Particle formation and multiphase morphologies in catalytic aqueous ethylene polymerization

Abstract

Aqueous polymer dispersions are vastly applied as eco-friendly adhesives, coatings and paints. Such dispersions are preferentially accessed by aqueous radical emulsion polymerization. Polymers, which are traditionally produced by insertion polymerization using oxophilic early transition metal-based catalysts, such as polyolefins with controlled microstructure (e.g. linear polyethylene), are not directly available as aqueous dispersions. Such dispersions can be accessed as secondary dispersions by means of emulsification of polymer melt or solution in water followed by evaporation of the organic solvent in the second case. However, the solubility of polyolefins is poor even in hot organic solvents and the synthesis of polyolefin dispersions by secondary dispersion procedure is limited to more soluble branched polyolefins or low-melting polyolefin waxes. In order to access high molecular weight polyolefin dispersions a direct synthesis approach is desirable. Additionally, a direct synthesis of polyolefin dispersions in the presence of dispersed filler can potentially enable formation of nanocomposite materials. It was previously shown, that the catalytic polymerization of ethylene to high molecular weight products is also possible using less oxophilic late transition metal catalysts. For exam-ple nickel or palladium based catalysts are active in ethylene polymerization in aqueous environ¬ments.[ ] Polyethylene dispersions consisting of high molecular weight linear poly-ethylene can be accessed directly using neutral complexes such as [κ2-(N,O)Ni(Me)(TPPTS)] as water-soluble catalyst precursors.[1b] However, this class of Ni(II)-based catalysts also suffers from deactivation in the aqueous environment. This affects the productivity and control of molecular weight. To resolve this issue, a deeper understanding of deactivation mechanisms under true pressure polymerization conditions and also of particle formation was necessary. In Chapter 3 the activation, deactivation and stabilization mechanisms of the water-soluble catalyst precursor [κ2-N,O-{2,6-(3',5'-(F3C)2C6H3)2C6H3-N=C(H)-(3,5-I2-2-O-C6H2)}NiCH3{P(3-C6H4SO3Na)3}] (1) and the derived catalytically active species were studied in aqueous environment. Upon dissolution of the water-soluble catalyst precursor in water the catalyst precursor is transformed into a water-insoluble catalyst precursor by abstraction of the water-soluble labile ligand (TPPTS). By exposing the catalyst precursor to the aqueous environment prior to the polymerization it was shown that the dissolution of the catalyst precursor (or rather the resulting hydrophobic Ni-species) in the aqueous mixture can be improved by addition of an ionic surfactant. However, independent of the presence of the surfactant the catalyst precursor (or its dissociation product) is deactivated in the aqeous environment within 2h. Using a 13C-labeled catalyst precursor at true pressure conditions it was shown, that in the presence of ethylene the catalyst precursor is rapidly activated by ethylene coordination (after 30 s: 20–28%, after 2 min: ca 45%). By comparison of chain numbers, accessed from polymer yields using the polymer’s molecular weight, and particle numbers generated per nickel center, accessed from polymer yields and the average particle size it could be shown that at early stages of the polymerization each particle consists of ~ 1 polymer chain and 1 polymer chain was initiated per ~ 10 catalyst precursor molecules. On the other hand, after longer polymerization times almost all particles consist of a single polymer chain which was initiated by 1 catalyst precursor, meaning that in the first minutes of the reaction almost all catalyst precursor is activated. The polymerization proceeds by an ethylene insertion-coordination mechanism and under conditions studied (at 15 °C, 40 bar and pH = 7) linear high molecular weight polyethylene is formed. However, in neutral aqueous conditions irreversible catalyst deactivation is observed directly after the addition of ethylene (as concluded from a rapid decrease of ethylene consumption). In contrast to the polymerization in organic solvents using lipophilic analogs of (1), on a molecular scale, instead of single-unsaturated chains typically issued from ß-hydrid elimination and chain transfer, fully saturated polymer chains are produced in aqueous polymerizations. The predominant deactivation pathway is a hydrolysis (or rather protolysis) reaction of the growing Ni-polymeryl species (and the catalyst precursor) and at pH < 7 only traces of polymer can be isolated. This deactivation reaction can be partially suppressed by reduction of the proton concentration in the reaction mixture for example by addition of strong bases, like alkali hydroxides, but due to the amphiphilic nature of water, hydrolysis cannot be totally prevented. Nevertheless, polymers with Mn = 7.2  105 g·mol-1 and Mw/Mn = 1.2 are obtained at pH 12.5 vs Mn = 1.0  105 g·mol-1 Mw/Mn = 2.1 at pH 7 after 1 h using 50 µmmol L-1 catalyst precursor at 40 bar and 15 °C. Alternatively, it was shown, that using a less dissociated water analog, D2O, as a solvent the reaction has an almost living character (TOF ~ 4000 ethylene (catalyst precursor)-1 h-1 for more than 24 h and increasing molecular weight) resulting in stable particle dispersions with up to 26 wt % PE. The catalyst lifetime can be also extended by addition of (slightly) basic, weakly coordinating ligands, like DMF or DBU. Here, the reaction is predominantly affected by the coordination effect, that is, the polymerization rates depend on the concentration of the additive (e.g. at 0.65 mol L-1 DMF polymerization rates of ~ 1 unit s-1 are observed vs 8 units s-1 at 0.13 mol L-1 DMF). As outlined above, the activation of the catalyst precursor results in formation of polymer chains which are dispersed in the form of nanoparticles in the aqueous phase. In sampling experiments it was demonstrated that the formation of the polyethylene particles goes along with rapid adsorption of the surfactant on the particle surface and the presence of the surfactant is a prerequisite for the formation of single lamella particles. At the first minutes of polymerization hexagonal lamellae are observed which gradually transform into much larger single crystal lozenges or more likely hollow pyramids. These observations show that the polyethylene particles produced by aqueous dispersion polymerization can evolve into crystals which are similar to those formed by crystallization from highly dilute polyethylene solutions at low undercoolings. Another aspect of the formation of the PE nanoplatelets in aqueous environment is the mode of crystallization at strong undercoolings. The particle thickness of ~ 7 nm which was observed in previous studies was higher, than theoretically predicted for such an extreme undercooling as present at 15 °C.