Polyvinyl butyral chemically modified with a silane agent in the molten state.
Polyvinyl butyral (PVB) is an amorphous random terpolymer of vinyl butyral, vinyl alcohol, and a small fraction of vinyl acetate, and its structure is shown in Fig. 1 [1, 2]. It is obtained by the condensation of butyraldehyde with polyvinyl alcohol (PVA), which is a derivative of poly vinyl acetate (PVAc), in the presence of an acid catalyst [3-5]. The vinyl butyral unit is hydrophobic and promotes good processability, toughness, elasticity, and compatibility with many polymers and plasticizers. Hydrophilic vinyl alcohol and vinyl acetate units are responsible for high adhesion of PVB to inorganic materials, such as glass. Commercial PVB contains approximately 17%-22% vinyl alcohol, l%-3% vinyl acetate, and 75%-82% vinyl butyral (w/w) .
PVB is mainly used in laminated safety glass in the automotive, aerospace and architectural industries. The PVB used in windshields is highly plasticized and different plasticizers, such as alkyl phthalate, dibutyl sebacate, and di-2-ethylhexanoate, of triethylene glycol may be present to different extents . These safety glasses consist of a sandwich of PVB film between two glass sheets, and its production generates a large volume of PVB waste , The recycling of PVB is hampered by the loss of plasticizers and/or degradation during the recycling process which can deteriorate the polymer properties [1, 2]. Ambrosio et al.  obtained composites with recycled PVB and leather fibers through extrusion, demonstrating that it is possible to recycle PVB. In this study, DMA analysis adopting frequency of 1 Hz showed a slight increase in the glass transition temperature ([T.sub.g]) of the extruded PVB ([T.sub.g] = 30[degrees]C) when compared with the PVB flake ([T.sub.g] = 27[degrees]C). These results indicated that the plasticizer remained in the composite after reprocessing.
Despite of its elasticity and toughness, the applications of PVB is limited because of its poor solvent resistance to organic solvents. Hydroxyls groups in vinyl alcohol fraction are mainly responsible for this limitation, and an improvement in the solvent resistance of PVB can enlarge its applications.
Several techniques have been used for chemical modification of polymers, fillers, and composites in order to improve specific properties [8-11]. Modification of natural fibers is adopted to improve the interfacial adhesion between the fibers and the polymer matrix to hence the properties of polymer composites . Polyolefins are chemically modified to increase some of its properties. Polyethylene can be crosslinked with peroxides, high energy electrons, gamma ray, or silane agent to increase its operating temperature and solvent resistance [9-13], while polypropylene (PP) can be modified with maleic anhydride (MA) to improve the interface between fillers and matrix . Chemical modification of neat polymers and composites mostly occurs in the molten state, in single or twin screw extruders or internal mixers [11, 14-18]. Internal mixers, like torque rheometers, are useful tools to perform chemical reactions in polymers during the mixing process. There are several examples of chemical modifications performed in a torque rheometer. Some examples involving elastomers and thermoplastics are ethylene-propylenediene terpolymer grafted with vinyloxyaminosilane (EPDM-g-VOS) , thermoplastic polyurethane (TPU) modified with 3amino-propyl trimethoxysilane (3-APTMS) , polyamide-6 (PA6) modified by chain extension with 2,2'-bis(2-oxazoline)  and PP, low-density polyethylene (LDPE), and high-density polyethylene (HDPE) modified with vinyl silane to obtain cross-linked polymers [17-20]. Torque rheometry is useful to follow the extent of crosslinking by monitoring the torque.
The traditional method of chemically modify polyolefin with vinyl silanes involves peroxides to generate free radicals on the polymer chain. This is carried out in two stages. First the peroxide free radical abstracts a hydrogen atom from the chain. Grafting occurs by addition of the vinyl group on to the chain radical with subsequent termination of the vinyl radical by abstraction of a further hydrogen atom. Afterward, crosslinking occurs by hydrolysis of the methoxy groups in the presence of water . In the case of PVB, its solvent resistance can be improved through crosslink reactions with silane agents. Metrokea and Henleyb modified PVB coating using a combination of bis(trimethoxysilylethyl)benzene and dibutyltin dilaurate as the hydrolysis catalyst . Chemical modification occurs through reactions between the hydroxyl group from PVB with silane agents. There are two main reactions that are normally involved with alkoxysilanos, hydrolysis and condensation [22, 23]. Both reactions are shown in Fig. 2.
