Hydrosilylation of terminal double bonds in polypropylene through reactive processing.
The hydrosilylation reaction, the addition of a silicon hydride to a multiple bond such as carbon-carbon, carbon-nitrogen, nitrogen-nitrogen, carbon-oxygen, and nitrogen-oxygen, was first reported in 1947 by Sommer et al. (1). Since then the reaction has been intensively studied and is currently the subject of 44 reviews. Hydrosilylation can be catalyzed radically, ionically, or by a metal complex such as platinum. rhodium, or palladium. One of the advantages of the platinum-catalyzed hydrosilylation is the tolerance and selectivity of the reaction. Oxiranes, acetals, esters, nitriles, amines, amides, nitro, ketone, carbamate, ether, isocyanate, phosphate, phosphonic dichloride, dialkoxy borane, sulfide, sulfone, or carborane groups can be present without reacting (2). Many of these functional groups are of considerable interest in polymer science, and it is therefore no coincidence that the application of hydrosilylation to functionalize polymers was soon investigated (2-5). However, these reactions have almost exclusively been investigated in solution. Although this offers certain advantages such as good temperature control, mild reaction conditions, and good homogenization, it is an energy-and labor-consuming process. The polymers have to be dissolved, precipitated after the reaction, and subsequently dried. The solvents have to be disposed of or they have to be recycled by distillation. Especially with polymers displaying a low solubility, this is a serious process limitation. Therefore, it is desirable to find a method to functionalize the polymer in the melt phase. This can be done through reactive extrusion. The advantage of performing a reaction in the extruder is that the polymer is functionalized and processed in the same step. An intermittent solidifying and remelting step is not necessary and, therefore, considerable savings of time, labor, energy, and equipment are accomplished.
Polypropylene is a major commodity polymer used in the industry. Its low price, good thermal and mechanical properties, chemical inertness, crystallinity, and hydrophobic character are desired in many applications. On the other hand, these features restrict its use in other highly profitable areas that are currently dominated by engineering plastics. Therefore, it is of interest to find a way to chemically alter polypropylene to increase features such as adhesion, chemical reactivity, or hydrophilicity. This would further open a path to the formation of interesting copolymers, the production of compatibilizers for inorganic fillers, or polymer blends.
Peroxide degradation of polypropylene, used to improve rheological properties (6, 7) leads to the formation of vinylidenes at the terminal site of the polymer chain. These vinylidenes may be selectively functionalized through hydrosilylation reactions. Two different pathways seem to be available to accomplish this task. The first is the radically induced hydrosilylation:
A peroxide initiator is used to form Si radicals, which attack the double bond and lead to the formation of a carbon radical. The reaction follows Farmer's rule (8), which states that the silyl radical will attack the carbon atom of the double bond bearing the most hydrogen atoms. The reason for this lies in the different stabilities of the carbon radicals. In the subsequent step, the newly formed carbon radical abstracts a hydrogen atom from a hydrosilane leading to the formation of the saturated organosilane and a new silyl radical, which can add to another double bond. This reaction mechanism is of special interest since peroxides are also used to perform the degradation of polypropylene and the subsequent formation of double bonds. However, it should be pointed out that radical reactions always bear the danger of side reactions and grafting along the polymer chain.
In 1957, Speier (9) discovered the activity of hexachloroplatinic acid as a hydrosilylation catalyst. It had such a high efficiency that it soon replaced all previous catalytic systems. The catalyst concentrations usually employed are in the order [10.sup.-5] mol platinum per mol of hydrosilane, although concentrations as low as [10.sup.-8] mol platinum per mol of hydrosilane have been reported (2). Today, even more reactive platinum catalysts are available. One of them is Karstedt's catalyst, platinum(0)-divinyltetramethyldisiloxane. As opposed to other highly reactive platinum olefin complexes, this catalyst can easily be derived from hexachloroplatinic acid (10) and it is commercially available.
