Silane Grafting and Moisture Crosslinking of Polypropylene.
Peroxide initiated vinylsilane grafting of polypropylene in an intensive mixer, and the subsequent water crosslinking process were studied. Different concentrations of vinyl trimethoxysilane and dicumyl peroxide were used. The materials obtained after mixing in the rheocord were hot pressed at 190[degrees]C. The melt viscosity of the obtained sheets, the melting enthalpy and melting temperature (DSC, differential scanning calorimetry), the mechanical properties and the thermal decomposition behavior (TG, thermogravimetric analysis) were studied. No evidence of grafting during the rheocord processing was observed. Nevertheless, for the hot pressed sheets with concentrations higher than 4 phr of vinyl silane an important increase in the melt viscosity was observed. This increase agrees with the change observed in the mechanical properties, which show a maximum for the water crosslinked samples containing 4 phr of vinyl silane. The modulus increases by 39% at 90[degrees]C and 33% at 130[degrees]C, while the ten sile strength rises by [sim]22% at both temperatures. The silane grafted water crosslinked samples show a more stable thermal behavior than both the silane grafted samples and the unmodified polypropylene.
Over the last few years a growing interest has developed regarding the process of crosslinking conventional polymers. Among the different possibilities to crosslink polymers, silane grafting of polyolefins has received the most attention because of the easy processing, low cost and capital investment and especially the important development of mechanical properties achieved in these materials [1-4].
Moisture curable polyethylene through silane grafting has been used commercially since the 1970s , and since then a great number of papers and patents on crosslinking polyethylene, ethylene propylene rubbers, polyvinyl chloride, etc., have been published, dealing with very different aspects such as formulation, processing, reaction kinetics, thermal and mechanical properties, etc. [1-8]. There are some recent patents [9-11] on the vinylsilane crosslinking of polypropylene, but surprisingly, no research publications have been found on this subject, apart from this one .
The usual procedure for crosslinking polyolefins through silane grafting is the following: The vinyl silane (generally vinyl trimethoxysilane for PE crosslinking)  is introduced into the polyolefin by compounding in an extruder at high temperatures in the presence of a peroxide. The peroxide acts as a source of free radicals, which abstracts a proton from the polymer backbone, giving reactive sites for the subsequent grafting of the silane (Fig. 1). Hydrolysis of the methoxy groups by water yields silanol groups, which are then free to undergo condensation in presence of a catalyst, to give chemical crosslinks.
In this work we have employed the above procedure to crosslink polypropylene. Vinyltrimethoxysilane and epoxytrimethyoxysilane, frequently used to crosslink polyolefins, have been employed in different concentrations. The thermal and mechanical properties of the silane grafted and silane grafted water crosslinked polypropylenes have been studied and compared with the unmodified material. The processing and melt viscosity of the compounds obtained are also compared.
A commercial isotactic polypropylene (PP) Isplen from Repsol was used, with a MFI of 0.96 (gr/10 min). The silanes used were vinyltrimethoxysilane (VS) and epoxytrimethoxysilane (ES), A-172 and A-187 respectively, supplied by Quimidroga. Dicumyl peroxide (DCMP) from Merck and dibutyltindilaurate (DBTDL) from Fluka were used without further purification.
2.2 Sample Preparation
All samples were compounded in a Haake Rheocord 9000 unit coupled to a Rheomix 600 mixer with two roller rotors. In all cases 50 g of polypropylene was introduced in the mixing chamber, and after it melted (approximately 5 minutes after introduction, as can be seen in Fig. 2), the corresponding quantities of the other components were added. The conditions used in the Rheocord were 175[degrees]C, 40 rpm, and the time of mixing was 20 min in all cases. Control samples were prepared for all series.
The samples obtained were hot pressed at 190[degrees]C and 30 bar for 3 min. and immediately quenched under the same pressure for 1 min, to obtain thin sheets, approximately 1mm thick. To complete the crosslinking process, half the samples were immersed in water at 90[degrees]C for 5 hours.
Sample designations and compositions are given in Table 1.
The sample designation was as follows: PP, number indicating the phr of silane (2, 4, 6, 8 or 12), two letters for the type of silane (VS for the vinyltrimethoxysilane and ES for the epoxitrimethoxysilane), and one additional letter "a," or "b," for the quantity of initiator and catalyst used (0.4 of DCMP and 0.6 of DBTDL for series "a" and 0.2 and 0.1 respectively for series b). Finally the W, when it appears, indicates that the sample was crosslinked in boiling water after preparation.
