffolds bearing
multiple oligopyridine chains have been used for the stereoselec-
tive assembly of multiple helicates. A single example has been
reported in which two Lehn-type bis(2,2ꢀ-bipyridine) ligands
have been linked by a polyethyleneoxy spacer and the resultant
ligand shown to form a hairpin dicopper(I) complex.9 One of
the most spectacular achievements has been the synthesis of
molecular trefoil knots using double-helical metal complexes as
a synthetic scaffold.10 Concepts of conventional and topological
chirality have played an integral role in the development of
strategies for the preparation of topologically novel compounds.
Complexes of stoichiometry [ML2] with two symmetrical
didentate (M four-coordinate pseudotetrahedral) or terdentate
ligands (M six-coordinate pseudooctahedral) each have two
orthogonal coordination planes defined by the donor atoms
of each ligand, exhibit D2d symmetry and are achiral and
typified by bis(bipy) and bis(terpy) complexes respectively.
Desymmetrising the ligands by substitution such that the two
outermost donor atoms of the ligands are no longer chemically
equivalent lowers the symmetry to C2 and the species become
chiral.11,12 Conversely, in most cases, symmetrical substitution
of the ligand has no effect upon the overall symmetry and,
for example, 4ꢀ-substituted 2,2ꢀ:6ꢀ,2ꢀ-terpyridines form achiral
D2d [ML2] complexes. We recently identified a novel situation
in which a ditopic ligand containing two terpy domains linked
by a flexible spacer gave rise to C2 symmetrical complexes, even
though the linker was attached to the 4ꢀ-positions of the terpy
(Fig. 1).13 In the same way that two mononuclear D2d units
may be linked to form the dinuclear double-helical structure
that is the key to the trefoil knot the new C2 unit may be
coupled to a D2d unit to give a new type of structure (B) that
is conceptually between a double helix (A) and a trefoil knot
(C) (Fig. 2, D2 representations of the structures presented in
each case). We use the notation PP or MM to describe the
Fig. 1 The prototype linked system resulting in the formation of C2
rather than D2d {M(terpy)2} motifs upon coordination to a single metal
centre.
Fig. 2 The relationship between the double helical (A), hairpin (B)
and trefoil-knot (C) structures. In each case a single (PP) enantiomer
has been depicted and for consistency, the trefoil knot is shown in the
(coordinated) D2 rather than the more familiar D3 form.
chirality at each metal centre; this is strictly redundant in a
helical system where the stereodescriptors P and M would
suffice, but this usage allows extension to the (non-chiral) PM
compound which leads to a macrocycle rather than a trefoil
knot upon dual closure. An interesting point is that although
each of these structures is chiral, A and B possess Euclidian
chirality whilst C is topologically chiral.14 In this paper, we
describe the preparation of hairpin double-helical structure of
this type incorporating 2,2ꢀ:6ꢀ,2ꢀꢀ:6ꢀꢀ,2ꢀꢀꢀ:6ꢀꢀꢀ,2ꢀꢀꢀꢀ-quinquepyridine
(qpy) metal-binding domains.15
Experimental
General
Commercially available chemicals were reagent grade and
were used without further purification; 2-(1-oxo-2-pyridin-
ioethyl)pyridine iodide16 and Ts(OCH2CH2)6OTs17 were pre-
pared by the literature methods. 1H and 13C NMR spectra were
recorded on Bruker 250, 400, 500 or 600 MHz spectrometers
and full assignments were made using COSY, ROESY, HMBC
and HMQC methods; chemical shifts are defined with TMS =
0 ppm for 1H and 13C spectra and are measured relative to
residual CHCl3 or CHD2CN. Infrared spectra were recorded
1 1 6 8
D a l t o n T r a n s . , 2 0 0 5 , 1 1 6 8 – 1 1 7 5
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Article Online
on Shimadzu FTIR-8400S or Mattson Genesis FTIR spec-
trophotometers with neat samples on a Golden Gate diamond
ATR accessory or in compressed KBr pellet respectively. UV-
VIS measurements were performed using a Varian Cary 5000
UV-VIS-NIR photospectrometer; spectra were recorded for ≈
10−5 M solutions in CHCl3 (ligands) or MeCN (coordination
compounds). EI mass spectra were recorded on a VG 70 SE
or a Kratos MS-50 mass spectrometer and electrospray (ES)
mass spectra on a Finnigan MAT LCQ mass spectrometer.
Maldi TOF mass spectra were recorded using a PerSeptive
Biosystems Voyager-RP or a PerSeptive Biosystems Voyager-
DE TM PRO Biospectrometry workstation. The microanalyses
were performed with a LECO CHN-900 apparatus.
1304m, 1250s, 1219m, 1204m, 1180s, 1150m, 1103m, 1034vs,
987vs, 949m, 810vs, 787s, 748m, 725m, 648m.
2-[3-(4-Hydroxyphenyl)-1-oxoprop-3-enyl]-6-[3-(4-methoxyphe-
nyl)-1-oxoprop-3-enyl]pyridine V. The most efficient route to
V was the reaction of I with 4-methoxybenzaldehyde (see text).
Compound I (6.90 g, 25.8 mmol) and 4-methoxybenzaldehyde
(30.0 ml, 33.6 g, 247 mmol) were heated in propanol (100 ml)
◦
and Et2NH (6 ml) at 80 C. After a few minutes, the reagents
had dissolved and the mixture was then heated at reflux for
2.5 h. During this period, further Et2NH (2 × 4 ml) was added.
After cooling the reaction mixture to room temperature, solvent
was removed in vacuo and a brown oil resulted. The product
and excess reagents were separated by column chromatography
(SiO2, CH2Cl2 : MeOH : Et2NH 20 : 1 : 0.1). The third fraction
was collected and V was isolated as an orange powder (yield:
3.6 g, 36%) (Found: C, 71.31; H, 5.22; N, 3.53. Calc. for
C24H19NO4·H2O: C, 71.45; H, 5.25; N, 3.47%); m/z (EI-MS)
385 [M]+; dH/ppm (400 MHz, CDCl3) 9.54 (b, OH), 8.36 (d, J
7.8 Hz, 2H, HC3,C5), 8.32 (overlapping d, J 15.9 Hz, 2H, Ha, Haꢀ),
8.07 (t, J 7.8 Hz, 1H, HC4), 8.00 (d, J 15.9 Hz, 1H, Hbꢀ), 7.99
(d, J 16.2 Hz, 1H, Hb), 7.72 (d, J 8.8 Hz, 2H, HG2), 7.68 (d, J
8.8 Hz, 2H, HF2), 6.98 (d, J 8.8 Hz, 2H, HG3), 6.92 (d, J 8.8 Hz,
2H, HF3), 3.89 (s, 3H, HOMe); dC/ppm (150 MHz, CDCl3) 189
Preparations
2-Acetyl-6-[3-(4-hydroxyphenyl)-1-oxoprop-3-enyl]pyridine I.
A solution of 2,6-diacetylpyridine (1.0 g, 6.1 mmol) and Et2NH
(2 ml) in propanol (20 ml) was heated to reflux. A solution
of 4-hydroxybenzaldehye (0.50 g, 4.1 mmol) in propan-1-ol
(10 ml) was then added dropwise over a period of an hour.
The mixture was heated at reflux for a further 2.5 h and
then cooled to room temperature. Solvent and Et2NH were
removed in vacuo. The dark brown, viscous residue was dissolved
in the minimum amount of CH2Cl2 and components were
separated by column chromatography (SiO2, CH2Cl2/MeOH
19/1). The first fraction was unreacted 2,6-diacetylpyridine
(0.55 g, 3.4 mmol). The second fraction (intense yellow) was
compound I, followed by unreacted 4-hydroxybenzaldehyde.