The mechanism of the platinum catalyzed hydrosilylation with platinum olefin complexes such as Speier's or Karstedt's catalyst was long believed to be of a homogeneous type, described by the mechanism of Chalk and Harrods (8). In 1986, Lewis (11, 12) proposed the following mechanism:
In this mechanism, the colloid formation is the initial step in the hydrosilylation reaction, when catalyzed by a low valence platinum catalyst. This colloid is stabilized by oxygen, which is adsorbed on the catalyst surface, and the platinum-oxygen complex can coordinate a silane. The Si-H bond is not broken during this addition. One must imagine that the bonding is similar to that of the non-classical [H.sub.2] coordination (12). The newly formed complex has an electrophilic character. Oxygen functions as cocatalyst and activates the complex by withdrawing electrons. The olefin can now attack the transition complex nucleophilically, and the new organosilane is formed.
Oxygen stability is a very important feature of the catalyst since it makes the application of expensive inert gas atmospheres unnecessary. However, the requirement of oxygen leads to problems of a different kind in reactions in the melt phase. At normal processing temperatures, oxygen reacts with the PP chain leading to degradation and formation of undesired oxidation products. Therefore, the oxygen necessary to activate the catalyst has to be introduced chemically, perhaps in the form of a peroxide (13). It is desirable to employ a peroxide with a long half-life to ensure that the catalyst is activated during the entire reaction course and to avoid excessive degradation resulting from peroxide decay.
The polypropylene used in these experiments was an isotactic grade from Montell Canada (KF 6100, MFI = 3g/10 min). A 2,5-dimethyl-2,5-(t-butylperoxy)-hexane peroxide (Lupersol 101) from Elf Ato-Chem was used to degrade the polypropylene and to generate the required terminal unsaturation. Dime-thylphenylsilane from Fluka, hydride-terminated polydimethylsiloxane (PDMS, Mw = 400 g/mol), and platinum-divinyltetramethyldisiloxane-complex (Karstedt's catalyst) from United Chemical Technologies Inc. were used as the model silane compounds and catalyst respectively. Anhydrous t-butylhydroperoxide, 5-6M in decane from Aldrich was used as a cocatalyst to stabilize the platinum colloid. Tetramethyldisiloxane from Fluka and styrene from Eastman were used in the functionalization experiments. Methanol and acetone from VWR, and toluene, 99.5% for general use, from BDH were used in the cleaning procedures. All chemicals and solvents were used as received.
To generate the terminal double bonds, polypropylene was degraded using Lupersol 101 in concentrations of 0.5-5 wt%. In the rest of this manuscript, the amount of peroxide used for the production of the degraded PP will be shown in parentheses, i.e. degraded polypropylene (5 wt%).
For degradation reactions in the batch mixer, the polypropylene was mixed with the desired amount of peroxide and added to the pre-heated mixer at temperatures in the range of 190-200 [degrees]C. To minimize access of oxygen as well as material loss due to foaming, the mixer was closed with either a metal or a polytetrafluoroethylene stamp. Although the bulk of the peroxide was consumed within 2 min, the degradation product was kneaded for 10 additional minutes to ensure the complete removal of volatiles.
For degradation reactions in the single screw extruder, the polypropylene was premixed with the desired amount of peroxide and the masterbatch was added through the hopper. The screw speed was adjusted to give a material residence time of 5 min. The products from the degradation experiments were either pelletized or converted to a powder form by dissolution in toluene and subsequent precipitation with methanol.
The reaction was first attempted in a hot press using a mold having a disk-type cavity. For that purpose, powdered polypropylene was loaded into the cavity and an excess of PDMS/platinum catalyst solution was added to it. The reaction was allowed to proceed for periods of 15 min to 1 hour at temperatures between 160 and 200 [degrees]C. After this reaction time, the mold was submerged in cold water and the reaction sample was removed from the mold for purification.
Hydrosilylation in the batch mixer
Degraded polypropylene was added to the batch mixer at temperatures ranging from 160 to 180 [degrees]C. The desired amount of platinum catalyst was added to the PDMS and mixed vigorously. After the colloid was formed, as indicated by the characteristic yellow color of the solution, the PDMS solution was added to the mixer with a syringe. Subsequently, the mixer was closed with a polytetrafluoroethylene stamp, and a nitrogen atmosphere was applied in order to avoid degradation and oxidation. A maximum reaction time of 2 hours was possible, and samples were collected for analysis at different time intervals.