2.3 Gel Content
To determine the gel content, discs of approximately 100 mg were cut from the sheets of each material. The discs were extracted by refluxing in decalin. For all samples studied there was no detectable quantitative gel formation. All the samples can be almost completely dissolved in decalin at elevated temperatures, and in all cases the majority of the sample precipitated after cooling (approximately 90% of the initial weight).
2.4 Melt Flow Index determination
The melt flow index (MFI) was determined in a Metrotec, extrusion plastometer according to the American National Standard ASTM D 1238-73 standard. The MFI is a measure of the zero shear viscosity, and consequently is related to the molecular weight and to the presence of different additives in the formulations.
The hot pressed sheets were crushed and introduced into the cylinder of the plastometer, which was heated at 170[degrees]C. After 10 min of preheating, a piston of 5 Kg was placed into the cylinder and the extruded samples were collected and weighed. Results are given as grams of sample extruded/10 minutes.
2.5 Thermal Properties
Thermogravimetric analysis (TG) was carried out with a Perkin-Elmer TGA7 thermal analyzer. The temperature was varied between 50[degrees]C and 600[degrees]C at 10[degrees]C/min, under a [N.sub.2] purge flow of 30 mL/min. Samples of approximately 7 mg were cut from the sheets obtained.
Differential scanning calorimetry was carried out using a Metier TA 3000-DSC20 system with TC10A microprocessor. Three consecutive scans were obtained to minimize the influence of possible residual stress in the material due to any specific thermal history. Samples of approximately 6 mg were cut from the sheets. A scanning rate of 10[degrees]C/min was used for both heating and cooling cycles, under a [N.sub.2] purge of 30 mL/min.
2.6 Mechanical Properties
The mechanical properties were studied by the analysis of the stress-strain curves obtained in an Instron 4031 universal testing machine. Measurements were carried out using a 0.1 KN load cell, with a testing speed of 50 mm/min, and at 20[degrees]C, 90[degrees]C and 130[degrees]C. Halterio type samples, according to the ASTM D 638-96, were cut from grafted and water cross-linked polymer sheets, obtaining specimens of 4 [mm.sup.2] cross-sectional area.
Figures 2 and 3 show the evolution of the torque with time during processing of the samples in the Haake Rheocord. The maximum torque is only dependent on the quantity of PP used, and as a consequence is similar in all cases. After the maximum. a plateau is observed that is also similar for all the samples used. A slight decrease in the torque can be observed at t = 5 min resulting from the addition of the other components in the formulations. Obviously this decrease is not observed in the case of the PP sample where no additional components were added, and where the torque reaches a maximun value of 3.71 N.m.
The samples containing only DCMP and DBTDL (PPa and PPb) are compared with the unmodified PP in Fig. 2. In both cases, a smooth decrease in the torque is initially observed, until it reaches the value of 1.04 and 0.50 N.m for PPb and PPa, respectively. This important decrease in the torque can only be explained by premature chain scission of the polypropylene resulting from the action of the DCMP (2). In fact, we have found in a separate work that, PVC processed with DCMP in similar conditions does not show a similar decrease (3). Consequently, by increasing the quantity of DCMP used, the observed torque decreases.
The decrease in the torque in the case of the samples containing DCMP, DBTDL and different quantities of VS is faster than for the sample containing only the initiator and the catalyst, but is not so apparent. Figure 3 shows the results for series b. An increase in the quantity of VS produces a decrease in the value of the torque observed. This value, related to the energy consumption during processing, falls to levels as low as 0.72, 0.45 and 0.31 N.m for samples with 4, 8 and 12 phr of VS, respectively. In the case of samples with ES, the decrease in the viscosity is not so important. In the case of the sample containing 8 phr of ES the torque reflects a value very similar to that for the PPb sample (Fig. 3). Although not shown here, similar results have been obtained for series a.
The decrease observed in the torque and thus in the viscosity of the mixture is indicative that up to this point no crosslinking processes have taken place, and can be explained either by a lubricating effect that small molecules such VS or ES may have on the polypropylene melt, or by the internal plasticization that grafting processes could promote in the polypropylene melt. The different behavior observed for VS and ES could be related to two facts: to the different polarity, and thus the lubricant capacity of both silanes, and to the different extent in the chemical grafting by the VS and ES.
3.2 Fusion Behavior
3.2.1 Melting Temperature and Melting Enthalpy
The melting peak temperatures and melting endotherms as a function of the concentration of the croslinking agent are shown in Table 2. The melting temperature and the melting enthalpy hardly vary with the heating scan. This suggests that the silane grafted and silane grafted water crosslinked PP could be reprocessed. It can be observed that the treatment suffered by the samples does not affect the thermal properties of polypropylene, the melting temperature and enthalpy remaining almost constant. Similar results have been found by other authors for silane crosslinking of polyethylene (2, 5). The almost constant melting temperature and enthalpy may be explained by high fractions of slightly crosslinked crystalline material. Only the amorphous region is crosslinked, leaving the melting properties of the crystalline material basically unchanged.
