Molecular threading

The successful synthesis of various catenanes by the strategy depicted in Figure 12 soon brought us to imagine a molecular trefoil knot synthesis by extending the former synthetic concept from one to two copper ions. As shown in Figure 14, two bis-chelating molecular threads (A) can be interlaced on two transition... 

Sauvage, J.P. (1990) Interlacing molecular threads on transition metals. Acc. Chem. Res. 23, 319-327. 

Let us now put the above principles into practice by considering the assembly of multiple-helical compounds. A simple chemical model for the formation of helicates involves the twisting of molecular threads , as shown in Figure 7-27. The incorporation of metal-binding domains into these threads allows the use of metal ions to control the twisting. 

The trick lies in recognising that the crossing points of the molecular threads correspond to a point at which the two threads are co-ordinated to a single metal ion. This would mean that the two helical structures in Figure 7-27 would be achieved by the incorporation of one- and two metal-binding domains, respectively. The first structure (7.44) arises from interaction with a single metal ion, the second (7.45) from interaction with two metal ions (Fig. 7-28). 

Figure 7-27. The twisting of molecular threads allows the assembly of helical structures. In compound 7.44 there is a single crossing of the molecular threads, whereas in 7.45 there are two such...

Figure 7-28. The use of metal ions to control the assembly of double-helical complexes. The twisting of the molecular threads is initiated by the co-ordination of metal-binding domains within the ligand to the metal ions. The assembly of the mononuclear compound 7.46 requires the incorporation of a single metal-binding domain in each molecular thread, whereas compound 7.47 requires two metal-binding domains per thread.

Figure 7-29. The generic features needed in a molecular thread designed to give structures such as 7.47, and an actual example of such a ligand, 7.48. A variety of spacer groups may be incorporated between the metal-binding domains.

Of course, it is quite possible to further extend these assembly processes to give doublehelical complexes with even more bond crossings. For example, a double-helical complex with three bond-crossings should result from the reaction of a molecular thread containing three metal-binding domains with three tetrahedral metal ions (Fig. 7-32). An example of the assembly of such a trinuclear double-helical complex is seen in the formation of 7.52 from the reaction of 7.51 with silver(i) salts (Fig. 7-33). 

Naturally, it is not necessary to limit the procedure to the use of tetrahedral metal centres. For example, it is possible to build double-helical structures from the interaction of molecular threads containing tridentate domains with metal ions possessing a preference for a six-co-ordinate octahedral geometry. An example of such a process is shown in Figu-... 

Figure 7-30. The reaction of 7.48 with copper(i) gives the desired double-helical dinuclear complex 7.49. The two molecular threads have been shaded differently in 7.49 to emphasise the helical structure.

Figure 7-31. The reaction of 7.50 with tetrahedral metal ions gives dinuclear double-helical complexes. The interplay of the metal and ligand requirements are emphasised in this process. Once again, the molecular threads have been shaded differently to clarify the structure of the product.

Figure 7-32. The interaction of a molecular thread containing three didentate metal-binding domains with tetrahedral metal ions should give a trinuclear double-helical complex.

Consider a difunctional molecular thread, with reactive sites at each end. These two reactive sites can react with some other difunctionalised molecule to give a cyclic structure (Fig. 7-37). This is exactly the type of process that we discussed at the beginning of Chapter 6. 

Now let us consider what happens if two such molecular threads containing didentate metal-binding domains are twisted into a helical arrangement after co-ordination to a tetrahedral metal centre. Reaction with the difunctional reagent could proceed in several ways. For example, the result could be the formation of a [2+2] macrocyclic complex as a result of the difunctional reagent linking together the two molecular threads (Fig. 7-38). 

Figure 7-37. The reaction of a difunctionalised molecular thread with another difunctionalised compound to give a macrocyclic product.

Figure 7-38. The formation of a co-ordinated [2+2] macrocyclic ligand in the reaction of a helical complex with a difunctional reagent. The difunctional reagent links together the two molecular threads. The black circles represent the sites at which the molecular thread has reacted with the difunctional reagent.

Figure 7-39. We have selected our molecular threads and our difunctional reagents such that it is only possible to form [1+1] cyclic structures. The consequence is the formation of a catenane, in which the two new cyclic molecules are interlinked. Removal of the metal ion would give the free catenane, but it is not possible to separate the two rings without the breaking of a bond.

Figure 7-40. Two views of the copper(i) complex of ligand 7.61 showing the arrangement of the reactive sites (indicated by arrows) such that cyclisation to give [1+1] macrocyclic products is favoured. One of the molecular threads has been shaded in each case.

To conclude this chapter, we will extend our investigation of metal-ion control over topology to the tying of molecular threads into knots. We commence by returning to some of our ideas about topology. The representation of a catenane in Figure 7-46 emphasises that a two-dimensional graph contains two points at which lines cross. 

Figure 7-47. Two cartoon views showing the formation of a catenane from a difunctional molecular thread. The first view emphasises the approach that we discussed in Section 7.5. The second view emphasises the formation of the precursor and the catenane in terms of the number of crossing points which must be drawn in the molecular thread.

Figure 7-50. The reaction of the molecular thread 7.67 with copper(i) salts gives a double-helical precursor with three crossing points. Reaction with ICH2(CH2OCH2)5CH2I gives the dinuclear trefoil knot 7.68.

Dietrich-Buchecker, C. 0., Sauvage, J.-P., Interlocking of molecular threads from the statistical approach to the templated synthesis of catenands. Chem. Rev. 1987, 87, 795-810. 

Preparation of the first rotaxanes was also accomplished by means of statistical threading of a chain through a cycle with following attaching of bulky substituents to the loose ends of the threaded chain [5], Statistical means of molecular threading were pushed out by much more efficient covalent templation procedures developed by Schill etal. [6,7],... 

The homoleptic Cu(l)2 complex and the bis-porphyrinic rotaxane 2+ both incorporate the same coordinating core Cu(dpp). In the rotaxane, one of the coordinating moieties (dpp) is incorporated in a macrocyde, which plays the role of a wheel. The second coordinating core is incorporated in the molecular thread - the axle of the rotaxane - the two ends of which bear the gold and the zinc porphyrin, respectively. 

The rotaxane 16 + (the rotaxane nomenclature used here is 16 "+, where N refers to the coordination number of the metal (4 or 5) and n to its charge) synthesized here is composed of two subunits a macrocycle and a molecular thread (Figure 15). The macrocycle1861 17 (Figure 15 (a) and (b)) possesses two different coordi-... 

Fig. 29. Template synthesis of the trefoil knot. The molecular thread is a metal-ligand that contains two chelating sites and the black dot represents a transition-metal ion in a pseudo-tetrahedral environment...

Figure 16 The PL spectra of anthracene (A) 10-substituted with a long molecular thread (ANTPEP 10-[3,5-di(terbutyl)phenoxy]decyl-2-( 2-[(9-anthrylcarbonyl)amino] acetate)) in a bisphenol A polycarbonate (PC) matrix at different concentrations shown in the figure. The PL spectrum in the dilute solution of dichloromethane (DCM) is displayed for comparison (curve 4). Molecular structures of the chemical compounds are shown in the upper part of the figure. Adapted from Ref. 94.

The first cyclodextrin-based rotaxanes were prepared by Ogino in 1981, l8a The idea was to use kinetically inert Co(III) complex fragments as stoppers. Therefore the molecular thread had to be functionalized with coordinating end groups. One of the synthetic routes is depicted in Figure 2.16a. 

Initial multithreading experiments involved macrocycle 58 and the various molecular threads of Figure 2.25, containing two chelating subunits.58 These... 

---