The hydrolysis of trialkoxysilanes to silanetriols occurs in several steps, and each succeeding step is faster than the first due to the relief of the steric hindrance at the silicon layer (Fig. 2a). The condensation reaction consists of bridging two molecules by crosslinking via siloxane bond formation. The siloxane bonds are created by the condensation of silanols from different molecules or by the reaction of silanol with another molecule of alkoxysilane, as shown in Fig. 2b . Siloxane bonds can also be formed by reactions of silanols and alkoxysilane from the same molecule.
The aim of this study was to improve the solvent resistance of PVB by crosslinking reactions to enlarge the application of neat PVB and as a flexible polymeric matrix for composites. The reactive mixture of PVB with the silano (vinyltrimethoxysilane [VTMS]) agent was performed in the molten state using an internal mixer. PVB was mixed with VTMS in the melt using an internal mixer and then compression molded. Another goal of the study was to determine the optimal chemical modification process parameters to obtain a PVB resistant to organic solvents that maintains its elasticity and toughness properties. Finally, a deeper investigation to better understand the crosslinking reactions between PVB and VTMS was performed.
The PVB used in this study (density of 1.20 g/[cm.sup.3]) was a recycled laminate glass material from the automotive industry. The VTMS is known as Dynasylan and was supplied by Evonic. The VTMS is a bifunctional organosilane with a reactive vinyl group and a hydrolyzable trimethoxysilyl group .
PVB was previously dried in an oven with circulating air at 70[degrees]C for 14 h to remove moisture and avoid hydrolyses during processing. The melt mixing of PVB and VTMS was performed in a internal mixer (Haake QC Polylab). The PVB modified with VTMS was compression molded by a hydraulic press (Marconi model MA-098) with a load of 10 tons for 10 min. Then, the sheets were stored in an atmosphere saturated with moisture at 60[degrees]C to favor the crosslinking process (hydrolysis as shown in Fig. 2a). The parameters adopted in this study can be seen in Table 1.
Characterization and Testing
The efficiency of the chemical modification with VTMS was evaluated by Soxhlet extraction. The extraction was performed based on the ASTM D2765 standard. Ethanol was used as the solvent, and the extraction temperature was 60[degrees]C for 8 h. Ethanol was selected as the PVB solvent because it is a typical substance found in many industrial environments and applications. To identify and quantify the additives in PVB, further Soxlet extraction was performed on PVB using hexane as the solvent at a temperature of 60[degrees]C for 8 h. In this extraction, only the additives of PVB were solubilized. The hexane was evaporated, and the extract was quantified and identified by Fourier transform infrared spectroscopy (FTIR) using a Nicolet spectrometer (model 4700). The FTIR spectrum was obtained from KBr flakes at room temperature and at wave numbers from 400 [cm.sup.-1] to 4,000 [cm.sup.-1].
Tensile tests of the modified PVB were performed in a universal testing machine (INSTRON 5569R1789) with a displacement speed of 500 mm/min. The ISO 527 type 1BA specimens were produced from the compression molded sheets to verify that the chemical modifications altered the mechanical properties of the PVB. A hardness test was performed according to the ASTM D2240 standard using a Bareinss machine (model Digitest). Dynamic mechanical analysis to verify the morphology of the crosslinked PVB clusters was performed in a dynamic mechanical analyzer (DMA) (TA Instruments, model DMA Q800). Single cantilever claws were used in flexural mode at a temperature of 40[degrees]C to scan the frequency range of 1 to 50 Hz.
The decomposition study of the modified PVB was performed on an apparatus from TA Instruments, model Q500, operating at a heating rate of 5 K/min, with temperatures ranging from 30[degrees]C to 600[degrees]C in a nitrogen atmosphere. The sample weight ranged from 13 to 16 mg.
RESULTS AND DISCUSSION
The silane agent adopted in this study has a reactive vinyl group and a hydrolyzed trialkoxysilane. Some studies have shown that the alkoxysilanes can directly react, without prehydrolysis, with hydroxyls groups from another substrates, for example, --Si--OH groups present in silica [26, 27]. Figure 3a represents the grafting reactions of VTMS and hydroxyls from PVB structure without previous hydrolysis. Grafting of the alkoxysilane in the PVB molecule can occur when the vinyl group of the silane reacts with any free radical formed during mixing. However, the most probable grafting mechanism is the reaction between the hydroxyl from PVB (--OH) and the alkoxysilane due the difference of electronegativity by transestrification (Fig. 3a). Transesterification is a reaction in which an alcohol displaces an alkoxide group to produce another alcohol molecule (Fig. 3a) . Then, the most probable most reaction is between the grafted alkoxysilane with a second or third hydroxyl group of the polymer, as is represented by Fig. 3b.
