2005 Elsevier. Reprinted with permission.

Transcription

1 IV Paavola, S., Löfgren, B., and Seppälä, J. V, Polymerization of hydroxyl functional polypropylene by metallocene catalysis, Eur. Polym. J. (2005) 41(12) Elsevier Reprinted with permission.

2 European Polymer Journal 41 (2005) EUROPEAN POLYMER JOURNAL Polymerization of hydroxyl functional polypropylene by metallocene catalysis Santeri Paavola, Barbro Löfgren, Jukka V. Seppälä * Polymer Technology, Helsinki University of Technology, P.O. Box 6100, FIN HUT, Finland Received 19 April 2005; received in revised form 11 May 2005; accepted 12 May 2005 Available online 31 August 2005 Abstract Propylene was copolymerized with 10-undecen-1-ol with use of dimethylsilanyl-bis-(2-methyl-4-phenyl-1-indenyl)zirconium dichloride as catalyst activated with methylaluminoxane (MAO) and triisobutylaluminum (TIBA). Comonomer incorporations as high as 2.0 mol% or 8.2 wt% were obtained without serious activity losses. Concentration of MAO, aluminum/comonomer ratio and pressure had some effect on polymerization activity and yield. However, changing the proportion of MAO in the cocatalyst mixture of MAO and TIBA proved to be most efficient way to enhance polymerization activity. Still, the result was a compromise between high functionality content, polymerization activity and molecular weight. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Propylene; Alcohol; Metallocene; Triisobutylaluminum; Methylaluminoxane 1. Introduction Polypropylene is rigid inexpensive material with excellent chemical resistance, impact strength and other properties typically associated with technical plastics. However, suitable applications are restricted by the lack of reactive groups. Introducing polar groups into normally non-polar polypropylene improves adhesion, paintability, affinity for dyes and printing agents, and compatibility [1]. Direct polymerization of polar comonomers using Ziegler Natta catalysts is limited by the intolerance of the catalysts to Lewis bases, which leads to catalyst * Corresponding author. Tel.: ; fax: address: jukka.seppala@hut.fi (J.V. Seppälä). deactivation, polymer degradation, and comonomer homopolymerization. Metallocene catalysts have an excellent comonomer response and give uniform comonomer distribution and stereoregular polypropylene with for example long linear a-olefins [2], but catalyst activity is not maintained when there is a functional group at the end of the long comonomer chain. Nevertheless, metallocene catalysts offer a way to polymerize alcohol functional comonomers with propylene without total catalyst decomposing [3 5]. In addition, the flexibility of metallocenes has enabled controlled polymerization with different polymerization mechanisms producing polypropylene-b-polymethylmethacrylate diblock copolymers [6]. Even with metallocenes the Lewis acid components (Zr, Al) tend to form complexes with the non-bonded electron pairs of heteroatoms in preference to react with p-electrons of the double bond of the incoming /$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi: /j.eurpolymj

3 2862 S. Paavola et al. / European Polymer Journal 41 (2005) monomer. The reaction with heteroatoms can best be eliminated by selecting catalyst components that are inert to functional groups, using an excess of aluminum alkyl compound to change hydroxyl groups into alkoxyl groups, or insulating the double bond from the heteroatom with a spacer group [4]. The undesired reaction can be further minimized by increasing steric hindrance about the heteroatom [4] or reducing the electron-donating ability of the heteroatom. A detailed survey of different routes to functional polypropylenes by exploitation of metallocenes was carried out by Sivaram and coworkers [7]. Pretreatment time has only minor effect on polymerization activity [5], but the effect of the ratio of aluminum to comonomer is not that thoroughly investigated. Many claim that stoichiometric ratio of triisobutylaluminum (TIBA) [8] or trimethylaluminum (TMA) should be sufficient, which would mean a total Al/comonomer ratio of 3 or 4 if commercial methylaluminoxane (MAO) is applied [9]. This maximizes the reactivity of the comonomer but at the expense of polymerization activity. In addition, the exact structure of the aluminum alkyl affects greatly polymerization activity as well as reactivity of the functional comonomer [10]. Our earlier results [11] demonstrated the possibility to use metallocenes to produce polypropylene containing functional groups, which even in small quantities enhanced adhesive properties. Although replacing major proportion of MAO with TIBA had many positive effects, we were not able to decrease aluminum/comonomer ratio below five without activity losses. We suggested that the loss in activity with the increase in comonomer concentration from 3 mmol/l to 30 mmol/ L might have been caused by decrease of ratio of aluminum to zirconium below reasonable level of In this work we aim to offer further evidence of that the comonomer concentration in the feed can be increased to 30 mmol/l without loss of activity if the concentration of MAO is simultaneously adjusted. In addition, the possibility to decrease the amount of TIBA closer to equimolar quantity of comonomer is investigated. 2. Experimental 2.1. Materials The catalyst used in the polymerizations was dimethylsilanyl-bis-(2-methyl-4-phenyl-1-indenyl)zirconium dichloride (Boulder Scientific Company). The comonomer, 10-undecen-1-ol, was obtained from Fluka and dried with molecular sieves before use. Toluene (p.a.) was obtained from Merck and purified by circulating it through 3-Å molecular sieves, BASF R3-11 copper catalyst, and activated Al 2 O 3 columns. Further moisture was removed by refluxing the toluene over sodium and benzophenone and distilling it under nitrogen before use. Methylaluminoxane (MAO, a 10 wt% solution in toluene) and triisobutylaluminum (TIBA) were obtained from Crompton Corporation and used as received. Propylene (grade 2.8 from AGA) was purified by conducting it through columns containing 3-Å molecular sieves, BASF R3-11 copper catalyst, and activated Al 2 O Polymerization procedure The polymerization experiments were carried out in a 0.5 L Büchi stainless steel autoclave reactor equipped with a propeller-like stirrer. The reactor was evacuated and flushed with nitrogen several times before the addition of toluene, MAO and TIBA. The total volume of the polymerization medium was 300 ml. After addition of the polar comonomer, the mixture was stirred for 15 min to allow the possible formation of complexes between comonomer and aluminum alkyl. The reaction mixture was then saturated with propylene. The polymerization was initiated by the addition of the toluene solution in which the catalyst was dissolved. The stirring speed was 500 min 1, polymerization temperature 70 C or 80 C, and propylene overpressure 1.0 bar, 2.0 bar, or 3.0 bar. Polymerization temperature was kept constant during reaction by circulating water in the reactor jacket. Partial pressure of the monomer was maintained constant with an electronic pressure controller. Propylene consumption was monitored during the polymerization. The polymerization was stopped by cutting off the propylene feed. The reactor was degassed, and the polymer was precipitated with a mixture of ethanol and a small amount of concentrated hydrochloric acid. The product was isolated by filtering, washed with ethanol and acetone, and dried in vacuum at 40 C over night Polymer characterization The molecular weights and molecular weight distributions of the polymers were determined with a Waters Alliance GPCV 2000 gel permeation chromatograph operating at 140 C and equipped with four Waters Styragel HMW columns (HMW 7, 2 HMW 6, and HMW 2) and a refractive-index detector. The solvent, 1,2,4-trichlorobenzene, was applied at a flow rate of 1.0 ml/min. The columns were calibrated with narrow molar mass distribution polystyrene standards using a universal calibration method. Polypropylene standard with known molar mass value was used as a reference in the selection of Mark-Houwink parameters K and a for the samples. Melting temperatures (T m ) and enthalpies (DH m ) were determined with a Mettler Toledo DSC 821 e differential scanning calorimeter. Indium was used for the calibration of the temperature scale. The melting