Hydrosilylation in the single screw extruder
The PDMS/catalyst solution was prepared as described in the previous section. A masterbatch of this solution and the degraded polypropylene was subsequently prepared and added to the hopper of the extruder. Reaction times varied between 2 and 10 min and the reaction was carried out at temperatures in the 170-230 [degrees]C range. Samples were taken from the air-cooled extrudate for further analysis.
FTIR analysis cannot distinguish between bound and unbounded siloxanes. Therefore, a thorough cleaning procedure had to be developed to ensure the complete removal of excess siloxanes prior to analysis. This was done by dissolution of the hydrosilylation product in boiling toluene and subsequent precipitation with methanol. Applying this cleaning procedure three times yielded a clean material that contained only chemically bonded silanes. If the precipitation was done with dry acetone, the unreacted Si-H group of the difunctional siloxane could be preserved.
Styrene Functionalization of Polypropylene
A 0.2 mol (26.86 g) amount of tetramethyldisiloxane and 0.02 mol (2.08 g) of styrene were added to a round-bottom flask together with 40 [[micro]liter] of Karstedt's catalyst. The reaction was allowed to proceed for 6 hours at temperatures of 70 [degrees]C to ensure complete conversion. The residual siloxane was removed by distillation and the reaction progress was followed by FTIR.
A 0.5 ml amount of the styryl functionalized tetramethyldisiloxane/catalyst solution was added to 0.6 g of degraded polypropylene (5 wt%) along with 10 [[micro]liter] of t-butyl-hydroperoxide. The mixture was heated in a hot press at 200 [degrees]C for 30 minutes. The product was cleaned following the previously outlined procedure. Subsequently, a film was pressed and analyzed by FTIR spectroscopy.
Fourier Transform Infrared Spectroscopy (FTIR)
Spectra were collected with a Nicolet 520 mid-range (500-4000 [cm.sup.-1]) FTIR. Because of its high sensitivity, FTIR proved to be the best analytical method to obtain reliable results during this research work. Sample material that had been cleaned following the procedure described above was pressed to a thin film in the hot press prior to analysis. Owing to the relatively low viscosity of the degraded polypropylene, it was also possible to take molten samples from the batch mixer or the extruder die and to spread a film between two spatula. Using this approach, films as thin as 10 [[micro]meter] thickness could be made.
1H NMR is very sensitive because of the high abundance of this isotope (99.985%) in nature (14). This was of great advantage for this project since the concentration of the chemically bound silane was very small. Spectra were taken using a Bruker AC-300 spectrometer. Because of the poor solubility of polypropylene, the samples had to be run at high temperatures up to 103 [degrees]C using deuterated toluene as solvent.
RESULTS AND DISCUSSION
Radically Induced Hydrosilylation
Since both the degradation of polypropylene and the hydrosilylation reaction are induced by peroxides, both reactions were attempted in parallel. Polypropylene, 5 wt% of peroxide and 10 wt% of polydimethylsiloxane were mixed to prepare a masterbatch. The mixture was loaded into the Haake mixer and the reaction was performed at 200 [degrees]C for 5 minutes. After cleaning as described above, the material was analyzed by FTIR, and the spectra is shown in Fig. 1. Incorporation of the siloxane is indicated by the Si-O-Si absorption at 1100-1030[cm.sup.-1]. Further proof is provided by the presence of the shoulders at 770 [cm.sup.-1] and 757[cm.sup.-1], which are attributed to the Si-Me groups. The shoulder at 888 [cm.sup.-1] also indicates that the terminal double bonds have not been consumed completely.
Performing the degradation and hydrosilylation in parallel has one large disadvantage, which is the high concentration of radicals. This always bears the possibility of side reactions such as grafting along the chain or undesired recombinations of two radical species. Degradation of the polypropylene with a high concentration of peroxide such as 5 wt% and subsequent hydrosilylation induced by catalytic amounts of peroxide avoids this problem. A hot press experiment was performed using 0.5 g of degraded polypropylene (5 wt%), 10 [[micro]liter] of Lupersol 101, and an excess of dimethylphenylsilane. The experiment was carried out at 160 [degrees]C for 15 minutes, and the spectrum of the product is shown in Fig. 2. It can be seen that the conversion of double bonds was high, as could be deduced by the extinction of the shoulder at 888 [cm.sup.-1], and that dimethylphenylsilane has been attached to the polymer as indicated by the bands at 1114 [cm.sup.-1] and at 730 [cm.sup.-1]. The high conversion of the terminal double bonds even at such a low peroxide concentration indicates that the silane adds preferentially to this position instead of the backbone. A detailed study of the radically induced hydrosilylation of terminal double bonds in polypropylene has been presented elsewhere (15).