The final fraction was 2,6-bis[3-(4-hydroxyphenyl)-1-oxoprop-
3-enyl]pyridine II.18 Analytically pure, yellow I was obtained by
recrystallization of I from hot MeOH (yield: 0.41 g, 55% based
on 2,6-diacetylpyridine consumed) (Found: C, 71.8; H, 4.9; N,
5.2. Calc. for C16H13NO3: C, 71.9; H, 4.9; N, 5.2%); mp 198 ◦C;
m/z (Maldi-TOF-MS) 267 [M]+; dH/ppm (250 MHz, CDCl3)
8.92 (br, 1H, OH), 8.36 (dd, J 7.8, 1.0 Hz, 1H, HC5), 8.22 (d, J
15.6 Hz, 1H, Ha), 8.22 (dd, J 7.8, 1.0 Hz, 1H, HC3), 8.02 (t, J
7.8 Hz, 1H, HC4), 7.96 (d, J 15.6 Hz, 1H, Hb), 7.64 (d, J 8.3 Hz,
2H, HG), 6.91 (d, J 8.3 Hz, 2H, HG3, 2.87 (s, 3H, HMe); IR
(m˜max/cm−1) 3376br, 3200br, 1698s, 1648s, 1594sh, 1582s, 1554vs,
1510vs, 1360s, 1283m, 1270m, 1218s, 1166s, 1047m, 994w, 818w.
(C O), 162 (CG4), 159 (CF4), 154 (CC2,C6), 146 (Cb,bꢀ), 138 (CC4),
=
131 (CF2,G2), 128 (CF1,G1), 126 (CC3,C5), 118 (Ca,aꢀ), 117 (CF3), 115
(CG3), 57 (CMe); IR (m˜max/cm−1) 3225w, 2962m, 2824m, 2754m,
2476m, 1651s, 1589vs, 1558vs, 1504vs, 1458s, 1389m, 1342s,
1250vs, 1165vs, 1034vs, 987s, 810vs, 740m, 694m, 663m.
4ꢀ -(4-Hydroxyphenyl)-4ꢀꢀꢀ-(4-methoxyphenyl)-2,2ꢀ:6ꢀ,2ꢀꢀ:6ꢀꢀ,2ꢀꢀꢀ:
6ꢀꢀꢀ,2ꢀꢀꢀꢀ-quinquepyridine VI. Compound V (1.45 g, 3.76 mmol),
2-(1-oxo-2-pyridinioethyl)pyridine iodide (2.61 g, 7.99 mmol)
and dry ammonium acetate (10 g, 0.13 mol) were heated to
90 ◦C in MeOH (12 ml) under argon. Almost all the solids
dissolved; 4 drops of freshly distilled CH3CO2H were added to
the reaction mixture. This was then heated at reflux overnight
and a beige precipitate formed. The solution was allowed to
cool to 5 ◦C and the precipitate that formed was collected
by filtration. The solid was washed with ice-cold MeOH and
dried in vacuo to yield VI (1.07 g, 49%) (Found: C, 75.40; H,
4.89; N, 11.67. Calc. for C38H27N5O2·H2O: C, 75.61; H, 4.84;
N, 11.60%); m/z (Maldi-TOF-MS) 587 [M + 2H]+; dH/ppm
(600 MHz, DMSO-d6) 9.94 (s, 1H, OH), 8.95 (d, J 1.8 Hz,
1H, HD3), 8.93 (d, J 1.7 Hz, 1H, HB5), 8.74 (m, 2H, HA6,E6),
8.71 (m, 2H, HC3,C5), 8.70 (d, J 1.7 Hz, 1H, HD3), 8.67 (m, 3H,
HA3,B5,E3), 8.24 (t, J 7.7 Hz, 1H, HC4), 8.02 (td, J 7.6, 1.9 Hz,
2H, HE4,A4), 7.98 (d, 2H, J 8.7 Hz, HG2), 7.89 (d, 2H, J 8.6 Hz,
HF2), 7.50 (ddd, J 10.8, 3.6, 1.2 Hz, 2H, HA5,E5), 7.17 (d, 2H, J
8.8 Hz, HG3), 7.01 (d, 2H, J 8.5 Hz, HF3), 3.89 (s, 3H, HOMe);
dC/ppm (150 MHz, DMSO-d6) 160.4 (CG4), 159.1 (CF4), 155.4
(CB2,C2,D2,B6,C6,D6), 154.8 (CA2,E2), 149.2 (CA6,E6), 149.1 (CB4), 149.0
(CD4), 138.6 (CC4), 137.4 (CA4,E4), 129.6 (CF2), 128.2 (CF3,G2,G3),
124.4 (CA5,E5), 121.0 (CA3,C3,C5,E3), 117.5 (CB3,B5,D3,D5), 116.2 (CF3),
114.8 (CG4), 56.0 (MeO); IR (m˜max/cm−1) 3330br, 1574s, 1512vs,
1466m, 1450m, 1381s, 1258vs, 1173s, 1103s, 1103m, 1072m,
1034m, 987m, 887m, 818vs, 787vs, 733vs, 663m.
2-Acetyl-6-[3-(4-methoxyphenyl)-1-oxoprop-3-enyl]pyridine
III. A solution of 2,6-diacetylpyridine (4.77 g, 29.2 mmol)
and 4-methoxybenzaldehye (3.55 ml, 3.98 g, 29.2 mmol) were
dissolved in propanol (50 ml) and the mixture was heated to
◦
120 C. Et2NH (2 ml) was added and the orange solution was
heated at reflux for 4 h. During this period, Et2NH (2 × 2 ml)
was added dropwise to the hot mixture. The reaction mixture
was cooled to room temperature and the yellow solid that had
formed was filtered from the orange solution. Separation of
the products by column chromatography was unsuccessful.
The crude product was therefore dissolved in hot MeOH;
the addition of a small amount of toluene resulted in the
preferential recrystallization of 2,6-bis[3-(4-methoxyphenyl)-1-
oxoprop-3-enyl]pyridine IV (see below). Crystals of IV were
separated by filtration. Solvent was removed from the filtrate
in vacuo, and the crude yellow product was redissolved in a
mixture of CH2Cl2 and toluene. This solution was heated, and
hexane was slowly added to the hot solution. On cooling to
room temperature, III precipitated as a yellow solid (yield:
2.06 g, 25%) (Found: C, 71.69; H, 5.57; N, 4.95. Calc. for
C17H15NO3·0.2H2O: C, 71.67; H, 5.45; N, 4.92%); m/z (EI-MS)
281 [M]+; dH/ppm (250 MHz, CDCl3) 8.36 (dd, J 7.6, 1.0 Hz,
1H, HC5), 8.23 (d, J 16.2 Hz, 1H, Ha), 8.22 (dd, J 7.6, 1.0 Hz,
1H, HC3), 8.02 (t, J 7.6 Hz, 1H, HC4), 7.98 (d, J 16.2 Hz, 1H, Hb),
7.68 (d, J 9.1 Hz, 2H, HG2), 6.97 (d, J 9.1 Hz, 2H, HG3), 3.87
(s, 3H, HOMe), 2.87 (s, 3H, HMe); IR (m˜max/cm−1) 3001w, 1697s,
1666s, 1589s, 1566s, 1512s, 1458m, 1442m, 1420m, 1342m,
2-[3-(4-Hydroxyphenyl)-1-oxoprop-3-enyl]-6-(1-oxo-3-phenyl)
prop-3-enyl)pyridine VII. Compound I (6.77 g, 25.3 mmol)
was suspended in propan-1-ol (100 ml). To this, Et2NH (30 ml)
was added causing the colour to change to red. The suspension
was heated to 80 ◦C and then a solution of benzaldehyde
(10.0 ml, 10.5 g, 98.9 mmol) in propan-1-ol (80 ml) was added
over a period of 10 min. The resultant solution was heated at
reflux overnight during which time its colour changed to brown.
