Lactide Cyclopolymerization by an Alumatrane-Inspired Catalyst
Control of molecular structure is an enduring motivation for chemists. From total synthesis(1) to self-assembly(2) to crystal growth,(3) the pursuit continues unabated. Polymer synthesis, in particular, has seen a revolution in control. It is now possible to precisely predetermine chain length, extent of cross-linking, comonomer incorporation, block length, stereochemistry and topology.(4) There has even been success in sequence control,(5) previously achieved only in biological systems. The difficulties inherent in the synthesis of pure macrocycle(6) have limited their availability, despite their compelling predicted properties. A variety of strategies have been employed but only two avoid linear material at all stages: catalyst-free ring equilibration and ring-expansion polymerizations (REP).(7) Our general catalyst design has been influenced by a recent example of the latter strategy wherein Grubbs used a ring-opening metathesis polymerization (ROMP) catalyst modified with a permanently tethered initiating group to polymerize cylcooctene yielding high-molecular-weight rings free of any linear contaminants.(8, 9) Taking a cyclic catalyst and a cyclic monomer as our design criteria, we focused on cyclic poly(lactic acid)(cPLA) as a target. PLA architectures are interesting due to their degradation profiles and because they can be renewably sourced, and the cyclic architecture of cPLA in particular may be advantageous in drug delivery.(10) Whereas cyclic lactide oligomer have been synthesized,(11-13) we sought to produce high-molecular-weight cPLA.(10)N-Heterocyclic carbenes,(14, 15) organotin compounds(16) and imidazoles(17) have previously been reported as effective catalysts for lactide REP to form cPLA. Herein we report an alumatrane-inspired catalyst, (N,N-bis(3,5-di-tert-butyl-2-benzyloxy)-2-(2-aminoethoxy)ethoxy)aluminum [(tBu-SalAmEE)Al, 1], that is active for controlled cPLA synthesis by lactide REP. Metallatranes, the trigonal bipyramidal complexes of tripodal tetradentate ligands,(18) have been widely used(19-21) and continue to be an area of active research.(22-28) Of particular relevance to this work are recent examples of monomeric alumatranes.(29-31) Alumatranes are most commonly dimeric, and the alumatrane dimers which have been examined for lactide polymerization are inactive.(32, 33) However, a monomeric alumatrane-isopropanol adduct capable of lactide polymerization in the melt has been reported.(32) Further inspiration was drawn from ligand hemilability where changes in hapticity(34) or denticity(35) can stabilize reactive catalytic intermediates or allow substrate coordination.