ARTICLES
central diimide residues in these septet sequences are shielded washed with methanol and dried at 80 8C for 3 h. The crude product was purified
by column chromatography (dichloromethane:methanol 99:1, v/v), giving 1 as a
pale yellow, crystalline solid (0.36 g, 27%). M.p. 309 8C; H NMR (700 MHz,
directly by two, one and no pyrenyl units, respectively. This is a
very different situation from that described for complexation of 8
to the same sequences, where the central, ‘observed’ diimide
1
CDCl :hexafluoropropan-2-ol 6:1, v/v) d (ppm) ¼ 8.49 (s, 4H), 7.97 (d, J ¼ 8.6 Hz,
3
4
H), 7.90 (d, J ¼ 9.0 Hz, 4H), 7.77 (d, J ¼ 7.9 Hz, 2H), 7.75 (s, 2H),
protons are shielded directly by two pyrenyl residues in all cases. 7.71 (d, J ¼ 8.5 Hz, 4H), 7.70 (t, J ¼ 7.9 Hz, 2H), 7.66 (d, J ¼ 7.9 Hz, 2H),
7
7
.60 (t, J ¼ 8.1 Hz, 2H), 7.33 (dm, 2H), 7.19 (dm, 2H), 7.18 (t, J ¼ 2.1 Hz, 2H),
The novel sequence selectivity discovered in this work for tweezer
molecule 9 (that is, ‘ISI’ rather than ‘SIS’) thus provides a simple
explanation for the extraordinarily high sensitivity to this tweezer
13
.17 (d, J ¼ 9.0 Hz, 4H); C NMR (62.5 MHz, CDCl :hexafluoropropan-2-ol 6:1,
3
v/v) d (ppm) ¼ 165.7, 165.5, 162.4, 155.8, 144.9, 140.9, 137.4, 137.3, 134.3, 132.8,
1
1
32.3, 131.5, 131.3, 130.5, 130.4, 129.9, 128.9, 128.4, 126.3, 123.7, 123.6, 123.4, 121.3,
20.1, 118.9, 118.8; IR (Nujol): 1,774, 1,713 (imide nC ¼ O), 1,369 (nC–N), 1,241
1
of H NMR resonances associated with long-range sequence infor-
22
21
mation . Conversely, the ‘anomalous’ behaviour of copolyimide (nC–O–C), 1,107 cm ; MS (MALDI-TOF): m/z ¼ 1,357. Calcd. for
1
þ
[
(
C70H36F N O S þ Na] 1,357. Analysis (calcd., found for C
H NMR resonances in the presence of tweezer molecule 9 can
6
4
14
2
62.97, 62.64), H (2.72, 2.88), N (4.19, 4.12%). Oligomers 2 and 3 were synthesized
now be seen, in conjunction with Table 1, to provide strong evidence
that this molecule does indeed display a general preference for
binding to the triplet sequence ‘ISI’ rather than ‘SIS’.
and isolated by an analogous method, replacing pyromellitic dianhydride with
,4,5,8-naphthalenetetracarboxylic dianhydride.