After cooling, solvent was removed in vacuo to give a brown oil.
The mixture was separated by column chromatography (SiO2,
CH2Cl2 : MeOH : Et2NH 20 : 1 : 0.1). Compound VII was
collected as the second fraction and was isolated as a yellow
powder (yield 3.35 g, 37%) (Found: C, 77.44; H, 5.10; N, 3.86;
D a l t o n T r a n s . , 2 0 0 5 , 1 1 6 8 – 1 1 7 5
1 1 6 9

View Article Online
C23H17NO3 requires C, 77.73; H, 4.82; N, 3.94%); m/z (EI-MS)
355 [M]+, 278 [M − Ph]+; dH/ppm (250 MHz, CDCl3) 8.50
(m, 4H, HC3, HC5, Ha, Haꢀ), 8.11–7.96 (m, 3H, HC4, Hb, Hbꢀ),
7.81–7.73 (m, 2H, HF2), 7.69 (d, J 8.7 Hz, 2H, HG2), 7.49–7.45
(m, 3H, HF3, HF4), 6.91 (d, J 8.5 Hz, 2H, HG3); IR (m˜max/cm−1)
3356br, 1659s, 1597s, 1558vs, 1504s, 1443m, 1335s, 1281m,
1234s, 1204s, 1165s, 1034vs, 980vs, 810vs, 764m, 741s, 679m.
solution of NH4PF6 in MeOH (1 ml) was added to the reaction
mixture. The brown precipitate that formed was collected by
filtration and was washed with ice-cold MeOH, water, ice-
cold MeOH again, and Et2O. The product was isolated as a
brown powder (19 mg, 34%) (Found C, 49.38; H, 3.29; N,
7.54; C76H54Cu2F18N10O4P3·6H2O requires C, 49.57; H, 3.61;
N, 7.61%); m/z (ES MS) 649 [M − L − Cu] +, 433 [M −
3PF6]3+; IR (m˜max/cm−1) 3094br, 1589s, 1543m, 1520m, 1481m,
1458m, 1420m, 1358m, 1242s, 1180s, 1018m, 818vs, 741m, 648m,
617m; UV-VIS (c = 111 lM and 11 lM) kmax/nm (MeCN) 226
(e/dm3 mol−1 cm−1 94 000), 292 (72 000), 317 (66 000), 351 (sh,
51 000), 467 (6400), 581 (2100).
4ꢀ -(4-Hydroxyphenyl)-4ꢀꢀꢀ-phenyl-2,2ꢀ:6ꢀ,2ꢀꢀ:6ꢀꢀ,2ꢀꢀꢀ:6ꢀꢀꢀ,2ꢀꢀꢀꢀ -quin-
quepyridine VIII. Compound VII (2.63 g, 7.40 mmol),
2-(1-oxo-2-pyridinioethyl)pyridine iodide (4.83 g, 14.8 mmol)
and ammonium acetate (5.98 g, 77.6 mmol) were dissolved in
dry MeOH (40 ml) under argon. An oil bath at 100 ◦C was used
to heat the reaction mixture. A brown solution formed. This
was then heated at reflux for 17 h to yield a beige precipitate.
After cooling to room temperature, the reaction mixture was
filtered. The precipitate was washed with ice-cold MeOH and
dried in vacuo. No further purification of VIII was necessary.
Yield: 3.64 g (89%) (Found: C, 78.82; H, 4.62; N, 12.52;
C37H25N5O·0.5H2O requires C, 78.70; H, 4.64; N, 12.40%); m/z
(MALDI-TOF MS) 579 [M + Na + H]+, 557 [M + 2H]+;
dH/ppm (600 MHz, DMSO-d6) 9.93 (s, 1H, OH), 9.04 (d, J
1.7 Hz, 1H, HB5), 8.98 (d, J 1.9 Hz, 1H, HD3), 8.79–8.70 (m,
6H, HC3,C5,A3,A6,E3,E6), 8.77 (s, 1H, HB3), 8.69 (s, 1H, HD5), 8.27 (t,
1H, J 7.7 Hz, HC4), 8.04 (m, 2H, HA4,E4), 8.02 (m, 2H, HF2), 7.89
(d, J 8.9 Hz, 2H, HG2), 7.64 (t, J 7.4 Hz, 2H, HF3), 7.59 (t, J
7.4 Hz, 1H, HF4), 7.56–7.52 (m, 2H, HA5,E5), 7.00 (d, J 8.8 Hz,
2H, HG3); IR (m˜max/cm−1) 3186br, 3055m, 1666m, 1605m, 1566s,
1543vs, 1520s, 1474s, 1434m, 1389vs, 1273s, 1219m, 1173m,
1103m, 1072m, 1041m, 995m, 887m, 833m, 818s, 787vs, 764vs,
733s, 694m, 678m, 656m, 640m, 617s.
[Cu2(IX)][PF6]3. Ligand IX (21 mg, 15 lmol) and
Cu(OAc)2·H2O (7.6 mg, 38 lmol) were dissolved in MeOH
(3 ml). The mixture was sonicated to give a green solution. A
filtered solution of NH4PF6 (56 mg, 34 lmol) in MeOH (1 ml)
was added to the reaction mixture. A green precipitate formed
and this was collected by filtration, was washed with ice-cold
MeOH and Et2O and was dried in vacuo to yield [Cu2(IX)][PF6]3
as a brown powder (22 mg, 74%) (Found C, 51.28; H, 3.62;
N, 6.75; C86H72Cu2F22.2N10O7P3.7 requires C, 51.11; H, 3.59; N,
6.93%); m/z (MALDI-TOF MS) 1629, [M + 2H − 2PF6]+,
1484 [M + 2H − 3PF6]+; IR (m˜max/cm−1) 3063br, 2870m, 1597s,
1574m, 1543m, 1520m, 1481m, 1450m, 1412m, 1242s, 1188s,
1111br, 1080m, 1011m, 833vs, 795s, 764s, 741s, 694m, 648m,
617m; UV-VIS (c = 122 lM and 12 lM) kmax/nm (MeCN) 224
(e/dm3 mol−1 cm−1 92 000), 289 (79 000), 316 (sh, 64 000), 351
(sh, 40 000), 472 (250), 574 (100).
Crystal structure determinations
Data were collected on an Enraf Nonius Kappa CCD in-
strument; data reduction, solution and refinement used the
progammes COLLECT19, SIR9220 and CRYSTALS.21
1,19-Bis[4-(4ꢀꢀꢀ -phenyl-2,2ꢀ:6ꢀ,2ꢀꢀ:6ꢀꢀ,2ꢀꢀꢀ:6ꢀꢀꢀ,2ꢀꢀꢀꢀ -quinquepyridin-
4ꢀ-yl)phenyl]-1,4,7,10,13,16,19-heptaoxanonadecane IX. Dry
VIII (319 mg, 0.574 mmol) and dry Cs2CO3 (107 g, 0.328 mmol)
were suspended in DMF (7 ml) under argon. The brown
reaction mixture was heated using an oil bath at 100 ◦C. A
solution of Ts(OCH2CH2)6OTs (169 mg, 0.286 mmol) in DMF
(5 ml) was added dropwise over a period of 5 min. The reaction
mixture was then heated at 80 ◦C for 20 h after which time a
white precipitate had formed. At this point, a CaCl2 drying
tube was added to the reflux condenser and and the mixture
heated at reflux for a further 70 h. During this period, more
precipitate formed. The reaction mixture was cooled to room
temperature. The solid was separated by filtration, washed
with DMF, water, MeOH and Et2O and finally dried in vacuo.
