Properties of a Vinyl Ester Resin Modi1ed with a Liquid Polymer KORO DE LA CABA ARANTXA ECEIZA CRISTINA MARIETA MARIA ANGELES CORCUERA PEDRO REMIRO IÑAKI MONDRAGON Departamento Ingeniería Química y del Medio Ambiente, Escuela Universitaria Politécnica, Universidad del País Vasco/Euskal Herriko Unibertsitatea, Plaza Europa 1. 20018 San Sebastián/Donostia, Spain (Received 9 July 20041 accepted 31 January 2005)
Abstract: A vinyl ester (VE) resin was modified with different concentrations of a liquid polymer, polyoxypropylenetriamine (POPTA), in order to study the changes produced during curing and in its final properties. Fourier transform infrared spectroscopy (FTIR) measurements were made to obtain both the styrene and vinyl ester unsaturations conversions during the cure of the resin. The glass transition region was analysed by dynamic mechanical thermal analysis (DMTA), which showed the constancy of glass transition temperature with modifier content. The mechanical properties of the mixtures were also analysed and the results have been related to the flexural behaviour of the cured neat resin.
Key Words: Vinyl ester resins, liquid modifiers, cure schedule, ultimate properties
1. INTRODUCTION Vinyl ester (VE) resins are produced from epoxy resins and unsaturated monocarboxylic acids. Their low room temperature viscosity coupled with rapid curing and relatively low cost make them suitable for various moulding processes [1, 2], such as the conventional unsaturated polyester resins. In addition, vinyl ester resins possess high mechanical strength as well as chemical and solvent resistance, which are not found in most of unsaturated polyester resins. Their hybrid molecular structure of epoxies and polyesters, allows them to challenge even epoxy resins in various application fields. The viscosity of the resin is controlled by adjusting the molecular weight of the vinyl ester and the amount of styrene, which is a reactive diluent and serves as a crosslinking
High Performance Polymers, 17: 605–616, 2005 1 12005 Sage Publications
DOI:10.1177/0954008305053206
606 K. DE LA CABA ET AL.
agent. VE resins are crosslinked by means of free radical mechanisms in presence of styrene, like unsaturated polyesters. Previous studies of such systems show that three types of processes are implied in the curing: resin vinylene group polymerization, resin and styrene C = C bond copolymerization, and styrene homopolymerization, leading to the formation of the so-called microgels, domains of high crosslink density [3–7]. Vinyl ester resins are used as thermoset matrix materials for reinforced composites1 however, they are brittle polymers. One method of toughening, which has achieved significant results with epoxy resins, is the modification of the resin with liquid rubbers, which are added to the resin and precipitate out from the solution during curing. Moreover, enhancement of the resin mechanical behavior can be obtained with relatively low rubber contents [8–11]. According to the literature [12], the initial compatibility of unsaturated polyesters and vinyl esters with most liquid rubbers is poor, making it rather difficult to obtain a rubber solution in a resin. To produce a successful toughening additive it is necessary to use a modifier which is compatible with uncured resin, so that it would readily dissolve in the liquid resin and remain in homogeneous solution until the beginning of the curing reaction. Polyoxypropylenetriamine (POPTA) has been found to be compatible with vinyl ester and such a mixture may produce phase separation under suitable curing conditions, as has been shown for unsaturated polyester (UP) resins in previous studies [13, 14]. The aim of this paper was to investigate suitable conditions to cure systems based on VE resins and analyse the influence of modifier content during the curing and on the final properties. Infrared spectroscopy was used to determine vinyl ester and styrene conversions during curing. Dynamic mechanical measurements were carried out in order to determine the matrix glass transition. Finally, the mechanical behavior of mixtures, which had been modified with different amounts of POPTA, wasexamined and the results have been related to morphological changes.
2. EXPERIMENTAL 2.1. Materials
Derakane 411–350 epoxy vinyl ester resin, provided by Dow Chemical, was used in this study. A 50 wt% methyl ethyl ketone peroxide (MEKP) solution (Diprometil LA-50-R) from Plastiform was used as initiator. The liquid modifier was polyoxypropylenetriamine (POPTA), which was generously given by Texaco, under the tradename of Jeffamine T5000. 2.2. Fourier transform infrared spectroscopy
The reaction kinetics of styrene monomer and vinyl ester C = C bonds was followed in an infrared spectrometer (Perkin-Elmer 16 PC) with a resolution of 2 cm21 in the transmis-
A VINYL ESTER RESIN MODIFIED WITH A LIQUID POLYMER 607
sion mode. After the resin, the modifier, and the initiator were mixed, one drop of mixture was placed between two KBr plates, which were then mounted on a sample holder and located in a temperature-controlled chamber to maintain the reaction temperature constant. Ten scans from 4000 to 400 cm21 were taken at each sampling time. 2.3. Rheological measurements
The ultimate glass transition of the cured resin was measured in a Metravib viscoanalyser. The weight of the sample, introduced into a steel cylinder, 10 mm in diameter, was around 4 g. Oscillatory shearing flow measurements at different isothermal temperatures were made with a device, 1 mm in diameter, at a frequency of 10 Hz for all experiments. The sample was then reheated from 20 to 2503 C at 33 C min21 in order to obtain the Tg4 value. 2.4. Dynamic mechanical thermal analysis
Dynamic mechanical tests were also carried out using a Metravib viscoanalyser and a three-point bending device with a span length of 44 mm. Specimens were machined to 60 mm 5 12 mm 5 5 mm from plaques prepared as described above. Tests at 10 Hz were made at a heating rate of 33 C min21 over the temperature range 20–2503 C. 2.5. Mechanical measurements
The mechanical properties of the cured mixtures were measured at 233 C in a mechanical testing machine, Model 4206 Instron, under conditions specified in the ASTM D-790 standard. Results were obtained using a three-point bending device with a span length of 64 mm. Specimens were machined to 80 mm 5 10 mm 5 5 mm. A 5 kN cell load was employed to produce the test rate of 1.7 mm min21 .