[ ] In sampling experiments followed by AFM and TEM analysis it could be shown that the nascent platelets are ~ 4 nm thin and their thickness increases to ~ 7 nm within first minutes of polymerization probably as a result of postcrystallization or lamellar thickening. To summarize, the above results provide an important insight into mechanisms which govern the effective polymerization of ethylene in aqueous environment, also showing the potential of the reaction for an effective large scale synthesis of polyolefin dispersions. Additionally, from the fundamental point of view, aqueous dispersion polymerization of ethylene has shown its potential as an adequate and elegant method for studying the formation and growth of polymer crystals at conditions not accessible by other methods. Polyethylene dispersions can also be used for the synthesis of nanocomposites. Based on the above insights on particle formation, in Chapter 4 multiphase polyethylene/polyethylene parti-cles consisting of polyethylenes with different microstructures, like linear, highly crystalline high molecular weight PE and amorphous, highly branched, low molecular weight PE or high molecular weight ethylene/norbornene copolymer were synthesized in a two-step method using appropriate Ni(II)-based catalysts. In the first step crystalline, 20 nm-large, single-lamel-la seed particles can be generated using a water-soluble catalyst precursor 1-TPPTS (from Chapter 3). Alternatively, seed particles can be obtained by miniemulsion polymeriza¬tion using lipophilic catalyst precursor analogs, giving in this case multilamellar almost spheri¬cal particles of 100–200 nm size. Here, the particle size can be partially controlled by the size of the miniemulsion droplets. In a second polymerization step the amorphous poly¬ethylene phase was synthesized using a second catalyst in the presence of the crystalline seed. When highly branched low molecular weight polyethylene was generated in presence of the crystalline nanoplatelets, core-shell particles were obtained. Besides, also some „cocrystalliza¬tion” of the low-molecular weight polyethylene with the crystalline core was observed. When larger, multilamellar crystalline particles were used as a core instead, low molecular weight polyethylene introduced in the second step “cocrystallized” with the core, forming needle structures on the surface of the core particles. Application of a high molecular weight, not-crystallizable ethylene-norbornene-copolymer in the second step led to the for¬mation of dumbbell-like structures consisting of one multilamellar crystalline particle and one amorphous copolymer particle. The obtained composites combine the high impact resistance of the crystalline part with the low minimum film formation temperature of the amorphous polymer and could be potentially applied as thin polyethylene coatings. Organic/inorganic polyethylene-based composites are also of interest, however, due to the high melting temperature and low solubility of polyethylene, production of such composites is often challenging. In Chapter 5 polyethylene nanocrystals were exemplarily used for the synthesis of polyethylene/graphene nanocomposites. Being a demanding 2-D material graphene does not easily disperse in viscous apolar media, like polyethylene melts. Application of highly dispersed polyethylene as matrix precursor enables the synthesis of homogeneous highly conductive polyethylene/graphene composites with low percolation thresholds. Starting from aqueous graphene dispersion, the polyethylene phase was introduced as a pre-synthesized dispersion or generated in situ in the presence of the dispersed graphene. After precipitation of the composite dispersion, followed by compression molding of the precipitated solid, composite samples were obtained. For such kinds of composites the quality of graphene (lateral extension, number of layers) strongly influences the properties of the composite (conductivity, homogeneity, percolation threshold). Using thermally reduced graphene with C-content ~ 94% polyethylene nanocrystal/graphene composites with conductivity σ = 2.6 × 10-5 S·cm-1 at ~ 2 wt % filler content were obtained, showing at the same time almost perfect filler distribution in the matrix. The synthesis protocol developed in this work can potentially be applied to any filler which can be dispersed in water and offers a direct approach to functional polyethylene composites and composite films derived thereof. Various aspects of the formation and application of the aqueous polyethylene dispersions which are presented in this work show, that aqueous polyethylene dispersions with a controlled microstructure are materials with a high potential which can be accessed at industrially relevant solid contents using readily accessible Ni(II)-based insertion polymerization catalyst precursors in appropriate conditions. The obtained dispersions are also suitable for the direct synthesis of multiphase polyethylene particles or even for the synthesis of highly conductive polyethylene/graphene nanocomposites. Thus, these findings reveal the possibility of direct syntheses of other functionalized polyethylene based nanomaterials in future. Introduction of comonomers could allow for even higher versatility of the polyethylene nanoparticles in composite applications.