4 S. Paavola et al. / European Polymer Journal 41 (2005) endotherms were measured upon reheating of the polymer sample to 190 C at a heating rate of 10 C/min. The 1 H NMR spectra of the polypropylene copolymers were recorded on a Varian Gemini MHz spectrometer at 120 C from samples dissolved in deuterated 1,1,2,2-tetrachloroethane-d 2. The solvent resonance was used as an internal reference and was assigned the chemical shift d 5.91 ppm. In the 13 C NMR measurements 1,2,4-trichlorobenzene/benzene-d 6 (90/10 w/w) was used as solvent and the [mmmm] methyl signal of polypropylene was used as an internal standard and was assigned the chemical shift d 21.8 ppm. 3. Results and discussion Polymerization conditions and results are presented in Table 1. The comonomer was the linear alcohol 10- undecen-1-ol, where a spacer reduces the interactions between the hydroxyl group and the reactive double bond. Three different propylene pressures were used: 1 bar, 2 bar, and 3 bar. A high concentration of aluminum alkyl was needed for complex formation with the alcohol comonomer to prevent catalyst deactivation. Typical aluminum/comonomer ratio was 4 but also lower ratio of 2 was tried. The aluminum/zirconium ratio was calculated from MAO as TIBA did not activate the catalyst. Two different MAO concentrations were tested: 10 mmol/l and 20 mmol/l. With higher concentration it was ensured that the Al/Zr ratio was at sufficiently high level for active polymerization. The rest of aluminum alkyl was TIBA ( mmol/l) so that the desired aluminum/ comonomer ratio was reached. An increase in comonomer incorporation lowered melting temperatures. Comonomer incorporations were determined from the 1 H NMR spectra. Two hydrogens attached to the carbon next to hydroxyl give a clear triplet at d ppm, which was used in determining the comonomer incorporations. The 1 H NMR spectrum of hydroxyl functional polypropylene (F3) is depicted in Fig. 1. Highest comonomer incorporations were achieved at the lowest propylene pressure. The obvious reason is the lower concentration of propylene monomer in toluene compared to that of the competing functional comonomer. Functionality content depended also on the concentration of comonomer in the feed and the concentration of MAO played some role as well. The 13 C NMR spectrum of hydroxyl functional polypropylene (F1) is depicted in Fig. 2. The side chain carbons were clearly visible at d 33.9 ppm (br), d 44.0 ppm (a), d 36.2 ppm (C9) and the carbons next to the hydroxyl group at d 63.0 ppm (C1) and offered further evidence of the reaction of the comonomer. Concentration of MAO, aluminum/comonomer ratio and pressure all affect polymerization activity and yield. Table 1 Polymerization results a M w /M n Comon. incorp. c M w, g/mol Yield, g Activity b T m, C DH m, kj/mol Al/Zr (MAO), mol/mol n (Zr), lmol Al ratio to comonomer, mol/mol TIBA, mmol/l MAO, mmol/l Alcohol, mmol/l Run Pressure, bar F , , /8.2 F , , , /3.6 F , , /4.1 F ,500 d , /6.0 F , , Traces F , Traces F7 e 3 (80 C) , , /2.3 Reactor volume 0.5 L, 300 ml toluene, polymerization temperature 70 C, polymerization time 30 min, stirring speed 500 rpm. b Activity measured as kg PP/(molZr*h). mol%/wt%. d Polymerization time only 25 min. Polymerization temperature 80 C [11]. a c e

5 2864 S. Paavola et al. / European Polymer Journal 41 (2005) Fig H NMR spectrum of functional polypropylene (F3). Solvent d 5.91 ppm. α br b a C9 c n C1 OH ppm Fig C NMR spectrum of functional polypropylene (F1). Solvent d ppm. When MAO concentration was doubled from 10 mmol/ L to 20 mmol/l, activity increased more than twofold. Although the activities are not comparable due to the differences in polymerization time, the yield was increased even when shorter polymerization times and almost half amount of catalyst was used. As suggested in an earlier publication [11], reasonable activities were achieved by keeping MAO/Zr ratio above 2000 and the activity was by no means limited by high comonomer concentration of 30 mmol/l in feed.