Melt Mixture of PVB-VTMS
Torque development during mixing is shown in Fig. 4. The first torque peak is due to the entry of the solidified polymer into the internal mixer. When PVB softens, the viscosity decreases and causes a reduction in torque. Conditions 1 and 4 (PVB without VTMS) only had one peak. Condition 1 was carried out at a higher temperature than condition 4; therefore, condition 1 had smaller torque values due to the polymer viscosity decrease with temperature. Observing curves 1 and 4, it is evident that PVB was completely soft after 2 minutes and that the torque was stable after 6-7 min of mixing. When VTMS was added to the PVB, a new torque peak was observed due to the viscosity increase of the mixture, likely caused by the dynamic crosslinking reactions , Dynamic crosslinking is a chemical crosslinking reaction that occurs in the internal mixer while polymer is under shear.
Comparing torque curves 2 and 3, processed with the same conditions, it is observed that the higher VTMS content of condition 3 increased the crosslinking rate, once the torque peak on curve 3 is formed 30 s before the torque peak on curve 2. However, the torque level of both peak are the same, indicating the same viscosity of the bulk for both VTMS contend. Condition 6, processed with higher rotation speed and with lower VTMS content and mixing temperature than condition 3, also presented a peak with same torque level, but the peak is broader indicating lower crosslinking rate. The same maximum torque level obtained in conditions 2, 3, and 6 indicates that the shear imposed by mixing limits the bulk viscosity (crosslinking degree). Temperature and VTMS content just have influence on crosslinking rate. Condition 3 has increase of torque after the crosslinking torque peak probably due to excess of VTMS. After the maximum torque level is achieved on condition 3, the shear stress due mixing break the polymeric chains and/or crosslinked domains decreasing the bulk viscosity. But, as condition 3 has excess of VTMS, crosslinking reactions can continue until the maximum torque level is achieved again. Conditions 4, 5, 7, 8, and 9 have no torque peak due to crosslinking reaction because the mixing time was too short.
Solubility Evaluation of the Modified PVB
The gel content shown in Fig. 5 was obtained by Soxhlet extraction and indicates the presence of chemical modification. The neat PVB (conditions 1 and 4) proved to be completely soluble in ethanol and the gel content was practically zero. Conditions 2 and 3, which presented a torque peak due to crosslinking (Fig. 4), have the highest gel content. After mixing, both conditions were highly crosslinked and compression molded failed.
Reducing the VTMS concentration and mixing temperature decreased the crosslinking rate and compression molding was easier. At 2 min (condition 5), there was no significant cross linking reaction (no torque peak), leading to a low gel content and low solvent resistance. With 7 min (condition 6), a torque peak at 4 min was noted, indicating the occurrence of crosslinking reactions. The gel content of condition 6 was approximately ten times higher than that of condition 5.
Conditions 7 and 8 represent the influence of the compression molding temperature on the gel content. Both conditions had very similar torque behavior, with no torque peak; therefore, no significant crosslinking reactions occurred during mixing. However, condition 8 had a gel content 35 times higher than condition 7. This suggests that the crosslinking reactions occurred during compression molding and that higher molding temperatures facilitated these reactions. These are static crosslinking reactions once polymer is not under shear.
In fact, comparing conditions 8 (VTMS = 0.92% [w/w]/ [T.sub.molding] = 180[degrees]C) and 9 (VTMS = 1.22% [w/w]/ [T.sub.molding] = 150[degrees]C), it is clear that the molding temperature had a greater influence on gel content than the VTMS concentration.
The modified PVB was characterized by tensile and hardness tests to evaluate if the crosslinking reactions compromised the elasticity and toughness of the PVB. The solvent resistance improvement by silane modification did not seem to significantly affect the mechanical properties of the PVB, as shown in Table 2.