Crystal data for III. C17H15NO3, M = 281.31, monoclinic,
space group P21/c, a = 12.394(7), b = 4.032(4), c = 28.900(16)
◦
3
−3
˚
˚
A, b = 90.73(4) ,U = 1444.3(17) A , Z = 4, Dc = 1.294 Mg m ,
l(Mo-Ka) = 0.089 mm−1, T = 293 K, 2494 reflections collected
on an Enraf Nonius Kappa CCD instrument. Refinement of
1877 reflections (191 parameters) with I > 1.0r(I) converged at
final R1 = 0.0725, wR2 = 0.0579, gof = 1.06.
Crystal data for [Cu2(VI)2][PF6]3·1.5Me2CO. C80.50H63Cu2-
¯
F18N10O5.50P3, M = 1820.43, triclinic, space group P1, a =
˚
11.709(1), b = 17.378(3), c = 19.3791(9) A, a = 92.837(7),
◦
3
˚
b = 92.355(6), c = 98.49(1) ,U = 3890.5(7) A , Z = 2,
1
A H NMR spectrum of crude IX indicated that the product
Dc = 1.554 Mg m−3, l(Mo-Ka) = 0.712 mm−1, T = 173 K,
28493 reflections collected on an Enraf Nonius Kappa CCD
instrument. Refinement of 15370 reflections (1288 parameters)
with I > 3.0r(I) converged at final R1 = 0.0618, wR2 = 0.0674,
gof = 1.14.
was partially protonated. It was therefore treated with dilute
aqueous NaOH (pH 11) and extracted with chloroform. The
solvent volume was reduced and was then chromatographed
(Al2O3, CHCl3 : Et2NH 42 : 1). Ligand IX was collected as the
second fraction and was isolated as a yellow powder (202 mg,
52%) (Found C, 74.93; H, 5.33; N, 10.06; C86H72N10O7·0.5H2O
requires C, 75.09; H, 5.42; N, 10.18%); m/z (MALDI-TOF MS)
1358 [M + 2H]+; dH/ppm (600 MHz, CDCl3) 8.92 (d, J 1.7 Hz,
2H, HB5/D3), 8.87 (d, J 1.6 Hz, 2H, HD3/B5), 8.73–8.64 (m, 16 H,
HB3,C3,C5,D5,A3,A6,E3,E6), 8.00 (t, 2H, J 7.7 Hz, HC4), 7.91–7.81 (m,
12H, HF2,A4,E4,G2), 7.53–7.46 (m, 6H, HF3,F4), 7.34 (ddd, J 7.4, 4.8,
1,1 Hz, 4H, HA5/E5), 7.31 (ddd, J 7.4, 4.7, 1,1 Hz, 4H, HE5/A5),
Crystal data for [Cu2(IX)][PF6]3·2.5Me2CO. C93.5H87Cu2-
¯
F18N10O9.50P3, M = 1481.18, triclinic, space group P1, a =
˚
15.6921(2), b = 16.0207(2), c = 21.2804(3) A, a = 103.9732(7),
◦
3
˚
b = 102.0324(8), c = 109.0858(7) ,U = 4659.0(1) A , Z = 2,
Dc = 1.472 Mg m−3, l(Mo-Ka) = 0.607 mm−1, T = 173 K,
21247 reflections collected on an Enraf Nonius Kappa CCD
instrument. Refinement of 12623 reflections (1355 parameters)
with I > 3.0r(I) converged at final R1 = 0.0588, wR2 = 0.0514,
gof = 1.16.
7.01 (d, J 8.6 Hz, 4H, HG3), 4.19 (t, J 4.8 Hz, 4H, CH2 ), 3.91 (t,
f
e
d
c
J 4.8 Hz, 4H, CH2 ), 3.77 (m, 4H, CH2 ), 3.73 (m, 4H, CH2 ),
3.69 (m, 8H, CH2a,b); IR (m˜max/cm−1) 3055m, 2870m, 1605s,
1582vs, 1543vs, 1512vs, 1474s, 1443m, 1420m, 1389vs, 1296m,
1250vs, 1180m, 1111vs, 1072vs, 1041vs, 987s, 941m, 887s, 817vs,
794vs, 764vs, 733vs, 694vs, 663s, 640m, 617vs; UV-VIS (c =
7.37 lM) kmax/nm (CHCl3) 257 (e/dm3 mol−1 cm−1 101 500),
280 (97 600), 307 sh (57 500), 316 sh (46 300).
CCDC reference numbers 241100, 259911 and 259912.
See http://www.rsc.org/suppdata/dt/b5/b500209e/ for cry-
stallographic data in CIF or other electronic format.
Electrochemistry
[Cu2(VI)2][PF6]3. Ligand VI (19 mg, 32 lmol) and
Cu(OAc)2·H2O (8.2 mg, 41 lmol) were dissolved in MeOH
(3 ml). The brown solution was filtered. After 10 min, a filtered
The voltammetric experiments (cyclic and differential pulse)
were measured with a BAS 100 B/W electrochemical work-
station and a three-electrode cell, consisting of a silver wire as
1 1 7 0
D a l t o n T r a n s . , 2 0 0 5 , 1 1 6 8 – 1 1 7 5

View Article Online
a pseudo-reference, a glassy carbon disk as the working and a
platinum wire as counter electrode. After gaining a full set of
voltammograms, ferrocene was added in order to determine
the exact position of the signals within the potential win-
dow. Tetra-n-butylammonium hexafluorophosphate (TBAPF6,
0.1 M) acted as the electrolyte. All solutions were degassed
thoroughly for at least 15 min with argon; an inert gas blanket
was maintained over the solution during the measurements. All
potentials given in this paper refer to the ferrocene–ferrocenium
redox couple at 0 V. The spectroelectrochemistry was measured
with a three-electrode configuration in a 1 mm cuvette with
platinum mesh as working electrode and counter electrode and
a silver wire as pseudo-reference electrode. The instrument used
for recording the spectra was an Agilent 8453 diode array
spectrophotometer.
mono-enones are particularly easy to detect by IR spectroscopy
=
as they exhibit two C O stretching modes; one is characteristic
cm−1
of an aryl ketone at ≈ 1698
and the other is assigned to
the enone and is shifted 30–40 cm−1 to lower energy. Reaction
of I with an excess of 4-methoxybenzaldehyde in propanol
at reflux in the presence of Et2NH gave the bis(enone) V in
36% yield. Compound V could be separated by column chro-
matography. The alternative route to V by reaction of III with
4-hydroxybenzaldehyde yielded mixtures which were relatively
difficult to separate by chromatography. These conversions are
summarised in Scheme 1, which also presents the nomenclature
used to distinguish the various aromatic rings.
In a Kro¨hnke reaction, the bis(enone) V reacted smoothly
with 2-(1-oxo-2-pyridinioethyl)pyridine iodide (Scheme 2) and
ammonium acetate in methanol at 90 ◦C to give the asymmetri-
cal qpy ligand VI in 49% yield. The ligand was characterized by
1H and 13C NMR spectroscopies and mass spectrometry. In the
Maldi-TOF mass spectrum, the highest mass peak at m/z 587
corresponded to [M + 2H]+. The 1H and 13C NMR spectra were
fully assigned (see Experimental section) using a combination
of COSY, ROESY, HMBC and HSQC technqiues. We discuss
the coordination behaviour of ligand III later in the paper.