3. RESULTS AND DISCUSSION 3.1. Characterization of the resin 1
H Nuclear magnetic resonance (NMR) spectroscopy, using aVarian VXR 300 MHz device, was used to characterize the resin. The spectrum is shown in figure 1. The two peaks appearing at 5.65 and 6.20 ppm are due to the protons of the vinyl group in vinyl ester, and the two peaks at 5.30 ppm and the two at 5.80 ppm correspond to the protons of the vinyl group in styrene. The peak located at 2.1 ppm is due to the methacrylate methyl end group in vinyl ester and the one at 1.7 ppm to the methyl group of bisphenol-A. NMR study was also used to determine the average molecular weight and the styrene content. The polymerization grade, n, can be calculated by taking the ratio of the area of the peak at 1.7 ppm to that at 2.1 ppm. This ratio was found to be 4, corresponding to a
608 K. DE LA CABA ET AL.
Figure 1. 1 H NMR spectrum for the vinyl ester resin.
VE average molecular weight of 1368 g mol21 . In addition, by calculating the ratio of the styrene vinyl group to that of VE group, taking into account the molecular weight of each monomer and the number of vinyl groups per monomer, this value was determined to be 47 wt% styrene, which is similar to the value determined by evacuating the styrene from the resin in an air-circulating oven at 1103 C for 2 h. The styrene content, calculated from the weight loss, was 46 wt% (DIN 16945). When curing thermosetting resins, the ultimate glass transition value, Tg4 , has to be taken into account when choosing the optimum post-cure temperature, because a lower temperature would lead to an incomplete range of cure. The Tg4 value was obtained by dynamic mechanical analysis, as shown in figure 2. Samples were cured for 90 min at 130, 150 and 1703 C and then reheated from 20 to 2503 C at 33 C min21 in order to obtain the Tg value. Glass transition temperatures were taken as the temperatures corresponding to the maximum values of tan 1 in the 2 relaxation. The values so obtained were a little higher
A VINYL ESTER RESIN MODIFIED WITH A LIQUID POLYMER 609
Figure 2. Loss factor variation upon temperature for the neat resin cured for 90 min at different temperatures: (■) 1303 C1 (6) 1503 C1 (▲) 1703 C.
than 1303 C, so that 1503 C was taken as the post-cure temperature to ensure complete cure of the mixtures. As a comparison, other VE matrices of similar styrene content present glass transitions in the same temperature range [15]. 3.2. Mixing and curing of the system
The VE resin was used as received. The POPTA was mixed with the resin by stirring the mixture and the resultant solution was blended with MEKP at room temperature. The concentration of the initiator employed was 2 wt% of total resin in all the cases, and the POPTA content used was 5, 10 and 20 wt%. All the resin plaques were cast in a mould which consisted of two parallel glass plates separated by a U-shaped 5 mm steel frame spacer and held together by clips. It was necessary to coat all parts of the mould with Frekote 44 release agent before use. The temperatures and times employed in the full cure schedule were: a pre-cure temperature of 803 C for 150 min, degassing under vacuum for a short time, and a post-cure temperature of 1503 C for 120 min. All samples were cooled slowly in the oven after post-cure. The mixtures are initially miscible but they form heterogeneous systems, as it is the usual situation in commercial modified VE systems, which are formed by a continuous VE-rich phase and a discontinuous modifier-rich phase. The neat resin-cured sample was a transparent yellow solid. When the modifier content added was higher than 5 wt%, the cured samples became translucent, with cloudiness increasing gradually with POPTA con-