Analysis of neat PVB (conditions 1 and 4) showed that the processing temperature did not alter the tensile strength and elongation at break, maintained its rubber-like properties. However, hardness increased 14.5% when the processing temperature was increased. These results should be analyzed with caution, however, since the specimens showed irregularities due to crosslinking. Conditions 2 and 3 were excessively crosslinked before compression molding and it was not possible to produce a tensile test specimen.
Influence of the Time and Process Steps on the Chemical Modification of the PVB
Condition 8 was replicated with different mixing times to verify which time is the ideal for reactive mixture of PVB and silane. Five mixing times were set based on torque and temperature monitoring curves obtained in the previously mixing (Fig. 6). Furthermore, it is necessary to prove if the crosslinking reactions occur during reactive mixing, compression molding or during storage in a humid atmosphere. For that, samples of PVB modified with different mixing times were tested by Soxhlet extraction after reactive mixing, after compression molding and after storage in a humid atmosphere.
Three reactions are likely to occur on mixing; grafting when VTMS first reacts with PVB chains; dynamic crosslinking which occurs between the grafted VTMS groups and between grafted VTMS and --OH from PVB under shear (mixture of polymeric bulk); breakage of chains due to temperature and shear. The mixer temperature was set at 150[degrees]C but due to viscous heating it rose to 180[degrees]C and remained stable (Fig. 6). Temperature positively contributed to the three reactions occurring during the mixing process. Grafting was the main reaction in the early stages but depended on the concentrations of the reagents and the temperature. The reactions involving the grafted chains are the dynamic crosslinking reactions, which increase the viscosity of the polymer due to the bonds between chains. Shear stress is proportional to the viscosity of the polymer; therefore, the crosslinking reactions cause increase of shear stress on the polymer bulk, resulting in a more intense chain breakage through mechanical energy absorption.
In Fig. 6, T1 represents the time to initiate grafting and crosslink formation. T1 also represents the gel time, where gel fraction appears . As the crosslinking reactions become more intense, the viscosity and torque increase and more chains or crosslinked clusters, which are highly crosslinked regions in the bulk, are broken or dispersed. At time T3, the maximum in rate of crosslinking is reached. At this point, the PVB bulk aspect changed from a plasticized polymer to a powder, as also observed by Verbois et al. in dynamic crosslinking of EVA . After that, viscosity and shearing of the melt is so high that the clusters are broken up or dispersed resulting in a decrease in melt viscosity and torque [30-32]. At lower viscosity, chain scission is less and equilibrium between it and crosslinking is achieved. At this point, a stable torque is achieved and is represented by time T5. Mixing times T2 and T4 are the intermediary stages of evolution of crosslinking structure.
Each of the five samples obtained at different mixing times (T1 to T5) were submitted to different processing stages and analyzed by Soxhlet extraction. The first stage is the mixing in the internal mixer (dynamic crosslinking), the second stage is the compression molding at 180[degrees]C (static crosslinking) and the third stage is the storage of the compression molded sample in a humid atmosphere at 60[degrees]C (static crosslinking). The nomenclature adopted based on mixing times and processing stages can be seen in Table 3.
The gel content of the samples (Fig. 7) was observed to increase with time from mixing up to the maximum crosslinking rate. There after it was constant although there was a reduction in viscosity and torque, stage T4-M to T5-M. The decrease in torque was caused either by the dispersion of the clusters or chain scission which modified the viscosity without reducing the gel content. At the equilibrium between chain scission and crosslinking the gel content was approximately 60%. This was the maximum gel content achieved under all conditions adopted
Static compression molding significantly increased the degree of crosslinking due to the absence of shear and chain scission. Sample T1-M, which was obtained with the shortest mixing time, had a gel content of 26% after mixing. However, the gel content increased to 70% after compression molding (T1-C), a similar result was obtained with the other samples (T2-C to T5-C). These results indicate that the maximum gel content is achieved during compression molding by static crosslinking. In addition, it is clear that contact with moisture did not significantly affect the gel content once the values obtained after molding and after storage in a humid atmosphere are practically the same. Therefore, storage in a humid atmosphere does not contribute significantly to achieving the maximum degree of crosslinking.