Results and discussion
Ligand syntheses
Our strategy was to prepare an asymmetrically-substituted
2,2ꢀ:6ꢀ,2ꢀꢀ:6ꢀꢀ,2ꢀꢀꢀ:6ꢀꢀꢀ,2ꢀꢀꢀꢀ-quinquepyridine (qpy) in which a hydroxy-
phenyl substituent was subsequently to be used as a nucleophile
in a reaction with a bis-electrophile to yield the desired bis-
qpy ligand. Our initial synthetic targets were the asymmet-
ric 2,2ꢀ:6ꢀ,2ꢀꢀ:6ꢀꢀ,2ꢀꢀꢀ:6ꢀꢀꢀ,2ꢀꢀꢀꢀ-quinquepyridine (qpy) ligands VI and
VIII. Compound VI was attractive as, in principle, it would allow
us to complete the conceptual relationship between double helix,
hairpin and trefoil knot by deprotection of the methyl group and
a second coupling reaction. Standard Kro¨hnke methodology16
was chosen for the final assembly of the ligands, but the initial
requirement was a route to the asymmetrical bis-enones V and
VII. Two strategies for compound V were explored and these are
summarised in Scheme 1. The reaction of 2,6-diacetylpyridine
with either 4-hydroxybenzaldehyde or 4-methoxybenzaldehyde
(in 1 : 1 molar ratio or with the 2,6-diacetylpyridine in excess)
in propanol in the presence of Et2NH yielded mixtures of the
enones I or III and the bis-enones II or IV. The mono- and bis-
enones were readily separated by column chromatography. The
Scheme 2 Synthetic route to ligand VI showing the ring labelling used.
Ligand VI contains a single qpy domain, and has a ‘labelled’
head and tail which give us the opportunity to investigate
prefences for head-to-head or ‘head-to-tail’ binding upon the
formation of dinuclear helicates.22 These preferences will be
crucial in determining the nature of hairpin helicates with linked
qpy domains.
We decided upon the use of a polyoxyethylene spacer to link
the two qpy domains, partly because of the successful use of such
spacers in the synthesis of catenanes and knots,1,2 and partly
because we hoped to use a “caesium effect” to maximise the
yield in the coupling reaction.23 We also decided that the most
successful approach would be to work with ligands in which
a simple phenyl rather than a 4-methoxyphenyl group acted as
the desymmetrising spectator, to avoid ether exchange reactions.
The strategy for the assembly of the new ligand IX with a phenyl
spectator in the 4ꢀꢀꢀ-position is shown in Scheme 3. The reaction
of compound I (Scheme 1) with benzaldehyde in the presence
of Et2NH in propanol at reflux gave, after chromatographic
workup, the asymmetric bis(enone) VII in moderate yield.
Compound VII was converted into the asymmetrical qpy ligand
VIII using standard Kro¨nke methodology. The ligand was
isolated in 89% yield and was characterized by spectroscopic
and mass spectrometric techniques, the 1H NMR spectrum
being assigned by COSY and ROESY methods. Following the
reaction of VIII with Cs2CO3 in DMF at 100 ◦C to generate
the phenoxide, the two qpy domains were linked by reaction
with Ts(OCH2CH2)6OTs in DMF at reflux for 90 h. Ligand
IX was obtained as a pale yellow solid in moderate yield,
Scheme 1 Synthetic routes to compounds I–V; showing the ring
labelling used.
D a l t o n T r a n s . , 2 0 0 5 , 1 1 6 8 – 1 1 7 5
1 1 7 1

View Article Online
Fig. 3 The molecular structure of III. Important bond lengths and
angles: C(1)–O(1) = 1.206(3), C(1)–C(3) = 1.483(4), C(3)–N(1) =
1.337(3), N(1)–C(7) = 1.336(3), C(7)–C(8) = 1.491(4), C(8)–O(2) =
1.217(3), C(8)–C(9)
=
1.464(3), C(9)–C(10)
=
1.323(3), C(10)–
˚
C(11) = 1.451(3), C(14)–O(3) = 1.363(3), O(3)–C(17) = 1.421(3) A;
O(1)–C(1)–C(2) = 121.9(3), O(1)–C(1)–C(3) = 120.1(3), C(2)–C(1)–
C(3)
=
118.0(3), C(3)–N(1)–C(7)
=
117.9(2), C(7)–C(8)–O(2)
=
119.6(2), O(2)–C(8)–C(9) = 122.4(2), C(7)–C(8)–C(9) = 118.0(2), C(8)–
C(9)–C(10) = 121.3(2), C(9)–C(10)–C(11) = 127.6(2), C(14)–O(3)–
C(17) = 117.3(2)◦.
Complex formation and characterisation
Two qpy ligands present a total of ten nitrogen donor atoms.
Each qpy ligand strand is expected to be partitioned into
a 2,2ꢀ:6ꢀ,2ꢀꢀ-terpyridine-like terdentate and a 2,2ꢀ-bipyridine-
like bidentate domain giving, in the absence of participation
by donor solvent molecules, compatibility with 4-, 5- or 6-
coordinate metal centres. Dicopper double helicates of qpy
ligands are known that have two copper(I), one copper(I) and
one copper(II) and two copper(II) centres.25–28 The latter two
coordination modes have been structurally characterised; in
the mixed oxidation state complex the total of ten nitrogen
donors satisfies the total coordination requirement of the four-
coordinate copper(I) and six-coordinate copper(II) centres and
a [4 + 6] helicate is formed. In the case of the dicopper(II)
compound, one metal is in a six-coordinate {Cu(terpy)2}
environment whilst the other is coordinated to four nitrogens
from the two qpy strands and additional solvent or counter-ion
donors.
Scheme 3 Synthetic route to ligand IX; ring labelling as used in the
Experimental section.
and was characterized by spectroscopic and mass spectrometric
techniques. The highest mass peak in the MALDI-TOF mass
spectrum at m/z 1358 was assigned to [M + H]+.
One of the principal difficulties we encountered in this
work was in distinguishing the asymmetric species from
mixtures of symmetric compounds. For example, the 1H
NMR spectum of VIII is essentially a superposition of the
two symmetrical compounds 4ꢀ,4ꢀꢀꢀ-diphenyl-2,2ꢀ:6ꢀ,2ꢀꢀ:6ꢀꢀ,2ꢀꢀꢀ:6ꢀꢀꢀ,
2ꢀꢀꢀꢀ-quinquepyridine and 4ꢀ,4ꢀꢀ-bis(4-hydroxyphenyl)-2,2ꢀ:6ꢀ,2ꢀꢀ:
6ꢀꢀ,2ꢀꢀꢀ:6ꢀꢀꢀ,2ꢀꢀꢀꢀ-quinquepyridine. Similarly, the asymmetric bis-
enones are almost indistinguishable from mixtures of the
symmetrical components and the mono-enones to mixtures of
2,6-diacetylpyridine and the bis-enone. Unambiguous proof of
the structure of the mono-enones came from a crystal structure
of III.
Ligand VI reacted readily with copper(II) acetate in methanol
and the mixed oxidation state complex [Cu2(VI)2][PF6]3 was
isolated as the hexafluorophosphate salt in 34% yield. The
brown colour suggested reduction of Cu(II) had taken place;
the formation of the mixed Cu(I)/Cu(II) complex was sup-
ported by elemental analysis and mass spectrometric data. The
electrospray MS showed an envelope at 433 with component
peaks at 0.33 mass unit separation; the envelope of peaks
was collectively assigned to [M − 3PF6]3+ and the isotopomer
distribution corresonded to that simluated for a dicopper species.
1
A broadened and shifted, but very poorly resolved H NMR
Crystal structure of compound III
spectrum (600 MHz in CD3CN) exhibited signals in a shift
range d +25 to +6 ppm, indicative of a paramagnetic species.
