The discovery by Sinn and Kaminsky that methyl alumoxane (MAO) can act as an efficient activator for metallocene olefin polymerization catalysts [1] has triggered tremendous developments in single-site olefin polymerization catalysis. [2] Presently, more than twenty years after the initial discovery, the actual nature of MAO and its mechanism of catalyst activation are still under debate. [3] Model studies on tert-butyl alumoxanes led Barron et al. to formulate the hypothesis that these alumoxanes consist of oligomeric (RAlO) n clusters, containing 4-coordinate Al in strained fused 4-membered Al 2 O 2 rings that can exhibit ™latent Lewis acidity∫ by ring opening, allowing Al to abstract an alkyl anion from the transition metal dialkyl catalyst precursor. [4, 5] The direct (structural, spectroscopic) study of these processes and their products is hampered by the apparent equilibrium nature of the alkyl transfer reaction, with the equilibrium constant strongly favoring the starting materials when well-defined alumoxanes with sterically demanding alkyl groups are used. Crystal structure determinations of the products resulting from the reaction of [tBuAlO] 6 with MeLi [5] and with RNH 2 [6] gave support for the proposed activation mechanism, but as yet the products of the activation of metallocene single-site olefin polymerization catalysts by alumoxane activators have eluded full characterization.
Recently we reported the synthesis and structural characterization of a well-defined boralumoxane species, [tBu 4 -Al 4 Ar 4 B 4 O 8 ] (1, Ar 2,6-diisopropylphenyl), and showed that this compound is able to activate [Cp 2 ZrMe 2 ] for catalytic ethene polymerization. [7] Although 1 is topologically quite different from the tBu-alumoxane species isolated by Barron et al., containing 3-coordinate B and 4-and 5-coordinate Al, it shares with these compounds the presence of a strained 4-membered Al 2 O 2 -ring assembly (edge sharing with a BAlO 2 [