Platinum Catalyzed Hydrosilylation
The catalytic reaction is shown schematically below:
In these experiments, the usual catalyst concentration was 2.64 x [10.sup.-4] g platinum per 1 gram of silane. The concentration of platinum in the polymer phase was of the order of [10.sup.-5] ppm. The actual catalytic species was a colloid formed by the reduction of the catalyst by the silane. The colloid formation could easily be observed since the solution adopts a characteristic yellow color as a result of this process. With polydimethylsiloxane and Karstedt's catalyst, the formation of the colloid was almost instantaneous. Usually the catalyst and the silane were combined prior to the introduction to the reaction vessel. The colloid was very stable and a silane/catalyst solution could be stored for several months in the refrigerator without a notable reduction of the catalytic activity.
At processing temperatures of polypropylene of 200 [degrees]C the polymer reacts with atmospheric oxygen. This leads to undesired degradation and to the formation of oxidation products such as methyl ketones, esters, aldehydes and [Gamma]-lactones. Also, reduction of the oxygen concentration in the melt inhibits the catalytic hydrosilylation, since oxygen acts as a cocatalyst. This was evident even after long reaction times, since the conversion of double bonds was very low and virtually no silane was incorporated. It was decided to introduce oxygen into the system in the form of a peroxide (13). The choice of the peroxide is crucial since high concentrations of radicals should be avoided. This is important not only to reduce the chance of radical side reactions but also to prove the catalytic activity of the platinum colloid. T-butyl hydroperoxide, having a half-life time of 60 min at 193 [degrees]C (16), appeared to be the most suitable candidate. It was applied in a 6 M solution in decane, while a t-butylhydroperoxide concentration of 0.1 wt% was mainly employed. To test the activating performance of the hydroperoxide, an experiment was performed in the batch mixer. A 200 g amount of polypropylene (0.5 wt%) was melted in the Haake mixer at 180 [degrees]C, A mixture of 5 g polydimethylsiloxane and 100 [[micro]liter] of Karstedt's catalyst as added and a sample was withdrawn after 30 min. Subsequently, 400 [[micro]liter] of t-butylhydroperoxide in acetone (dried over Ca[Cl.sub.2]) was added with a syringe and another sample was withdrawn after an additional 30 min. Both samples were cleaned following the usual procedure. Figure 3 shows the spectra of these two samples. It can be seen clearly that the reaction does not proceed without the addition of the t-butylhydroperoxide cocatalyst, as no siloxane absorption can be observed in the corresponding spectrum (dotted line). After the addition of the t-butylhydroperoxide, hydrosilylation of the polypropylene takes place, and the spectrum (solid line) indicates the double bond (888 [cm.sup.-1]) conversion and the silane incorporation (10301100 [cm.sup.-1]).
Additional evidence for the attachment of the siloxane onto the polypropylene may be obtained from 1H NMR measurements. Figure 4 shows the spectrum of a sample from a run carried out on the Haake mixer. The singlet at 0.4 ppm can be assigned to the protons of the methylene carbon, which is connected to the siloxane. The multiplet between 0.27 ppm and 0.2 ppm is due to the hydrogens of the siloxane methyl groups.
One advantage of reactive extrusion is that the polymer is functionalized and processed in one step. However, this requires that all reactants remaining in the product do not harm its quality. In the case of the platinum-catalyzed hydrosilylation reaction, the influence of the platinum colloid on the stability of the polypropylene has to be considered. To investigate this issue a thermogravimetric experiment was performed. A sample of degraded polypropylene (5 wt%), which had been catalytically hydrosilylated with PDMS in the single screw extruder, was heated in a helium atmosphere to 500 [degrees]C with a temperature ramp rate of 20 [degrees]C/min. As can be seen from the thermogram [ILLUSTRATION FOR FIGURE 5 OMITTED], the onset of the thermal degradation was observed at 389.7 [degrees]C. At processing temperatures of 200 [degrees]C the material was stable; no degradation was observed. For comparison, degraded polypropylene (5 wt%) was investigated under the same conditions. Here, the onset temperature for thermal degradation was 407.8 [degrees]C. This shows that the presence of the platinum catalyst has no detrimental influence on the thermal stability of polypropylene.