The mixing time changes the crosslinked structure of the melt . Shear and temperature, provided by the internal mixer, break, and disperse the crosslinked clusters formed during dynamic crosslinking, resulting in a decrease of viscosity (torque reduction after mixing time T3) without a significant impairment of gel content (as when comparing T3-M and T5M). This structure comprises unmodified chains to larger clusters of high molecular weights and can be imagined as a continuum of clusters of different sizes. These structures allow the processing of crosslinked PVB, where the unmodified chains and the microsize clusters act as a thermoplastic phase . During compression molding, static crosslinking occurs all over the PVB and may join clusters, forming a different structure. Depending on the mixing time adopted, the ratio of dynamic and static crosslinking may vary. For example, T1-C has a ratio of static crosslinked larger than T3-C, which may result in different final structures. These differences may alter various properties of the modified PVB (Fig. 7).
According to Menard , dynamic mechanical analysis can be a qualitative indicator of the relative differences in molecular weight (MW) and molecular weight distribution (MWD) in polymers. This approach was developed using the Doi -Edwards theory , where the crossover point between the storage modulus (E') and the complex viscosity ([[eta].sup.*]) can be related to the MW and MWD of polymers. The complex viscosity ([[eta].sup.*]) was calculated by Eq. 7, where E" is the loss modulus and co is the frequency.
[absolute value of [[eta].sup.*]] = [square root of [(E').sup.2] + [(E").sup.2]/[omega] (1)
As molecular weight increases, the viscosity also increases, and the crossover point moves upward (toward higher viscosity). As the MWD increases, the frequency at which the material starts acting elastic increases and the point moves toward higher frequency . Therefore, considering that samples containing larger crosslinked clusters have higher values of MW and the greater the variation of clusters sizes, the larger is the polymer MWD, the structure of modified PVB formed during the mixing process was analyzed by DMA (Fig. 8) ,
Although the samples presented a similar crosslinking degree, analysis shows that T3-C has a higher MW and a lower MWD than T1-C, while T5-C presented intermediate values. This may indicate that at T1-C and at the beginning of dynamic crosslinking (gel point), there are small clusters with different sizes (low MW, high MWD) that grow to form larger and more uniformed clusters (high MW, low MWD) in T3-C. The shear inside the mixer becomes high enough to break and disperse the large clusters, resulting in an intermediate structure in T5-C. A representation of the structures formed by dynamical crosslinking is shown in Fig. 9.
Concluding, the ideal mixing time should be set before the maximum crosslinking rate (torque peak) is achieved, to avoid excessive mixing times and degradation. The crosslink degree, achieved before the torque peak, is not high enough to prejudice post processing, and crosslinking reactions will continue during molding, ensuring maximum solvent resistance. These are important results once they indicated that melt mixture of PVB-VTMS could be performed in an extruder without excessive crosslinking and viscosity increase. What would allow subsequent molding (compression or injection) to desired form, where maximum crosslinking degree is achieved, ensuring solvent resistance.
The maximum gel content obtained by the samples was approximately 78% (w/w). Therefore, there was approximately 22% (w/w) of material that was not extensively crosslinked by VTMS, avoiding solubilization. The PVB used in this study was a residue from the automotive industry, so the exact composition of the polymer was unknown. According to the literature, the PVB used in this application contains 10%--35% (w/w) of plasticizer . A new Soxhlet extraction was performed in neat PVB to quantify and identify the type of plasticizer contained in the PVB. The solvent used was hexane because it can solubilize the most common additives used in PVB but does not interact with the polymer . After extraction, hexane was evaporated, leaving an extracted of 15.6% (w/w). This shows that most of the polymer was modified by VTMS and that the material not affected was basically the plasticizer and other additives. The extracted sample was analyzed by FTIR, and the spectrum can be seen in Fig. 10.
According to the literature, the most common plasticizers used in PVB for sandwiched laminated glass are alkyl phthalate, dibutyl sebacate, triethylene glycol bis (2-ethylhexanoate), or dihexyl adipate [1, 5-7]. The FTIR spectrum of the hexane soluble extracts after the solvent evaporation was consistent with that of an aliphatic ester. The intense bands at 3,447, 2,958, 1,734, 1,172 and 1,137 [cm.sup.-1] are characteristic of ester groups and are indistinguishable from the spectrum of dibutyl sebacate. This spectrum is in agreement with the experiments conducted by Dhaliwal , who showed that dibutyl sebacate was the main plasticizer used in PVB.
Thermogravimetric Analysis of Modified PVB
The thermal stability of the T1-C to T5-C samples was evaluated by thermogravimetric analysis (TG). The results are shown in Fig. 11. Weight loss occurred in three different regions.