1
In reality the situation is not quite so clear-cut, because even the
Synthesis of the ferrocenyl tweezer molecule 8. A suspension of 5-bromoisophthalic
diimide resonance arising from the ‘non-9-binding’ sequence acid (0.245 g, 1.00 mmol) in thionyl chloride (10 ml) was stirred at reflux under dry
nitrogen for 2 h. Evaporation of excess thionyl chloride under reduced pressure gave an
[
–SFSISFS–] does undergo a small upfield shift in the presence of
off-white solid. This residue was dissolved in dichloromethane (50 ml), and
-pyrenemethylamine hydrochloride (0.536 g, 2.00 mmol) and triethylamine (1 ml)
were added to the solution. The mixture was stirred at room temperature overnight,
this tweezer. It is already known that 9 can bind weakly to a
diimide residue , and because the NMR experiments described
1
22
here all show fast exchange between bound and unbound tweezer and the resulting solid was filtered, washed with water (3 × 30 ml) and methanol
molecules, it seems that the strong ‘ISI’ complexation proposed (2 × 40 ml), and dried to yield a cream crystalline solid (0.659 g, 95%). To a
suspension of this bromo-substituted tweezer molecule (0.336 g, 0.500 mmol) in
for tweezer 9 will be dynamically superimposed on much weaker
toluene (500 ml) was added tetrakis(triphenylphosphine)palladium(0) (0.029 g,
‘
SIS’ binding, with both mechanisms operating simultaneously on
0
.025 mmol), ferroceneboronic acid (0.120 g, 0.520 mmol) and sodium carbonate
0.106 g, 1.00 mmol) in water (5 ml). The mixture was heated under reflux for
stant K , measured (using the UV–Vis dilution method, based on 2 days, cooled and diluted with toluene (200 ml) and water (200 ml). The organic
the NMR timescale. Consistent with this, the 1:1 association con-
(
a
the charge-transfer band at 551 nm) for the binding of tweezer mol- phase was washed with 5% HCl, then with water, dried over MgSO , and evaporated
4
to dryness. The crude product was purified by column chromatography
(dichloromethane as eluent) to give 8 as a yellow, crystalline solid (0.284 g, 73%
ecule 9 to the chain-folding oligomer 2, was found in this work to be
2
1
21 22
780 M , more than five times higher than the value (140 M
)
of
1
yield), m.p. 243 8C; H NMR (250 MHz, DMSO-d ) d (ppm) ¼ 9.41 (t, J ¼ 5.4 Hz,
6
Ka for the binding of 9 to the simple diimide 10. In the context of
2H), 8.54 (d, J ¼ 9.3 Hz, 2H), 8.33–8.25 (m, 9H), 8.21 (d, J ¼ 1.5 Hz, 2H), 8.16 (s, 4H),
this 500% increase in binding constant, it should be noted that 8.15–8.04 (m, 4H), 5.28 (d, J ¼ 5.4 Hz, 4H), 4.88 (s, 2H), 4.38 (s, 2H), 4.01
13
the experimental error in binding constants determined by the (s, 5H); C NMR (62.5 MHz, DMSO-d
) d (ppm) ¼ 166.2, 140.3, 135.0, 133.2,
31.1, 130.7, 130.5, 128.5, 128.0, 127.8, 127.4, 127.3, 127.1, 126.6, 125.6, 125.5, 125.1,
24.6, 124.4, 124.3, 123.6, 83.7, 69.8, 69.7, 67.0, 43.5; IR (Nujol): 1,620 (nC ¼ O),
6
1
1
1
UV–Vis dilution method is normally of the order of 15% (ref. 44).
We have thus established, through studies of fully defined oligo-
mers and their tweezer complexes, that two different designs of
21
,376 (nC-N), 845 cm ; MS (MALDI-TOF): m/z ¼ 777. Calcd. for
þ
þ
[
C52H36FeN O þ H] , 777; for [C
H
FeN O þ Na] , calcd. 799; found 799;
2
2
52 36
2 2
tweezer molecules (exemplified by 8 and 9) have structures that Analysis (calcd., found for C52
H FeN O ): C (80.41, 79.95), H (4.67, 4.66),
36 2 2
are complementary to two different triplet sequences in chain- N (3.61, 3.59%).
folding polyimide sulfones. This situation is faintly reminiscent of
the relationship between the different tRNAs (each having a differ- Received 13 January 2010; accepted 10 May 2010;
ent ‘binding codon’) and mRNA (ref. 45). Correspondingly, homo- published online 27 June 2010
polyimide 12 can be represented as the sequence . . .SISISISISIS. . .,
with tweezer molecules 6, 7 and 8 binding to the triplets ‘SIS’, while
the present work shows that tweezer molecule 9 binds preferen- 1. Watson, J. D. & Crick, F. H. C. A structure for deoxyribose nucleic acid. Nature
tially to the sequence ‘ISI’, so that the ‘reading frame’ for 9 can be
thought of as being shifted to the left or right by a single
monomer residue (Fig. 7).