The catalytic reaction was also studied in a single screw extruder and a [2.sup.2] full factorial experimental design was performed to investigate the effects of screw speed and temperature. Experimental conditions are shown in Table 1. The degraded polypropylene, siloxane, platinum catalyst, and t-butylhydroperoxide cocatalyst were added via the extruder hopper as a masterbatch. Samples of the reaction product were taken at the die and spread to a film, which was subsequently analyzed by FTIR without further cleaning. The quantitative evaluation was done by measuring the peak height of the vinyl absorption at 888 [cm.sup.-1] and comparing it with the polypropylene backbone peak at 973 [cm.sup.-1]. This resulted in a relative concentration of double bonds. The results are summarized in Table 2 along with the factorial design conditions. According to the nomenclature of a full factorial design (17), "- -" symbolizes the experiment with both parameters at their lower level, "+ +" symbolizes the experiment with both parameters at their upper level, and "+ -" and " - +" symbolize the experiments with one parameter at the upper level and one parameter at the lower level. For evaluation of the experimental error, the experiment was repeated four times at the center point of the design plan. The calculated effects and F-test results are summarized in Table 3. From the experimental results, it could be concluded that the screw speed, and with it the residence time, had the largest influence on the double bond conversion and thus on the silane incorporation at the 95% confidence level (F-distribution value = 10.13). Somewhat surprising was the small temperature effect, since the statistical analysis proved it to be not significant. The two-factor interaction, however, was significant. Its negative value showed that the interaction leads to a reduction of double bonds, that is, an increase in the hydrosilylation product. An explanation of this result might be that the hydrosilylation reaction is relatively fast at the processing temperature, such that diffusion and mixing become the rate-determining step. Then, the results suggest that the interaction of screw speed and temperature increase the conversion of double bonds by improving the mixing between the PP and PDMS phases, possibly by lowering the viscosity of the PP domains.
Table 1. Conditions for [2.sup.2] Full Factorial Experimental Design. Variable Value screw speed 10-50 Rpm temperature 170-210 [degrees]C silane polydimethylsiloxane (4 g) degraded polypropylene (5 wt%) 40 g Pt-catalyst 40 [[micro]liter] t-butylhydroperoxide 80 [[micro]liter] die diameter 1 mm Table 2. Relative Concentration of Double Bonds in the [2.sup.2] Factorial Experimental Design. Experiment Screw Speed Temperature Relative (Rpm) ([degrees]c) Conc. center point 1 30 190 0.1184 center point 2 30 190 0.1137 center point 3 30 190 0.1173 center point 4 30 190 0.1098 - - 10 170 0.0988 + - 50 170 0.1370 - + 10 210 0.1060 + + 50 210 0.1187
Hydrosilylation of terminal double bonds in polypropylene with organofunctionalized silanes may be achieved via the same catalytic mechanism. An organo-functionalized silane may be prepared by using a difunctional tetramethyldisiloxane and an olefin with the desired functionality (4). The use of excess tetramethyldisiloxane guarantees the formation of the desired monoadduct. This technique was applied to produce a styrene-functionalized tetramethyldisiloxane as described earlier following the reaction scheme:
Table 3. Significance of Effects (Center Point Standard Deviation [[Sigma].sup.2] = 1.515 x [10.sup.-5]). Factor Effect F-Test Results mean ([[Beta].sub.0]) 0.11512 screw speed ([[Beta].sub.1]) 0.01272 [F.sub.1] = 42.01 temperature ([[Beta].sub.2]) -0.002775 [F.sub.2] = 2.04 interaction ([[Beta].sub.1, 2]) -0.006375 [F.sub.1,2] = 10.74
Styrene is used as a model substance since the phenylic ring can easily be detected by FTIR. Fig. 6 shows the spectra of samples after cleaning. It can be seen that the characteristic double bond absorption at 888 [cm.sup.-1] is nearly extinguished and that the Si-O-Si absorption at 1050 [cm.sup.-1] indicates the incorporation of tetramethyldisiloxane. The success of the reaction is further supported by the occurance of a peak at 699 [cm.sup.-1], which indicates the presence of the styrene phenyl ring in the polymer phase. It should be emphasized that no additional platinum catalyst had to be used. The platinum, which had been used in the production of the styryl-siloxane, could be activated sufficiently enough with t-butylhydroperoxide to catalyze the hydrosilylation of the polypropylene.