Dhaliwal and Hay  studied the thermal decomposition of PVB using a thermogravimetric unit attached to a Thermolab mass spectrometer and the degradation from 200[degrees]C to 250[degrees]C using FTIR analysis. They observed a weight loss of 20-25% in temperatures below 250[degrees]C. From 200[degrees]C to 250[degrees]C, the infrared (IR) spectrum showed a large reduction in the intensities of the carbon hydrogen stretching and bending modes and the carboxyl absorption band at 1,730 [cm.sup.-1]; however, little or no change was observed in the bands attributed to butyral groups. Based on these results, they concluded that the weight loss below 250[degrees]C was due to the loss of plasticizer. Above 260[degrees]C, the major products of decomposition were initially butyraldehyde and butenal, which were obtained by elimination of butyral groups. Above 380[degrees]C, there was also acetic acid as a minor component, from the elimination of acetate. The low concentration of the acetic acid in the volatile was consistent with its content in the original PVB.
Based on this study and observing Fig. 11, the weight loss of approximately 17% from 175[degrees]C to 250[degrees]C can be associated with the loss of plasticizer. This amount of plasticizer is in agreement with the value estimated through the Soxhlet extraction with hexane, which was 15.6% (w/w). The crosslinking reactions with VTMS had no significant influence on this decomposition stage. However, the size of the crosslinked clusters seems to influence the initial decomposition of butyral groups above 260[degrees]C (Fig. 1 lb). The PVB molecule is composed of approximately 80% (w/ w) of butyral groups and 20% (w/w) of vinyl alcohol , which is responsible for the crosslinking reaction with VTMS. Therefore, the crosslinked clusters were agglomerates of butyral groups linked by the vinyl alcohol and VTMS. Decomposition of the butyral groups inside the large clusters demands more energy, resulting in the displacement of the maximum decomposition rate. Samples T1-C, T4-C, and T5-C, which have small clusters with good dispersion, had an almost identical behavior as the neat PVB, with a maximum decomposition rate of 293 [+ or -] 2[degrees]C. Samples T2-C and T3-C, which have larger clusters, had a maximum decomposition rate at 305[degrees]C and 303[degrees]C, respectively. These samples also exhibited certain decomposition at approximately 290[degrees]C, likely due to the small clusters still present in the structure. The elimination of butyral groups continued above 325[degrees]C, and acetate group decomposition occurred at 380[degrees]C.
The objective of this study was to improve the solvent resistance of PVB through crosslinking reactions with a silane agent (VTMS) to enlarge the recycling of PVB as a flexible polymeric matrix for composites.
The reactions between PVB and VTMS were effective to improve the solvent resistance of the polymer against organic solvents. Soxhlet extraction showed that the gel content of modified PVB can reach approximately 78%. Mechanical tests indicated that tensile strength, elongation at break and hardness of PVB were not significantly altered by chemical modification and that the polymer maintained its rubbery aspects.
A more detailed study of chemical modification showed that the crosslinking reactions occurred dynamically in the mixer and statically during compression molding. The maximum gel content obtained by dynamic crosslinking for certain process parameters was approximately 60%, while static crosslinking reached a value larger than 70%, independent of the previous processing. Torque monitoring suggested that the crosslinked clusters formed during mixing were broken and dispersed due to the high shear stress imposed by the bulk viscosity increase, causing the torque to decrease over a long mixing time. Depending on the mixing time adopted, the ratio of dynamic and static crosslinking varied, resulting in a different final structure. Frequency scan obtained by DMA indicated that dynamic crosslinking reactions, in a first stage, formed bigger and more uniformed clusters (high MW with low MWD) until the viscosity of the bulk is so high, that shear stress is enough to break clusters, resulting in smaller clusters with different sizes (lower MW with higher MWD). Based on these results, the ideal mixing time should be set before the maximum crosslinking rate (torque peak) is achieved. The crosslink degree achieved before the torque peak is not high enough to prejudice the molding to the desired form, and crosslinking reactions will continue during post processing, ensuring maximum solvent resistance.
The FTIR analysis of the possible plasticizers present in PVB indicated dibutyl sebacate as the main component. Thermogravimetric analysis of the modified PVB showed that the plasticizer lost may begin at 175[degrees]C, and the chemical modification does not affect this weight loss. However, the size of the crosslinked clusters influenced the initial decomposition of butyral groups, once the larger clusters showed displacement of the maximum decomposition rate.