References
171, 737–738 (1953).
2
.
.
Wilkins, M. H. F., Stokes, A. R. & Wilson, H. R. Molecular structure of
deoxypentose nucleic acids. Nature 171, 738–740 (1953).
Franklin R. E. & Gosling, R. G. Molecular configuration in sodium
thymonucleate. Nature 171, 740–741 (1953).
3
Methods
4. Crick, F. H. C. On protein synthesis. Symp. Soc. Exptl Biol. 12, 138–163 (1958).
5. Nirenberg, M. Historical review: deciphering the genetic code—a personal
account. Trends Biochem. Sci. 29, 46–54 (2004).
6. Dawkins, C. R. The Blind Watchmaker (Longmans, 1986).
7. Lodish, H. et al. Molecular Cell Biology 6th edn (Freeman, 2007).
8. Burd, C. & Weck, M. Self-sorting in polymers. Macromolecules 38,
7225–7230 (2005).
9. Weck, M. Side-chain functionalized supramolecular polymers. Polym. Int. 56,
453–460 (2007).
10. Harmata, M. Chiral molecular tweezers. Acc. Chem. Res. 37, 862–873 (2004).
11. Kl a¨ rner, F.-G. & Kahlert, B. Molecular tweezers and clips as synthetic receptors.
Molecular recognition and dynamics in receptor2substrate complexes. Acc.
Chem. Res. 36, 919–932 (2003).
12. Chen, C. W. & Whitlock, H. W. Molecular tweezers—a simple model of
bifunctional intercalation. J. Am. Chem. Soc. 100, 4921–4922 (1978).
13. Zimmerman, S. C. & VanZyl, C. M. Rigid molecular tweezers—synthesis,
characterization and complexation chemistry of a diacridine. J. Am. Chem. Soc.
109, 7894–7896 (1987).
14. Kurebayashi, H., Haino, T., Usui, S. & Fukazawa, Y. Structure of supramolecular
complex of flexible molecular tweezers and planar guest in solution. Tetrahedron
57, 8667–8674 (2001).
Materials and instrumentation. Synthetic procedures were carried out under
an atmosphere of dry nitrogen, unless otherwise specified. Commercial solvents
and reagents were used without purification, unless otherwise stated. N,N-
(ref. 46), tweezer molecules 6 (ref. 43), 7 (ref. 43) and 9 (ref. 22), model imide 10
(
(
ref. 21), macrocycle 11 (ref. 46), homopolymer 12 (ref. 19) and copolymer 13
ref. 21) were prepared according to literature procedures. Proton and C NMR
13
spectra were recorded on Bruker AV-700 and DPX 250 MHz spectrometers,
respectively. Computational modelling (molecular mechanics with charge
equilibration, Cerius2, Accelrys, San Diego) was carried out on an SGI-O2
Single-crystal X-ray data for complexes [8 þ1] and [9þ11] were measured on
an Oxford Diffraction X-Calibur CCD diffractometer using Cu-Ka radiation.
Structure solution and refinement were carried out using the SHELXS-97 suite
of programmes.
Synthesis of oligoimides. The synthesis and isolation of oligomer 1 is given as an
example. A mixture of pyromellitic dianhydride (0.436 g, 20.0 mmol), diamine 4
(0.648 g, 10.0 mmol) and 3-(trifluoromethyl)aniline 5 (0.646 g, 40.0 mmol) in
DMAc (120 ml) was stirred at room temperature for 0.5 h to give a clear solution. To
this was added a mixture of acetic anhydride (1 ml) and pyridine (0.5 ml), and the
reaction mixture was heated at 100 8C under dry nitrogen for 14 h, before cooling
to room temperature and pouring into methanol. The precipitate was filtered,
15. Balzani, V. et al. Host2guest complexes between an aromatic molecular tweezer
′
and symmetric and unsymmetric dendrimers with a 4,4 -bipyridinium core.
J. Am. Chem. Soc. 128, 637–648 (2006).
659
©
2010 Macmillan Publishers Limited. All rights reserved.