6 S. Paavola et al. / European Polymer Journal 41 (2005) Aluminum/comonomer ratio had no significant effect on the activity if it was kept above four. When it was decreased from four in run F5 to two in run F6, the activity dropped dramatically to less than 2% of the activity in run F5 and yield was almost zero. TIBA does not solely form complexes with the alcohol comonomer, but has also other functions. Even though TIBA does not activate the catalyst, it acts as scavenger for impurities and is therefore consumed. Hence, stoichiometric aluminum/comonomer ratio, which is frequently in use, is not sufficient. The effect of pressure on molecular weight and comonomer incorporation is depicted in Fig. 3. Molecular weight was almost proportional to propylene pressure. Pressure change from 1 bar to 2 bar and further to 3 bar caused molecular weight to increase from 53,000 g/mol to 104,000 g/mol and further to 132,000 g/mol. In contrast, the comonomer incorporation decreased from 2 mol% to 0.9 mol% and further to just traces of comonomer. At elevated pressure, the propylene concentration in toluene was higher and that was why molecular weight increased and the comonomer had to compete with propylene monomer to gain access to the growing polymer chain. High aluminum alkyl concentrations and especially increasing MAO concentration from 10 mmol/l to 20 mmol/l resulted in a gel-like network between toluene and the aluminum alkyl. This phenomenon was enhanced by low temperature, high yield and molecular weight of the produced polymer as the product was no longer soluble in toluene. Therefore it was impossible to conduct polymerizations using high MAO concentrations at high pressure. Comonomer incorporation [mol%] Polymerization pressure [bar] Molecular weight [g/mol] Fig. 3. Dependence of comonomer incorporation (as columns) and molecular weight (as points j) on polymerization pressure. One goal was to perform a full series of polymerization where only pressure would change. Unfortunately, because of the gel formation with high MAO concentration at high pressure, it was not possible, but the effect of pressure could still be evaluated by comparing runs F1 to F2 and F3 to F5 done in different pressures but otherwise at similar polymerization conditions. The pressure affected the polymerization activity and the molecular weight in a similar way. Polymerization activity increased steadily along with pressure. When pressure was doubled from 1 bar to 2 bar, the activity increased by 70%. When the pressure was further elevated to 3 bar, the activity increased still 30% over the level at 2 bar. The initial reaction between the aluminum alkyl and the comonomer is immediate, but slightly higher polymerization activities were observed by Aaltonen et al. [5] when pretreatment time was extended to several hours. However, the reactivity of the comonomer was not enhanced. Long pretreatment times are impractical and therefore uncommon in industrial processes. The exact structure of MAO or the complex of aluminum alkyl and comonomer is still somewhat obscure. Turunen et al. [12] reported that with excess of aluminum alkyl, 10-undecen-1-ol (HOR) tends to form dimeric complexes [R 0 2 AlOR] 2 and [R 0 AlðORÞ 2 ] 2 with aluminum alkyls (R 0 3Al) instead of tetrameric aluminum compound containing a central six-coordinated aluminum atom. This more stable tetrameric aluminum compound was more common with equimolar aluminum/ comonomer ratio and became more dominant when the amount of comonomer was increased. Polymerization results (F5) were also compared to our earlier results (F7) obtained at higher polymerization temperature, but otherwise at similar conditions. There was no change in polymerization activity, but molecular weight was almost 50% higher as more propylene is dissolved into toluene at lower temperature. This caused also the drastic decrease in comonomer incorporation up to only traces of comonomer. Polypropylenes with high functionality contents suitable for various applications could still be obtained at this lower temperature by adjusting polymerization