Isomeric pyridyl-thiazole donor units for metal ion recognition in bi- and tri-metallic helicatesw

Underlying recognition phenomena in multi-component self-assembly processes can be exploited for the site-specific inclusion of different metal ions into polymetallic arrays. In the field of metallosupramolecular chemistry this is typically accomplished by programming either (i) the denticity of or (ii) the nature of the donor atoms in the binding sites of a polytopic ligand, prior to self-assembly with target metal ions. Segmental oligo-N-heterocyclic ligands that contain both tridentate and bidentate chelates, or the ability to partition as such, exemplify the first of these approaches. They form double-stranded hetero-bimetallic helicates with metal monoand dications due to the preference of the former (e.g. Ag, Cu) for tetrahedral coordination geometry and that of the latter (e.g. Cu, Co, Fe) for octahedral coordination geometry. When two such ligands are arranged appropriately, e.g. a co-aligned head-to-head (HH-) fashion in the case of a simple ditopic ligand, the resulting assembly features fourand six-coordinate sites into which the respective M and M ions are accommodated according to their respective geometry preferences. The second strategy— namely, the use of a polytopic ligand whose binding sites vary with regards to the nature of their donor atoms—has been applied with considerable success by Piguet and Bunzli et al. for the challenging task of selectively incorporating different lanthanide (Ln) trications into triple-stranded helicate arrays. They demonstrated that the self-assembly of ditopic ligands comprised of a triimine N3and a diimine/amide N2O-domain, with various Ln–Ln0 pairs, can produce up to 90% of the desired hetero-bimetallic helicate. Indeed, extensions of this work further lead to the isolation of triand tetranuclear hetero-bimetallic complexes in much higher yield than predicted from a purely statistical standpoint. A related approach saw the use of a ditopic catechol/thiocatechol ligand which, upon reaction with Ti and Mo, formed a heterobimetallic triple-stranded helicate. In this paper we use a new class of ditopic segmental pyridyl-thiazole (py-tz) N-donor ligand (Scheme 1) to demonstrate an alternative strategy for selectively introducing different metals into polynuclear arrays. The simplest of these ligands, L, contains two tridentate N3 binding domains which are structural isomers of one another. Self-assembly with Hg or Zn ions gives various isomers of a dinuclear double-stranded complex in solution. In the presence of both ions, however, only one species is formed in which the py-py-tz sequences bind to Zn and the py-tz-py sequences bind to Hg. The metal/site specificity is attributed to the divergent nature of the three N-donors in each tridentate domain, which varies according to the order in which the heterocycles appear in the sequence. When combined with a simple bidentate chelate for binding tetrahedral Cu, the Zn/Hg recognition effects displayed by the two tridentate units can be exploited for the self-assembly of a Zn/Hg/Cu hetero-trimetallic helicate. Ligand L (Scheme 1) was prepared by reaction of its methylene hydroxyand chloro-substituted py-py-tz and py-tz-py constituents, respectively, in a Williamson ether synthesis. Purification of the reaction mixture gave L as a colourless solid (see ESIw). Reaction of L with one equivalent of Zn(ClO4)2 6H2O in MeCN gives a colourless solution for which ESI mass spectroscopy showed an intense peak at m/z 1469 corresponding to the dizinc(II) species [Zn2(L )2(ClO4)3] . The H NMR (CD3CN) spectrum features two major sets of resonances, in addition to a third minor set which accounts for o5% of the total ligand (peak integration also suggests that one of the