No attempt was made to further assign this spectrum. A crystal
structure determination of the complex was undertaken and
provided confirmation of a mixed valence copper complex.
Brown, X-ray quality crystals of [Cu2(VI)2][PF6]3·1.5Me2CO
were grown by diffusion of Et2O into an acetone solution of
the complex. In the cation [Cu2(VI)2]3+, the two ligands form
a double helical arrangement embracing two copper centres.
Cations with both PP and MM chirality are present in the lattice.
The five pyridine rings in each ligand are partitioned into bipy-
and terpy-coordination domains. The copper(II) centre is in a
distorted octahedral environment, while the copper(I), is in a
distorted tetrahedral environment. The Cu · · · Cu separation is
Several crystals of III suitable for X-ray crystallography grew
fortuitously from a hexane/ethyl acetate/diethylamine solvent
mixture during attempts to purify the crude material by column
chromatography. The molecular structure of III is shown in
Fig. 3, and selected bond distances and angles are given in the
figure caption. All intramolecular bond lengths and angles are
1
typical, and as expected from the H NMR spectrum (3JHH
=
16.2 Hz), thealkenehas an E stereochemistry. With the exception
of the hydrogen atoms of the two methyl groups, all the atoms in
a molecule of III are close to being coplanar; the angle between
the least squares planes of the pyridine and aryl rings is 9.1◦. The
molecules form a herring-bone assembly along the b-axis. Along
the c-axis, molecules are assembled in stacks, each molecule
being off-set with respect to the next; van der Waals interactions
operate between aromatic rings with closest pyridine–pyridine
˚
3.94 A, too long to be considered a bonding interaction. Two
views of the dicopper cation are given in Fig. 4. The cation
suffers from several disorders. In one of the methoxy groups,
˚
˚
contacts being 3.43 A, and aryl–aryl contacts of 3.65 A.
1 1 7 2
D a l t o n T r a n s . , 2 0 0 5 , 1 1 6 8 – 1 1 7 5

View Article Online
with ice-cold MeOH, H2O and Et2O, the colour darkened and
[Cu2(IX)][PF6]3 was isolated as a brown powder in 74% yield.
The colour change indicates that initially a dicopper(II) species
is formed, that is readily reduced to a mixed-valence species.
This same scenario is observed when copper(II) reacts with the
parent qpy ligand.24,25 The elemental analysis of the product
indicated that the bulk sample was a mixture of Cu(II)Cu(II) and
Cu(II)Cu(I) species (see Experimental section). In the MALDI-
TOF mass spectrum, the highest mass envelopes at m/z 1629
and 1484 were assigned to the ions [M + 2H − 2PF6]+ and
[M + 2H − 3PF6]+; the isotope patterns observed matched those
1
simulated. The H NMR spectrum of the complex exhibited
broad signals over the range d +25 to +3.5 ppm. Although the
spectrum could not be fully assigned, the lowest field signals
can be attributed with some confidence to protons closest to the
paramagnetic metal centres, while signals at d 3.9, 3.8, 3.6 and
3.5 were assigned to CH2 protons within the polyether chain
which is spatially remote from the metal centres.
Single crystals of [Cu2(IX)][PF6]3·2.5Me2CO were grown by
diffusion of Et2O vapour into an acetone solution of the com-
plex. The [Cu2(IX)]3+ cation is chiral and the unit cell contains
both PP and MM enantiomers related by an inversion centre.
Fig. 5a shows the structure of the PP-enantiomer. The two qpy
coordination domains of ligand IX form a helical assembly
around the two copper centres (Fig. 5b). As in the structure
of [Cu2(VI)2]3+, each qpy ligand strand is partitioned into a
bipy- and terpy-subunit, providing appropriate coordination
environments for a copper(I) centre (distorted tetrahedral Cu(2),
RN–Cu–N angles = 664.3◦) and a copper(II) centre (distorted
octahedrals). The terpy unit of one ligand strand consists of
rings C, D and E (see Scheme 3) while that of the second strand
is made up of rings A, B and C. The two bipy domains are made
up of rings A and B of strand 1 and rings D and E of strand
2. The polyether chain forms a hairpin conformation, but the
extent to which this preorganises the two qpy strands for helicate
formation must be viewed with caution (see below). The two Cu
Fig.
4
(a) Structure of the P-enantiomer of the cation in
[Cu2(VI)2][PF6]3·1.5Me2CO with selected atom labelling. (b) View of
the cation looking down the Cu–Cu axis showing the helical array
of the two ligands. Selected bond lengths and angles: Cu(1)–N(4) =
2.037(3), Cu(1)–N(5) = 1.986(3), Cu(1)–N(9) = 2.011(3), Cu(1)–N(10) =
2.011(3), Cu(2)–N(1) = 2.186(3), Cu(2)–N(2) = 1.963(3), Cu(2)–N(3) =
2.357(3), Cu(2)–N(6) = 2.104(3), Cu(2)–N(7) = 1.932(3), Cu(2)–N(8) =
˚
2.243(3) A; N(4)–Cu(1)–N(5) = 81.4(1), N(4)–Cu(1)–N(9) = 133.9(1),
N(5)–Cu(1)–N(9) = 123.6(1), N(4)–Cu(1)–N(10) = 109.8(1), N(5)–
Cu(1)– N(10) = 134.7(1), N(9)–Cu(1)–N(10) = 80.7(1), N(1)–Cu(2)–
N(2) = 77.9(1), N(2)–Cu(2)–N(3) = 75.7(1), N(3)–Cu(2)–N(6) = 83.0(1),
N(1)–Cu(2)–N(7)
=
98.7(1), N(3)–Cu(2)–N(7)
=
107.7(1), N(6)–
Cu(2)–N(7) = 78.7(1), N(1)–Cu(2)–N(8) = 83.6(1), N(2)–Cu(2)–N(8) =
102.7(1), N(3)–Cu(2)–N(8) = 100.3(1), N(7)–Cu(2)–N(8) = 76.9(1),
N(1)–Cu(2)–N(6) = 104.4(1), N(2)–Cu(2)–N(6) = 101.9(1)◦.
˚
centres are non-bonded and the separation (4.0 A) is very similar
24,25
˚
to that in [Cu2(qpy)2][PF6]3
(3.96 A) and [Cu2(VI)2][PF6]3 (see
there are two methyl positions (C130 and C180) with 71.5 and
28.5% occupancies respectively; the major occupancy (C130) is
shown in Fig. 4. Each of the C6 rings in the 4-hydroxyphenyl
substituents is disordered over two positions, each with a
50% occupancy. The two phenyl rings of the 4-methoxyphenyl
substituents are mutually staggered and there are weak van
der Waals interactions between them (closest distance between
above). In the hairpin complex, as in [Cu2(VI)2][PF6]3 there is
the possibility of forming head-to-head (HH) and head-to-tail
(HT). Once again, in the solid state, a remarkable regioselectivity
is observed and only the HT stereoisomer is observed. There is
extensive interstrand p-stacking within the double helix with
most significant interactions between D pyridine rings of each
strand and between the aryl rings G (Scheme 3 for numbering
and Fig. 5b).
˚
C atoms is 3.90 A). A number of features are of interest in
this structure. Firstly, the introduction of the substituents has
remarkably little effect on the gross structural features of the heli-
cates as shown by a comparison with the analogous mixed oxida-
tion state helicates with 2,2ꢀ:6ꢀ,2ꢀꢀ:6ꢀꢀ,2ꢀꢀꢀ:6ꢀꢀꢀ,2ꢀꢀꢀꢀ-quinquepyridine
We had originally speculated about the possibilty of binding
Group 1 metals to the polyoxyethylene spacer which might
act as a pseudo-crown ether. However, as the space-filling
representation in Fig. 5c shows, the binding cavity is more
apparent than real. The role of the caesium carbonate in
the synthesis remains speculative, although the yields were
significantly and reproducably higher than when potassium or
sodium carbonate were used as base.