pressure and MAO proportion in the cocatalyst mixture. Higher functionality contents might be needed for use as compatibilizer in polymer blends [13] or adherent in polymer composites [14]. Another possible route to a high functionality content might have been to increase the concentration of comonomer in feed. On the other hand, even very low functionality contents were sufficient for enhancing adhesion towards metals and coatings [11] why there might not be need for increase in comonomer incorporation from this barely detectable level. Looking at the properties of polypropylenes instead of the polymerization conditions, the polypropylene F7 prepared at higher temperature had more

7 2866 S. Paavola et al. / European Polymer Journal 41 (2005) resemblance with F2, regardless of the difference in polymerization temperature, pressure and MAO concentration. Polymerization is limited by several factors. The polymerization activity has to be reasonable high in order to achieve some polymer, and therefore the aluminum ratio to comonomer has to be kept over critical level as well as ratio of MAO to zirconium. Also the polymerization pressure limits activity. However, if the TIBA concentration and especially the MAO concentration are high, these aluminum alkyls tend to form gellike network with toluene at least in low polymerization temperatures. On the other hand, the molecular weight has to be at sufficiently high level for technical applications and increase in temperature has a negative effect on the molecular weight even in the presence of TIBA. Still, apart from keeping major portion of aluminum alkyl as MAO instead of TIBA, all these improvements in the polymerization conditions are done at the expense of the desired high functionality content. The result is a delicate balance between polymerization activity, molecular weight and functionality content. 4. Conclusions Propylene-co-10-undecen-1-ol copolymers were successfully polymerized with comonomer incorporations of 2.0 mol% or 8.2 wt% at highest. Copolymerization was confirmed with NMR and DSC measurements. It was possible to polymerize functional polypropylene with reasonable activity by adjusting the proportion of MAO in the mixture of MAO and TIBA used as cocatalyst. Further increase in comonomer incorporation was able to be performed at the expense of molecular weight and activity by decreasing the propylene concentration, i.e. polymerization pressure. Hence, a compromise between polymerization activity, functionality content and molecular weight must be searched in order to obtain functional polypropylene with optimal properties for different technical applications. Acknowledgements Funding from the Technology Development Centre (TEKES) and the Academy of Finland (SA ) is gratefully acknowledged. References [1] Mülhaupt R, Duschek T, Rieger B. Makromol Chem, Macromol Symp 1991;48/49: [2] Van Reenen AJ, Brull R, Wahner UM, Raubenheimer HG, Sanderson RD, Pasch H. J Polym Sci, Part A: Chem 2000;38(22): [3] Löfgren B, Seppälä JV. New functionalized olefin copolymers synthesised by metallocenes and novel organometallic catalysts. In: Scheirs J, Kaminsky W, editors. Metallocenebased polyolefins, vol. 2. Chichester, England: John Wiley & Sons Ltd; p [4] Hakala K, Löfgren B, Helaja T. Eur Polym J 1999;34(8): [5] Aaltonen P, Fink G, Löfgren B, Seppälä JV. Macromolecules 1996;29: [6] Jin J, Chen EY-X. Macromol Chem Phys 2002;203: [7] Yanjarappa MJ, Sivaram S. Prog Polym Sci 2002;27: [8] Hagihara H, Tsuchihara K, Takeuchi K, Murata M, Ozaki H, Shiono T. J Polym Sci, Part A: Polym Chem 2004;42: [9] Santos JM, Ribeiro MR, Portela MF, Pereira SG, Nunes TG, Deffieux A. Macromol Chem Phys 2001;202: [10] Imuta J-I, Kashiwa N, Toda Y. J Am Chem Soc 2002;124(7): [11] Paavola S, Uotila R, Löfgren B, Seppälä JV. React Funct Polym 2004;61(1): [12] Turunen J, Pakkanen TT, Löfgren B. J Mol Catal A: Chem 1997;123: [13] Hippi U, Korhonen M, Paavola S, Seppälä J. Macromol Mater Eng 2004;289: [14] Ristolainen N, Vainio U, Paavola S, Torkkeli M, Serimaa R, Seppälä J. J Polym Sci, Part B: Polym Phys 2005;43(14):