Ethylene polymerization using catalysts derived from activation of zirconocene aluminohydride complexes, supported on silica, pretreated with methylaluminoxane (MAO), is described. The novel catalyst compositions were compared to those using conventional zirconocene dichloride complexes and characterized by SEM/EDX and DRIFT spectroscopy. Supported catalysts were prepared which featured various surface Al:Zr ratios. When using excess MAO as both activator and scavenger, the catalysts containing the most Zr per g of support gave rise to the most active formulations; the high activities in the presence of excess MAO are due, in part, to catalyst leaching prior to and/or during polymerization. When tri-isobutylaluminum (TIBAL) was used as scavenger, the supported catalysts that featured a higher surface Al:Zr ratio had higher activity than those prepared at the lower Al:Zr ratios. The activity of the aluminohydride complexes was significantly higher than that of the corresponding dichloride complexes, when activated by MAO while the in the presence of TIBAL, there was little difference in performance between the two catalyst precursors.Active supported catalysts for ethylene polymerization are derived from reaction of zirconocene aluminohydride complexes with methylaluminoxane-treated silica. Depending on the method of activation, polymerization activity and polymer MW are significantly increased compared with the corresponding dichloride complexes.

The use of the chelating diboranes o-C6F4[B(C6F5)2]2 (1) and o-C6F4(9-BC12F8)2 (2: 9-BC12F8 = 1,2,3,4,5,6,7,8-octafluoro-9-borafluorene) for the polymerization of isobutene (IB) in aqueous suspension or in hydrocarbon solution was studied. Polymerizations in aqueous suspension provided polymer of moderate MW and at variable conversion and were dependent on temperature, mode of diborane addition, the presence of surfactant, and the acidity of and nature of the anion present in the aqueous phase. The T dependence of MW over the T range −80 to −20 °C was studied in aqueous suspension, and higher MW polymer was formed at lower T. The hydrolysis and methanolysis of diboranes 1 and 2 was studied by NMR spectroscopy. Reactions of diborane 1 with excess MeOH or water afford solutions containing oxonium acids [o-C6F4{B(C6F5)2}2(μ-OR)][(ROH)nH] (7: R = H, n > 2; 3: R = Me, n = 3). When diborane 1 is present in excess over water or MeOH, degradation of the diborane is observed. In this case the products are o-C6F4{B(C6F5)2}H (5) and (C6F5)2BOH 7 or (C6F5)2BOMe 4, respectively. In the case of diborole 2, o-C6F4(9-BC12F8)B(2-C12F8-2′′-H)(μ-OH)·7H2O (17) and o-C6F4(9-BC12F8)B(2-C12F8-2′′-H)(μ-OMe) (11) were isolated from reactions of 2 with water and MeOH, respectively, and were characterized by X-ray crystallography. None of these degradation products effect IB polymerization in aqueous suspension. As a model for initiation of polymerization, the reaction of diborole 2 with 1,1-diphenylethylene (DPE) was studied. Addition of MeOH at low T results in efficient formation of the ion-pair [Ph2CMe][o-C6F4(9-BC12F8)2(μ-OMe)] via protonation of DPE. Polymerizations in hydrocarbon media were exothermic and rapid and gave quantitative yields of polymer even at very low concentrations of diborane 1. The T dependence of MW was studied in hydrocarbon solution and showed non-Arrhenius behavior. This was explained by competitive chain transfer to monomer at elevated T and chain transfer to molecular water at lower T.

Modeling of ethylene polymerization using density functional theory was undertaken for both generic and substituted nickel iminophosphonamide (PN2) and amidinate (CN2) complexes. The more highly substituted complexes were studied using quantum mechanics/molecular mechanics (QM/MM) techniques so as to probe the role of steric effects on insertion and chain-transfer processes. For the generic systems H2P(NSiH3)2NiR(L) and HC(NSiH3)2NiR(L) (R = alkyl; L = C2H4), insertion had a higher barrier in the PN2 versus CN2 complex. The energy of ethylene binding was strongly affected by the nature of the R group. This was shown to be a function of agostic stabilization of the alkyl group in the absence of monomer. Insertion barriers are also strongly dependent on the nature of the alkyl group, particularly in the case of the sterically hindered Keim catalyst, which was modeled by (Me3Si)2NP(Me)(NSiMe3)2NiR(L) and QM/MM techniques. Degenerate chain transfer was systematically studied in the case of the generic CN2 complex HC(NSiH3)2NiEt(C2H4) and proceeds through five-coordinate intermediates with distorted trigonal-bipyramidal geometries. The highest-energy intermediate corresponds to a bis(ethylene)−NiH complex, where loss of ethylene would constitute (degenerate) chain transfer. Intermediates in the analogous PN2 complexes lie higher in energy, and thus these complexes should provide higher molecular weight material, as observed experimentally. β-H elimination/chain walking was also investigated using both generic and substituted complexes. The ground states in these reactions are agostic alkyls, while the ethylene−NiH complex, in which ethylene is perpendicular to the square plane, is a weakly bound intermediate. These intermediates are related to those formed during chain transfer by binding of the monomer.