The chemical modification of PVB through melt mixing with VTMS was efficient to improve the solvent resistance to ethanol without deeply compromising the flexible features of the polymer. The ideal process should melt mix the components without excessive dynamic crosslinks to allow the molding in which the maximum crosslinking degree and solvent resistance are reached through static crosslinking. The possibility to modify PVB in traditional process in industry, such as extrusion (melt mixture) and compression or injection molding (static crosslinking) is an important contribution to enlarge recycled PVB applicability.
[1.] T.S. Valera and N.R. Demarquette, Eur. Polym. J., 44, 755 (2008).
[2.] A.K. Dhaliwal and J.N. Hay, Thermochim. Acta, 391, 245 (2002).
[3.] M. Hajian, G.A. Koohmareh, and M. Rastgoo, J. Appl. Polym. Sci., 115, 3592 (2010).
[4.] M.D. Fernandez, MJ. Fernandez, and P. Hoces, J. Appl. Polym. Sci., 102, 5007 (2006).
[5.] N.M.S. El-Din and M.W. Sabaa, Polym. Degrad. Stabil., 47, 283 (1995).
[6.] D. Morais, T.S. Valera, and N.R. Demarquette, Macromol. Symp., 245-246, 208 (2006).
[7.] J.D. Ambrosio, A.A. Lucas, H. Otaguro, and L.C. Costa, Polym. Compos., 32, 776 (2011).
[8.] D. Gheysari and A. Behjat, Eur. Polym../., 37, 2011 (2001).
[9.] S. Rouif, Nuci. Instrum. Meth. Phy. Res. Sect. B, 236, 68 (2005).
[10.] M.J. John and R.D. Anandjiwala. Polym. Compos., 29, 187 (2008).
[11.] T. Bremner and A. Rudin, J. Appl. Polym. Sci., 49, 785 (1993).
[12.] M. Horn and H.-J Kotzsch, U.S. Patent 5,282,998 (1994).
[13.] J. Cha and J.L. White, Polym. Eng. Sci., 41, 1227 (2001).
[14.] M.S.C. Kumar and M. Alagar, Eur. Polym. J., 38, 2023 (2002).
[15.] S. Dassin, M. Dumon, F. Mechin, and J.P. Pascault, Polym. Eng. Sci, 42, 1724 (2002).
[16.] C. Lu, T. Chen, X. Zhao, X. Ren, and X. Cai, J. Polym. Sci.: Part B: Polym. Phys., 45, 1976 (2007).
[17.] Z. Wang, X. Wu, Z. Gui, Y. Hu, and W. Fan, Polym. Int., 54, 442 (2005).
[18.] H.C. Kuan, J.F. Kuan, C.C.M. Ma, and J.M. Huang, J. Appl. Polym. Sci., 96, 2383 (2005).
[19.] F.W. Fabris, F.C. Stedile, R.S. Mauler, and S.M.B. Nachtigall, Eur. Polym. J., 40, 1119 (2004).
[20.] B.I. Chaudhary, S.S. Sengupta, J.M. Cogen, and M. Curto, Polym. Eng. Sci., 51, 237 (2011).
[21.] T.L. Metroke and M.V. Henley, Progr. Org. Coat., 69, 470 (2010).
[22.] F. Beari, M. Brand, P. Jenkner, R. Lehnert, H.J. Mettemich, J. Monkiewicz, and H.W. Siesler, J. Organomet. Chem., 625, 208 (2001).
[23.] H. Schmidt, H. Scholze, and A. Kaiser, J. Non-Cryst. Solids, 63, 1 (1984).
[24.] A. Ghosh-Dastidar, S.S. Sengupta, J.M. Cogen, L.H. Gross and S.F. Shurott, "Experimental and Theoretical Aspects of Kinetics in the Moisture Cross-Linking Reaction for Power Cables," in Proceedings of International Wire and Cable Symposium, Somerset, New Jersey (2007).
[25.] A.B. Pessanha, M.C.G. Rocha, and A.H.M.F.T. da Silva, Polimeros: Ciencia e Tecnologia, 21, 53 (2011).
[26.] A.V. Krasnoslobodtsev and S.N. Smirnov, Langmuir, 18, 3181 (2002).
[27.] Y. Xie, C.A.S. Hill, Z. Xiao, H. Militz, and C. Mai, Compos.: Part A, 41, 806 (2010).