This experimental work proved the feasibility of the hydrosilylation of terminal double bonds in polypropylene in the melt phase. If platinum is used to catalyze the reaction, a peroxide cocatalyst is necessary to maintain the catalytic activity, and t-butylhydroperoxide has been shown to be a useful cocatalyst. The reaction time of the hydrosilylation reaction is short enough to be completed in a screw extruder. This makes possible the simultaneous extrusion and modification of polypropylene. Platinum residuals in the product did not seem to influence its thermal stability. The hydrosilylation reaction could also be induced by peroxide radicals. Degradation and hydrosilylation can be performed simultaneously by using high peroxide concentrations, while previously degraded polypropylene could be hydrosilylated with catalytic amounts of a peroxide. Finally, the preparation of terminally functionalized polypropylenes through hydrosilylation with organofunctionalized silanes was demonstrated in a case study involving the terminal attachment of a styrene molecule. Further details on this process may be found in a recent patent application (18).
1. L. H. Sommer, E. W. Pietrusza and F. C. Whitmore, J. Am. Chem. Soc., 69, 188 (1947).
2. I. Ojima, in The Chemistry of Organic Silicon Compounds, S. Patai, Z. Rappoport, eds., p. 1479, John Wiley and Sons Ltd., New York (1989).
3. X. Guo, R. Farwaha, and G. L. Rempel, Macromolecules, 23, 5048 (1990).
4. J. Hazziza-Lasker, G. He'lary, and G. Sauvet, Makromol. Chem., Macromol. Symp., 47, 383 (1991).
5. J. C. Saam and C. W. Macosko, Polym. Bull. 18, 463 (1995).
6. M. Dorn, Adv. Polym. Techn., 5(2), 87 (1995).
7. C. Tzoganakis, J. Vlachopoulos, and A. E. Hamielec, Polyrm. Eng. Sci., 28, 170 (1988).
8. B. Marciniec, J. Gulinski, W. Urbaniac, and Z. W. Kornetka, in Comprehensive Handbook On Hydrosilylation, B. Marciniec, ed., Pergamon Press, Oxford, England (1992).
9. J. L. Speier, Adv. Organometallic Chem., 17, 407 (1979).
10. B. D. Karstedt, U.S. Patent 3,715,334, General Electric Company (1973).
11. L. N. Lewis and N. Lewis, J. Am. Chem. Soc., 108, 7228 (1986).
12. L. N. Lewis, J. Am. Chem. Soc., 112, 5998 (1990).
13. H. M. Bank, U.S. Patent 5,359,113, Dow Corning Corporation (1994).
14. R. B. Taylor, B. Parbhoo, and D. M. Fillmore, in The Analytical Chemistry of Silicones, A. L. Smith, ed., 12, 347, John Wiley and Sons, New York (1991).
15. G. N. Shearer and C. Tzoganakis, J. Appl. Polym. Sci., 65, 439 (1997).
16. Elf Atochem Organic Peroxide Half-Life Data Brochure (1992).
17. G. E. P. Box, W. G. Hunter, and S. J. Hunter, Statistics For Experimenters, John Wiley & Sons Inc., New York (1995).
18. C. Tzoganakis and H. Malz, Hydrosilylation of Polypropylene, International Patent Application PCT/CA97/00412, July 12, 1997, pursuant to U.S. Provisional 60/019,678.
|Printer friendly Cite/link Email Feedback|
|Author:||Malz, Hauke; Tzoganakis, Costas|
|Publication:||Polymer Engineering and Science|
|Date:||Dec 1, 1998|
|Previous Article:||Rheological and mechanical properties of polyethylene blends.|
|Next Article:||Pressureless sintering of polycarbonate powder compacted at ambient temperature.|