3.2.2 Melt Viscosity
During sample processing in the Rheocord, no evidence of either crosslinking or grafting was found. The MFI of these samples was measured from the sheets obtained after molding in the hot press.
For the unmodified PP sample a typical value of 0.96 g/10min was obtained. An important decrease in the viscosity of the samples can be observed when a small quantity of DCMP is added, for the samples PPb and PPa, with MFI values of 8 and 48, respectively. The same behavior was observed in the torque curves (Fig. 2). Chain scission caused by the peroxide may be the reason for the important decrease in the viscosity of these samples.
Table 3 shows the MFI results for series b. Although not shown here, a similar trend was obtained for series a. For the series b, the sample containing 2 phr of VS (PP2VSb) shows an increase in the MFI with respect to the PPb sample. This high value, which means a very low viscosity in the melt for this sample, could be due to the lubricant effect promoted by the addition of a small quantity of VS. During the mixing of polypropylene with VS (Fig. 3), a lubricant effect was noticed, which decreases the torque with increasing VS content. The behavior observed for the PP2VSb sample was similar to that previously observed.
When these samples are reprocessed, two factors affect the viscosity of the melt; the decrease in the viscosity due to the lubricanting effect of the VS molecules, and the strong increase in the viscosity as consequence of crosslinking, which may occur during the hot pressing of the samples. For the PP2VSb sample, the lubricant effect dominates, and the quantity of silane may not be sufficient to promote crosslinking. From the 4 phr sample to the most concentrated, partial crosslinking has taken place and therefore, these samples present a higher viscosity and lower MFI. When the VS concentration is increased, or when the conditions for crosslinking are favored (immersion in hot water or an increase in the DCMP concentration), the MFI obtained decreases.
3.3 Mechanical Properties
Mechanical testing of silane grafted and silane grafted-water crosslinked samples was carried out to determine the mechanical properties. Tables 4, 5 and 6 show the results obtained at 20[degrees]C, 90[degrees]C and 130[degrees]C, under the conditions described in the Experimental section. Three samples were tested in each case. The numbers between brackets in Tables 4 to 6 correspond to the standard deviations.
At 20[degrees]C (Table 4) it is clear that the peroxide induces a very drastic change in the molecular structure of the resin, which hinders the normal flow behavior of the polypropylene. The addition of VS affects the elongation at break very slightly, which is always maintained at a level lower than 98% of that of the original PP.
On the other hand, the modulus and the ultimate and maximum tensile strength (Table 4) are slightly increased with respect to the PP sample with the addition of the VS molecules. The maximum in properties is reached when 4 phr of VS is added and when the samples are treated in hot water. In this case, Young's modulus and maximum tensile strength are increased by 16.5 and 11.5 % with respect to the unmodified PP sample.
At higher temperatures all samples show fluency. The maximum displacement allowed by the apparatus when operating with the hot camera is 30 cm, but all the specimens tested showed an elongation higher than 1500%. Therefore, elongation at break and ultimate tensile strength could not be measured at 90[degrees]C or 130[degrees]C.
Nevertheless, as expected, the development of modulus and tensile strength at 90[degrees]C and 130[degrees]C is higher with respect to the PP sample (Tables 5 and 6). The same trend as that observed above (a small change in properties for samples not treated in water, a higher development of properties in water treated samples, and a maximum at 4 or 6 phr VS), is observed at 90[degrees]C and 130[degrees]C. The improvement of the modulus is 39% at 90[degrees]C and 33% at 130[degrees]C, while the improvement in the tensile strength is [sim]22% in both cases for series b. For series a, mechanical properties were only determined at 90[degrees]C, obtaining an improvement in the modulus and the tensile strength of 39% and 29% respectively, compared to the values for PP sample at this temperature. Once again, the highest results are obtained with increasing quantity of DCMP.
3.4 Decomposition Behavior
Figure 4 shows the thermograms of the PP and PPb samples. The sample containing the DCMP starts its decomposition at temperatures somewhat lower than the sample of PP. At 350[degrees]C, for example, the sample PPb has lost 3.6% of its initial weight, while the PP sample has only lost 0.5% (see Table 7). As mentioned above, the DCMP present in the PPb sample could generate some free radicals and premature chain scission in the sample and, consequently, a lower decomposition temperature and lower torque during processing. Figure 4 also shows the derivative weight loss curves (DTG) of the samples, where for practical purposes, only one symmetrical peak can be observed, indicating a single decomposition process. Similar DTG curves are observed for the other samples tested.