[28.] I. Hasegawa and S. Sakka, Bull. Chem. Soc. Jpn., 61, 4087 (1988).
[29.] A. Thitithammawong, C. Nakason, K. Sahakaro, and J. Noordermeer, Polym. Test., 26, 537 (2007).
[30.] A. Verbois, P. Cassagnau, A. Michel, J. Guillet, and C. Raveyre, Polym. Int., 53, 523 (2004).
[31.] A. Msakni, P. Chaumont, and P. Cassagnau, Polym. Eng. Sci., 46, 1530 (2006).
[32.] C.H. Wu and A.C. Su, Polym. Eng. Sci., 31, 1629 (1991).
[33.] K.P. Menard, Dynamic Mechanical Analysis: A Practical Introduction, 2nd ed., CRC Press, USA (2008).
[34.] K.I.R.R. Rahalkar, Rheol. Acta, 28, 166 (1989).
Marilia Sonego, (1) Lidiane Cristina Costa, (2) Jose Donato Ambrosio (1,2)
(1) Center for Characterization and Development of Materials,, Federal University of Sao Carlos (UFSCar), P.O. Box 388, Sao Carlos, SP, 13565-970, Brazil
(2) Department of Materials Engineering, Federal University of Sao Carlos (UFSCar), P.O. Box 388, Sao Carlos, SP, 13565-970, Brazil
Correspondence to: J.D. Ambrosio; e-mail: donaIo@ccdm.ufscar.br
Published online in Wiley Online Library (wileyonlinelibrary.com).
TABLE 1. Parameters adopted during melt mixing and compression molding of PVB modified with VTMO. VTMO [T.sub.process] [T.sub.2min] Condition (wt%) ([degrees]C) ([degrees]C) 1 0 180 178.4 2 1.22 180 177.8 3 1.55 180 185.5 4 0 150 167.0 5 0.6 150 168.2 6 0.6 150 168,5 7 0.92 150 166.2 8 0.92 150 166.2 9 1.22 150 165.2 Rotation [t.sub.mix] [T.sub.molding] Condition (rpm) (min) ([degrees]C) 1 50 7 180 2 50 2 180 3 50 7 180 4 100 2 150 5 100 2 150 6 100 7 150 7 100 2 150 8 100 2 180 9 100 2 150 TABLE 2. Mechanical properties of neat PVB and PVB modified with VTMO. Tensile Elongation Condition strength (MPa) at break (%) 1 21.6 [+ or -] 2.1 233 [+ or -] 9 2 -- -- 3 -- -- 4 20.7 [+ or -] 1.0 231 [+ or -] 10 5 18.5 [+ or -] 2.5 213 [+ or -] 14 6 8.16 [+ or -] 1.4 (a) 134 [+ or -] 9 (a) 7 18.9 [+ or -] 0.8 210 [+ or -] 13 8 19.0 [+ or -] 1.8 222 [+ or -] 16 9 19.3 [+ or -] 1.5 221 [+ or -] 12 Hardness Condition shore A 1 64.6 [+ or -] 0.5 2 -- 3 -- 4 56.5 [+ or -] 1.0 5 56.5 [+ or -] 1.7 6 51.8 [+ or -] 1.3 (a) 7 57.2 [+ or -] 2.1 8 59.1 [+ or -] 1.3 9 58.9 [+ or -] 2.2 (a) Irregular specimen: crosslinking affected conformation. TABLE 3. Nomenclature adopted based on mixing times and chemical modification stages. Mixing time Stages T1 T2 T3 T4 T5 Mixing T1-M T2-M T3-M T4-M T5-M Mixing and T1-C T2-C T3-C T4-C T5-C compression molding Mixing, compression T1-H T2-H T3-H T4-H T5-H molding, and humid storage
Please note: Some tables or figures were omitted from this article.
|Printer friendly Cite/link Email Feedback|
|Author:||Sonego, Marilia; Costa, Lidiane Cristina; Ambrosio, Jose Donato|
|Publication:||Polymer Engineering and Science|
|Date:||Sep 1, 2016|
|Previous Article:||Preparation and characterization of a novel hydrophilic PVDF/PVA UF membrane modified by carboxylated multiwalled carbon nanotubes.|
|Next Article:||Critical role of depressurization and effects of saturation conditions in the formation of [beta]-crystal during isotactic polypropylene foaming with...|