(559f) Ru-Catalyzed Ring-Opening Metathesis Polymerization Route to Narrow-Distribution Polyethylene

Myers, S. B., Princeton University

Synthetic routes to the production of linear polyethylene with a narrow molecular weight distribution have long been of interest. In addition to direct polymerization of the ethylene monomer, alternate methods first synthesize an unsaturated precursor polymer which then yields polyethylene upon hydrogenation. Both the ring opening metathesis polymerization (ROMP) of unsubstituted cycloolefins and the classic living anionic polymerization of butadiene, each followed by hydrogenation, can create narrowly distributed polyethylene. The ROMP methods, however, avoid the short-branched structure afforded by the anionic polymerization of butadiene, and create ?perfect? polyethylene. Various synthetic methods to produce polyethylene precursors have been developed using ROMP, but require gaseous monomers [1], extreme reaction conditions (-45C ) [2], or very expensive catalysts [3].

Here, we evaluate the ?living? ROMP of cyclopentene using a robust, commercially-available, ruthenium-based initiator (?Grubbs first generation?), Ru(=CHPh)Cl2(PCy3)2, in the presence of excess tricyclohexyl phosphine (PCy3). Tricyclohexyl phosphine serves as a polymerization controller by limiting the activity of the Ru initiator. In solution, the PCy3 ligands on the initiator reversibly dissociate to yield an active catalyst site; the addition of excess phosphine shifts the ligand binding equilibrium towards attachment [4]. In the absence of excess phosphine, the polymerization of cyclopentene remains uncontrolled, exhibiting Mw/Mn values of 1.5-2. Through tuning the phosphine to initiator ratio, we are able to synthesize well-defined polymers of narrow molecular weight distribution (Mw/Mn = 1.05 ? 1.15) over a wide range of Mn (9000-100,000 g/mol). The excess phosphine appears to slow propagation relative to initiation, and reduce the frequency of secondary metathesis events, described in more detail below.

The reaction kinetics for this system are modeled as first-order in both effective monomer and initiator concentrations. Due to its low ring strain, cyclopentene ROMP is an equilibrium reaction. As a result, a 1.3M equilibrium monomer concentration is required for propagation, but is not available for reaction. To ensure that the monomer concentration remains well above 1.3M and avoid a broadening of the distribution through propagation-depropagation equilibrium, these polymerizations must be taken to relatively low conversion (15%). The effective initiator concentration depends on the fraction of time the initiator is active, which is directly related to the initiator-ligand binding equilibrium constant and the amount of excess phosphine present. Our data were adequately described by this model, which quantitatively captured the increase in molecular weight with increasing monomer to initiator ratio, increasing polymerization time, and decreasing phosphine to initiator ratio. At the low conversion employed, the model suggests little effect of the initiator concentration on the polymer molecular weight, which was confirmed experimentally.

Ideally, the molecular weight distributions of the polymer products would remain narrow across the entire range of suitable reaction conditions, but this was not universally observed. Most notably, high values of monomer to initiator ratio resulted in intramolecular metathesis, or macrocycle formation [5], observed through an increase in polydispersity with conversion, and an imposed limit on the achievable molecular weight. This effect was reduced by increasing the phosphine to initiator ratio, slowing the rate of intramolecular metathesis relative to propagation.

[1] Z. Wu, D.R. Wheeler, and R.H. Grubbs, J. Am. Chem. Soc., 114, 146 (1992).

[2] P. Dounis and W.J. Feast, Polymer, 37, 2547 (1996).

[3] S.T. Trzaska, L-B.W. Lee, and R.A. Register, Macromolecules, 33, 9215 (2000).

[4] C.W. Bielawski and R.H. Grubbs, Macromolecules, 34, 8838 (2001).

[5] K.J. Ivin, A.M. Kenwright, and E. Khosravi, Chem. Commun., 1209 (1999).