24,25
and 6,6ꢀꢀꢀꢀ-dimethyl-4ꢀ,4ꢀꢀꢀ-diphenyl-
˚
˚
(Cu · · · Cu, 3.957 A)
ꢀ
ꢀ
ꢀꢀ ꢀꢀ ꢀꢀꢀ ꢀꢀꢀ ꢀꢀꢀꢀ
2,2 :6 ,2 :6 ,2 :6 ,2 -quinquepyridine (Cu · · · Cu, 3.957 A)28 al-
though the helical pitch, as measured through the torsion angles
between bipy and terpy domains varies slightly.
Although the substituents do not greatly change the basic
helical structure, they exert a remarkable degree of regiospeci-
ficity. Although we cannot comment on the solution equilibria,
the solid state crystal contains only the head-to-tail (HT)
regioisomer in which each metal centre is coordinated to the
B ring of one ligand strand and the D ring of the other.
Although the substituents are similar in size, the hydroxyphenyl
residue can become involved in classical hydrogen bonding, but
although the two hydroxy groups are on the same side, as seen in
Electrochemistry and spectroelectrochemistry
Even though the structural properties of the hairpin complex
closely resembled those of the model complexes with simple
qpy ligands, we hoped that the electrochemical properties might
give an insight into the dynamic properties. In particular, we
hoped that rearrangements involving ligand dissociation would
be retarded through the linking of the two qpy domains.
The electrochemical properties of complex [Cu2(IX)][PF6]3
and its non-linked analogue [Cu2(VI)2][PF6]3 were mea-
sured, and were found to very closely resemble those of
[Cu2(qpy)2][PF6]3.24,25 Two peaks associated with metal-centred
processes were observed for both compounds at moderate
potentials (−0.44 V, 0.13 V for [Cu2(VI)2][PF6]3 and −0.43 V,
0.16 V for [Cu2(IX)][PF6]3 (determined by differential pulse
˚
Fig. 4b, the O · · · O distance of 6.1 A precludes intramolecular
hydrogen bonding. However, extensive hydrogen bonding with
the counterions and solvent is apparent with HO · · · X (X = F,
˚
acetone O) distances in the range 2.816–2.900 A.
The hairpin ligand IX reacted with copper(II) acetate in
methanol to give a green solution from which a green solid
was precipitated by the addition of NH4PF6. After washing
D a l t o n T r a n s . , 2 0 0 5 , 1 1 6 8 – 1 1 7 5
1 1 7 3

View Article Online
Fig. 6 Cyclic voltammograms for [Cu2(IX)][PF6]3 with a scan speed
of 100 mV s−1 (dotted line) and 50 mV s−1 (dashed line) measured in
acetonitrile.
fully reduced state, Cu(I)Cu(I), are particularly slow. The reason
for this behaviour is unknown, but we speculate that the reduced
flexibility of the bridged ligand in [Cu2(IX)][PF6]3 makes it more
difficult for the complex to adjust its structure to accommodate
a change in oxidation state: i.e. the ligand is more pre-organised.
This gives rise to a higher reorganisation energy on changing
redox state, leading to slower electron transfer rates.29 In order
to examine the stability of the complexes in different oxidation
states, spectroelectrochemical experiments were performed. For
[Cu2(VI)2][PF6]3 and [Cu2(IX)][PF6]3, the development of UV-
VIS absorption spectra was measured with applied negative
potentials or applied positive potentials in order to generate
the Cu(I)Cu(I) and Cu(II)Cu(II) states, respectively. The results
are presented in Fig. 7.
Fig.
5 (a) Structure of the PP-enantiomer of the cation in
[Cu2(IX)][PF6]3·2.5Me2CO with selected atom labelling. (b) View of the
cation looking down the Cu–Cu axis showing the helical assembly of
the two ligand strands. Selected bond lengths and angles: Cu(1)–N(1) =
2.117(3), Cu(1)–N(2) = 1.950(3), Cu(1)–N(3) = 2.240(3), Cu(1)–N6 =
2.201(3), Cu(1)–N(7) = 1.973(3), Cu(1)–N(8) = 2.354(3), Cu(2)–N(4) =
2.027(3), Cu(2)–N(5) = 2.006(3), Cu(2)–N(9) = 2.035(3), Cu(2)–N(10 =
˚
2.001(3) A; N(1)–Cu(1)–N(2) = 79.3(1), N(2)–Cu(1)–N(3) = 77.1(1),
N(1)–Cu(1)–N(6) = 103.6(1), N(2)–Cu(1)–N(6) = 96.0(1), N(3)–Cu(1)–
N(6) = 84.9(1), N(1)–Cu(1)–N(7) = 99.3(1), N(3)–Cu(1)–N(7) =
104.8(1), N(6)–Cu(1)–N(7) = 78.3(1), N(1)–Cu(1)–N(8) = 80.7(1),
N(2)–Cu(1)–N(8) = 110.5(1), N(3)–Cu(1)–N(8) = 102.1(1), N(7)–
Cu(1)–N(8) = 75.2(1), N(4)–Cu(2)–N(5) = 81.4(1), N(4)–Cu(2)–N(9) =
130.0(1), N(5)–Cu(2)–N(9) = 114.1(1), N(4)–Cu(2)–N(10) = 119.8(1),
N(5)–Cu(2)–N(10) = 137.1(1), N(9)–Cu(2)–N(10) = 81.9(1)◦. (c) A
space-filling representation of the hairpin complex showing that
the polyoxyethylene linker is tightly folded and has no crown-like
metal-binding cavity.
Fig. 7 Evolution of UV-VIS absorption spectra under applied po-
tentials for [Cu2(VI)2][PF6]3 and [Cu2(IX)][PF6]3 (A: [Cu2(VI)2][PF6]3,
oxidative conditions; B: [Cu2(VI)2][PF6]3, reductive conditions; C:
[Cu2(IX)][PF6]3, oxidative conditions; D: [Cu2(IX)][PF6]3, reductive
conditions). The applied potentials were chosen 200 mV more positive
and more negative, respectively, than the according redox potentials.
voltammetry, potentials quoted versus Fc/Fc+), indicating that
all three redox states (Cu(I)Cu(I), Cu(I)Cu(II) and Cu(II)Cu(II)
are readily accessible. As in [Cu2(qpy)2][PF6]3,24,25 both metal-
centred processes for [Cu2(VI)2][PF6]3 are reversible. In contrast,
the metal-centred processes for [Cu2(IX)][PF6]3 are irreversible
at a scan rate of 100 mV s−1. At slower scan rates, the redox
processes of [Cu2(IX)][PF6]3 become more reversible, with a
smaller difference between anodic and cathodic waves in the
cyclic voltammogram (Fig. 6). A higher reversibility at lower
scan rates is in contrast to most irreversible processes, in which
the redox process competes with chemical reactions of the
reduced or oxidised state of the compound. Therefore, it is
proposed that the complex is chemically stable in all redox states,
but that the electron-transfer processes, commencing from the
The major features of the spectra in Fig. 7 between 400 and
750 nm are due to a MLCT transitions of the copper(I) species
of the two complexes. Oxidation to copper(II) quenches these
absorptions for both complexes (Fig. 6, A and C). Upon re-
reduction of copper(II) to copper(I) restores the original features
of the spectra with an absorption maximum at 475 nm and
a shoulder at longer wavelengths (around 600 nm). The over-
laid spectra of complexes [Cu2(VI)2][PF6]3 and [Cu2(IX)][PF6]3
exhibit two clean isosbestic points, at 416 and 738 nm for
[Cu2(IX)][PF6]3 and 429 and 720 nm for [Cu2(VI)2][PF6]3. The
latter values closely resemble the results reported previously for
[Cu2(qpy)2][PF6]3.24,25
The reversibility of the spectroelectrochemistry indicates that
despite the irreversible behaviour seen in the voltammetric
1 1 7 4
D a l t o n T r a n s . , 2 0 0 5 , 1 1 6 8 – 1 1 7 5

View Article Online
experiments, [Cu2(IX)][PF6]3 is stable in all oxidation states,
supporting the conclusion that the electron transfer rates are
slow due to demanding re-organisation processes.