Figure 5 shows the comparison of the initial step in the decomposition of the PP sample with the water treated and untreated samples containing 6 and 12 phr of VS in the series b. In samples containing VS there is a small weight loss at temperatures as low as 120[degrees]C-l70[degrees]C, probably due to the volatilization of the unreacted VS remaining in the sample. Samples treated in water show lower weight loss than those untreated (Table 7), showing the greater extent of the silane reaction.
Figure 6 shows the complete decomposition process for the samples containing VS. The presence of a residue after decomposition can be observed in the case of the water untreated samples, which is not present in the corresponding water treated samples and which increases with the VS content. Moreover, in the water treated samples, a more stable behavior is observed than in the case of the polypropylene sample. It can be concluded that silane crosslinking slightly improves the thermal stability of PP. Higher VS content produces more stable samples, in agreement with the idea of a more stable three-dimensional network.
No crosslinking process is observed during compounding of the PP samples in the rheocord. A decrease in viscosity with the peroxide concentration and a lubricant effect of the vinylsilane were observed.
The melt viscosity of samples after hot pressing shows an increase in viscosity for the samples containing only DCMP, and for samples with low concentrations of vinyl silane. These results agree with the results obtained in the Rheocord. Nevertheless, for the sample with 4 phr vinyl silane and for the more concentrated samples, a large increase in the viscosity is observed. Grafting, and perhaps partial crosslinking, can occur during the hot pressing of the samples at 190[degrees]C, with concentrations of vinyl silane higher than 4 phr. Nevertherless, no gel content was observed in any of the samples. The network formed may be of a weak nature.
The mechanical properties of the samples at 20[degrees]C, 90[degrees]C and 130[degrees]C show maxima at 4 or 6 phr. The development of the mechanical properties is higher at the greater temperatures. An increase of [sim]30% can be observed in the modulus and in the tensile strength at 90[degrees]C.
Silane grafting and silane grafting water crosslinking slightly improve the thermal stability.
We are grateful to the Comision Interministerial de Ciencia y Tecnologia (CICYT) for financial support, MAT 96-0615 and MAT 99-1179.
(1.) Permanent address: Departamento de Ingenieria Quimica, Universidad de Alicante, Alicante 03080, Spain.
(*.) Corresponding author.
(1.) M. Hidalgo, L. Gonzalez, and C. Mijangos, J. Appl. Polym. Sci., 61, 1251 (1996).
(2.) M. Celina and G. A. George, Polym. Deg. Stab., 48, 297 (1995).
(3.) M. Hidalgo, M. I. Beltran, H. Reinecke, and C. Mijangos, J. Appl. Polym. Sci., 70, 865 (1998).
(4.) A. K. Sen. B. Mukherjee, A. S. Bhattacharyys, P. P. De, and A. K. Bhowmick, J. Appl. Polym. Sci., 44, 1153 (1992).
(5.) Y. T. Shieh and K. I. Shiao, J. Appl. Polym. Sci., 70, 1075 (1998).
(6.) M. M. P. Lee, C. H. Hsieh, K. M. Yeh, C. I. Chang, and P. L. C. Hao, J. Vinyl Tech., 2 (2), 119 (1980).
(7.) W. K. Wong and D. C. Varrall, Polymer, 35 (25), 5447 (1994).
(8.) M. I. Beltran, J. C. Gracia, A. Marcilla, M. Hidalgo, and C. Mijangos, Polym. Deg. Stab., 65, 65 (1999).
(9.) M. Nakamura, M. Sasakura, and S. Yamamoto, Patent JP 08323855 A2 10 Dec 1996.
(10.) M. Koji, Patent JP 09095590 A2 8 Apr 1997.
(11.) V. Boulgakov, E. Pikous, and G. Djavakhchhvili, Eur. Pat. Appl., EP 776930 Al 4 Jun 1997.
(12.) J. Toynbee, Polymer, 35 (2), 438 (1994).
|Printer friendly Cite/link Email Feedback|
|Author:||BELTRAN, M.; MIJANGOS, C.|
|Publication:||Polymer Engineering and Science|
|Date:||Jul 1, 2000|
|Previous Article:||Mechanical Properties and Interfacial Interaction of [CaCO.sub.3] Filled HDPE Compatibilized With HDPE Functionalized by Ultraviolet Irradiation [*].|
|Next Article:||Copolymerization of Carbon Dioxide and Propylene Oxide Using Various Zinc Glutarate Derivatives as Catalysts.|