Recent advances in polyolefin manufacture have focused on the production of differentiated commodity polyolefins, specialty polyolefins, and polyolefins hybrids. What differentiates these new polyolefin types from commodity polyolefins is that their molecular architectures are much more complex and often contain long chain branches (LCBs), leading to unique properties that make them competitive with specialty polymers.
One approach to produce these novel polyolefins is to use one or two single-site catalysts in two CSTRs in series. The first CSTR is used to make semicrystalline polymer chains, some of which must be vinyl-terminated (macromonomers). These macromonomers are then incorporated, via terminal branching, onto the chains growing in the second CSTR, becoming LCBs. If the backbone and the macromonomer have different compositions, they are called cross-products. Since it is not possible to incorporate all macromonomers, the final polymer will consist of a complex mixture of linear chains made by the two catalysts, homogeneous-branched chains (that is, chains where the backbone and all LCBs are of the same type), and cross-product macromolecules. The cross-product will add rather special properties to the polymer and, depending on its molecular architecture, the final product may act as a thermoplastic elastomer (TPE). Developing polymer reactor models for different catalyst combinations can help understand the details of these complex syntheses and to control the properties and fractions of linear chains, homogeneous-branched chains, and cross-products.
Two mathematical models were developed in this thesis for the solution polymerization of olefins with two single-site catalysts to predict the microstructure of long chain branched polyolefins. The first model was developed for a semi-batch reactor and the second one for two CSTRs in series. The models can predict the fractions of different polymer populations made in CSTRs and semibatch reactors, as well as their average chain lengths and LCB densities. Simulation results show that CSTRs are more efficient than semi-batch reactors to make polymers with high LCB densities and/or cross product fraction.
Simulation results also show that to increase the weight percent of cross-product using a linear-catalyst and a LCB-catalyst, the rate of macromonomer formation of the linear-catalyst should be high. The fraction of cross-product can be increased even further when both catalysts are capable of incorporating macromonomers to form LCB-chains because; in this case, both catalysts can form cross-product chains. Monomer concentration has no effect on cross-product mass fraction and polydispersity index, but increasing monomer concentration will decrease LCB density and increase the average chain lengths. Catalyst deactivation also has a great impact on polymer properties: LCB density, polydispersity index, cross-product fraction, and average chain lengths will all decrease by increasing the catalyst deactivation rate of both catalysts.
Simulation results for two CSTRs in series shows that increasing residence time in the second CSTR will lead to higher cross-product formation and LCB density. This rate of increase is more significant if the residence time in the second CSTR is similar to that of the first CSTR. The catalyst feed policy also has a great impact on polymer properties. We found out that feeding the linear-catalyst and the LCB-catalyst in equal amounts to the first CSTR and just adding the LCB-catalyst to the second CSTR is the preferred catalyst injection method for making polymer with a high mass fraction of cross-product, high chain length averages, and lower polydispersity index (PDI).
These simulation studies indicate that detailed polymerization kinetics for each catalyst is needed in order to synthesize these novel polyolefins. In the experimental part of this thesis, ethylene polymerization kinetics studies were performed first with two individual metallocene catalysts, then with both of them simultaneously.
First, ethylene polymerization with rac-Et(Ind)2ZrCl2/MAO was carried out in a semi-batch solution reactor. Reaction temperature, monomer, MAO, and catalyst concentrations were the factors studied to establish a framework to predict catalyst decay, polymer yield and molecular weight averages. The polymerization order with respect to ethylene and catalyst concentration was found to be first order. Chain transfer to monomer was the dominating chain transfer reaction while β-hydride elimination was negligible. An increase in MAO concentration led to a decrease in molecular weight. Catalyst decay could be described with a first order mechanism. At low MAO concentration this catalyst could make polymer with about one vinyl group per chain.
A similar ethylene polymerization kinetics study using dimethylsilyl(N-tert-butylamido)-(tetramethylcyclopentadienyl)-titanium dichloride (CGC-Ti)/MAO system showed that the polymerization order with respect to catalyst concentration was first order, but first order catalyst decay failed to explain catalyst deactivation. The polymerization order with respect to ethylene concentration was not unity for the whole range of ethylene concentration. The trigger mechanism, along with reversible first order activation and deactivation with MAO and first order thermal decay, could explain the effect of time, monomer and catalyst concentration on the rate of polymerization. Decrease in MAO concentration increased the amount of polymer chains with terminal vinyl groups and consequently led to polymers with LCBs. Decreasing monomer concentration at low MAO concentration also led to production of polymer chain with more long chain branching.
Ethylene homopolymerization and copolymerization with 1-octene were conducted using combined catalysts system at low and high MAO concentrations. Reactivity ratios were calculated and polymer samples with bimodal MWDs were obtained but no increase in LCB frequency or cross product formation was detected using carbon-13 nuclear magnetic resonance (13C NMR) and high-temperature gel permeation chromatography (GPC) coupled with a viscosity detector.
In order to promote the formation of cross-product macromolecules, 1,9-decadiene was copolymerized with ethylene using the Et(Ind)2ZrCl2/MAO to make tailored macromonomers with pendant 1-octenyl branches. The macromonomers ranged from having 1 to 6.5 vinyl groups per chain. These macromonomers were then incorporated into growing ethylene/1-butene or ethylene/1-octene copolymer chains using a titanium-based constrained geometry catalyst (CGC-Ti) to form branch block polymer chains with amorphous main backbone having short chain branch density (SCBD) up to 50 per 1 000 carbon atoms, and high crystalinity long chain branches with SCBD up to 3/1000 C atoms (cross product). Increase in polymerization time or catalyst concentration in the second stage of polymerization was observed to increase the cross-product weight fraction. We also observed that an increase in ethylene pressure during the second stage of polymerization, while 1-butene concentration was constant, favoured the formation of cross product. When 1-octene was used as comonomer in the second stage of polymerization, the presence of more pendant vinyl groups in the macromonomer led to increased long chain branching.
Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:OWTU.10012/6041 |
Date | 24 June 2011 |
Creators | Mehdiabadi, Saeid |
Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
Language | English |
Detected Language | English |
Type | Thesis or Dissertation |
Page generated in 0.0024 seconds