8 J. Libman, Y. Tor and A. Shanzer, J. Am. Chem. Soc., 1987, 109,
5880.
9 T. Nabeshima, T. Inaba, N. Furukawa, T. Hosoya and Yumihiko
Yanot, Inorg. Chem., 1993, 32, 1407.
10 Molecular Catenenanes, Rotaxanes and Knots, eds. J.-P. Sauvage and
C. Dietrich-Buchecker, Wiley-VCH, Weinheim, 1999.
11 E. C. Constable, C. E. Housecroft, T. Kulke, C. Lazzarini, E.
Schofield and Y. Zimmermann, J. Chem. Soc., Dalton Trans., 2001,
2864.
Conclusions
We have prepared a novel ligand in which two qpy metal-
binding domains are linked by a polyoxyethylene spacer. The
new ligand reacts with copper(II) salts to form dinuclear double
helicates with a hairpin structure. The hairpin structure is formed
regioselectively as the HT stereoisomer. Ligand IX is sterically
more demanding than VI, and this results in a significantly
slower geometrical reorganisation within the dinuclear complex
leading to sluggish electron transfer rates upon switching of
redox states. In future work, we will extend these studies to the
preparation of qpy-based trefoil knots.
12 A. von Zelewsky, Stereochemistry of Coordination Compounds, Wiley,
Chichester, 1996.
13 H. S. Chow, E. C. Constable and C. E. Housecroft, Dalton Trans.,
2003, 4568.
14 J.-C. Chambron, C. Dietrich-Buchecker and J.-P. Sauvage, Top. Curr.
Chem., 1993, 165, 131.
15 E. C. Constable, I. A. Hougen, C. E. Housecroft, M. Neuburger, S.
Schaffner and L. A. Whall, Inorg. Chem. Commun., 2004, 7, 1128.
16 F. Kro¨hnke, Synthesis, 1976, 1.
17 M. P. Andrews and G. A. Ozin, Inorg. Chem., 1986, 25, 2587.
18 L. A. Whall, PhD Thesis, University of Basel, 1998.
19 COLLECT Software, Nonius BV, Delft, The Netherlands, 1997–
2001.
20 A. Altomare, G. Cascarano, G. Giacovazzo, A. Guagliardi, M. C.
Burla, G. Polidori and M. Camalli, J. Appl. Crystallogr., 1994, 27,
435.
Acknowledgements
We should like to thank the Swiss National Science Foundation,
the University of Basel and the Portlandcement Fabrik for
financial support.
21 D. J. Watkin, C. K. Prout, J. R. Carruthers, P. W. Betteridge and
R. I. Cooper, CRYSTALS, Issue 11, Chemical Crystallography
Laboratory, Oxford, UK, 2001.
References
22 E. C. Constable, F. R. Heirtzler, M. Neuburger and M. Zehnder,
Supramol. Chem., 1995, 5, 197; E. C. Constable, F. R. Heirtzler, M.
Neuburger and M. Zehnder, Chem. Commun., 1996, 5, 933; E. C.
Constable, F. R. Heirtzler, M. Neuburger and M. Zehnder, J. Am.
Chem. Soc., 1997, 119, 5606; G. Baum, E. C. Constable, D. Fenske,
C. E. Housecroft, T. Kulke, M. Neuburger and M. Zehnder, J. Chem.
Soc., Dalton Trans., 2000, 119, 945.
23 A. Ostrowicki, E. Koepp and F. Vo¨gtle, Top. Curr. Chem., 1992, 161,
37.
24 M. Barley, E. C. Constable, S. A. Corr, R. C. S. McQueen, J. C.
Nutkins and M. D. Ward, J. Chem. Soc., Dalton Trans., 1988,
2655.
1 Perspectives in Supramolecular Chemistry Vol. 5, Transition Metals
in Supramolecular Chemistry, ed. J.-P. Sauvage, Wiley, Chichester,
1999.
2 Comprehensive Supramolecular Chemistry, eds. J. L. Atwood,
J. E. D. Davies, D. D. MacNicol, F. Vo¨gtle and J.-M. Lehn, Pergamon,
Oxford, 1996, vol. 7.
3 G. F. Swiegers and T. J. Malefetse, Chem. Rev., 2000, 100, 3483;
Coord. Chem. Rev., 2002, 225, 91.
4 C. Piguet, G. Bernardinelli and G. Hopfgartner, Chem. Rev., 1997,
97, 2005.
5 G. F. Swiegers and T. J. Malefetse, Chem. Rev., 2000, 100, 3483;
Coord. Chem. Rev., 2002, 225, 91.
6 N. Fatin-Rouge, S. Blanc, A. Pfeil, A. Rigault, A. M. Albrecht-Gary
and J. M. Lehn, Helv. Chim. Acta, 2001, 84, 1694; T. M. Garrett, U.
Koert, J. M. Lehn, A. Rigault, D. Meyer and J. Fischer, J. Chem.
Soc., Chem. Commun., 1990, 84, 557; J. M. Lehn and A. Rigault,
Angew. Chem., Int. Ed. Engl., 1988, 27, 1095; J. M. Lehn, A. Rigault,
J. Siegel, J. Harrowfield, B. Chevrier and D. Moras, Proc. Natl. Acad.
Sci., USA, 1987, 84, 2565; M. M. Harding, U. Koert, J. M. Lehn, A.
Marquis-Rigault, C. Piguet and J. Siegel, Helv. Chim. Acta, 1991, 74,
594.
25 E. C. Constable, M. G. B. Drew and M. D. Ward, Chem. Commun.,
1987, 1600.
26 K. T. Potts, M. Keshavarz-K, F. S. Tham, H. D. Abruna and C. R.
Arana, Inorg. Chem., 1993, 32, 4422; K. T. Potts, M. Keshavarz-K,
F. S. Tham, H. D. Abruna and C. Arana, Inorg. Chem., 1993, 32,
4436.
27 E. C. Constable, A. J. Edwards, P. R. Raithby, D. R. Smith, J. V.
Walker and L. Whall, Chem. Commun., 1996, 2551.
28 W. Dai, H. Hu, X. Wei, S. Zhu, D. Wang, K. Yu, N. K. Dalley and
X. Kou, Polyhedron, 1997, 16, 2059.
7 R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi, C. R. Woods and
J. S. Siegel, Eur. J. Org. Chem., 2001, 173; C. R. Woods, M. Benaglia,
F. Cozzi and J. S. Siegel, Angew. Chem., Int. Ed. Engl., 1996, 35, 1830.
29 A. J. Bard and L. R. Faulkner, Electrochemical Methods: Fundamen-
tals and Applications, John Wiley & Sons: New York, 1980.
D a l t o n T r a n s . , 2 0 0 5 , 1 1 6 8 – 1 1 7 5
1 1 7 5