610 K. DE LA CABA ET AL.
tent. Higher modifier concentrations, 15 and 20 wt%, produced completely opaque white samples. These features of the cured samples correlated with the morphology generated by phase separation induced by curing [16–18]. 3.3. Cure kinetics
In order to analyse the conversion evolution of the mixtures prepared at the cure schedule defined above, Fourier transform infrared spectrophotometry was used. The conversion of styrene, X St , and vinyl ester, X VE , C = C bonds were measured by following the changes in the height of their characteristic peaks. In this study, C = O peak at 1730 cm21 was chosen as the internal standard to correct for thickness changes in the sample during reaction. The consumption of VE resin C = C bonds has been determined from the peak at 945 cm21 , whereas that of styrene C = C bonds was determined from the peak at 910 cm21 . The determination of conversions was carried out taking into account the following equations: X VE 7 1 2
At 39454 A0 39454
X St 7 1 2
At 39104 A0 39104
where A0 and At are the normalized absorbances before the reaction starts and after a certain time t. Total conversion, X, of C = C double bonds was determined as: X7
S V
X St 8 X VE 1 8 VS
where S/V is the styrene/vinyl ester double bonds molar ratio [19]. Figure 3 shows the evolution of the individual VE and styrene conversion for the neat resin cured at 803 C for 150 min followed by 120 min at 1503 C. The routes of conversion of styrene and vinyl ester double bonds are similar at the beginning of the reaction. This is due to the fact that at these early stages of the reaction, the crosslinking density of the reacting system would not be sufficiently high to appreciably influence the propagation mechanisms of styrene and vinyl ester C = C bonds. When the crosslinking density increases, the mobility of the small styrene molecules is less affected than that of C = C units in large vinyl ester molecules. Thus, the conversion of styrene is higher than that of the vinyl ester. The cessation of reaction at values lower than full conversion is due to diffusion limitations and has been widely observed [20–24]. The fact that the mobility of small styrene molecules is less affected by the increase of crosslinking density has also been shown for a similar system, unsaturated polyester (UP), in a previous study [25]. A possible reason could be the formation of microgels during cure, so that double bonds of VE can get
A VINYL ESTER RESIN MODIFIED WITH A LIQUID POLYMER 611
Figure 3. VE (O) and styrene (6) C = C bonds conversion for the neat resin.
trapped inside them owing to the compactness of microgels, as can happen for UP systems [26–30]. Comparing the effect of modifier addition on both conversions, it can be seen in figures 4 and 5 that VE conversion increases with modifier content, whereas styrene conversion decreases. The addition of POPTA could cause a dilution effect which would increase the VE intramolecular reaction, leading to an increase in compactness of the microgels, which, on the other hand, could cause a decrease in styrene conversion due to the greater difficulty in find VE double bonds. 3.4. Dynamic mechanic thermal analysis
Figure 6 reports the loss factor variation upon temperature obtained by dynamic mechanic thermal analysis for the neat resin and for the POPTA-containing mixtures. The Tg values of all mixtures were taken as the temperatures corresponding to the maximum values of the 2 relaxation. This maximum appears around 1233 C for all samples. In the case of the neat resin, this value is slightly lower that the one calculated by the same technique but with different cure conditions (130, 150 or 1703 C for 90 min). This could indicate that the isothermal cure conditions employed have a great influence in the network formed, as has been observed for VE resins by other authors [31]. The constancy of the glass transition temperature with modifier content can be justified by the constancy of the global conversion calculated by FTIR, as shown in table 1.
612 K. DE LA CABA ET AL.
Figure 4. VE C = C bonds conversion for the 5 wt% (■ ), 10 wt% (6) and 20 wt% (▲) POPTA-modified mixtures through curing at 803 C for 150 min and post-curing at 1503 C for 120 min.
Figure 5. Styrene C = C bonds conversion for the 5 w t% (■ ), 10 wt% (6) and 20 wt% (▲) POPTA-modified mixtures through curing at 803 C for 150 min and post-curing at 1503 C for 120 min.
A VINYL ESTER RESIN MODIFIED WITH A LIQUID POLYMER 613
Figure 6. Variation of loss factor and storage modulus upon temperature for the neat matrix (■ ) and for mixtures modified with (6) 5, (▲) 10, and (▼) 20 wt% POPTA.
Table 1. Conversion and Tg values calculated by FTIR and DMTA analysis, respectively.
%wt POPTA X VE X St X Tg (3 C)
0 0.77 0.99 0.93 123
5 0.78 0.96 0.91 123
10 0.79 0.94 0.90 123
20 0.80 0.90 0.88 120
However, the width of the peak increased as the modifier content was higher, which could indicate that the structure of the network formed would be different, due to differences in VE and styrene double-bond conversions, although global conversion is similar. This broadness of the peak could be assigned to a part of the crosslinked network which is less tight and highly irregular due to the incorporation of the liquid modifier, as has been shown by other authors [32]. 3.5. Mechanical properties
Figure 7 is a plot of the stress–strain curves obtained from flexural testing. The mixtures modified with 2 and 5 wt % POPTA showed a similar behavior to that associated with vinyl ester systems, whereas those with the highest contents of POPTA exhibited ductile
614 K. DE LA CABA ET AL.
Figure 7. Stress–strain curves for the neat resin and modified mixtures.
Table 2. Summary of mechanical properties.
%wt POPTA 5 (MPa) 6 (%) E (MPa)
0 141 7.63 3501
2 124 7.94 3126
5 91 8.13 2795
10 88 12.4 2426
15 62 17.5 1688
20 41 21.3 1093
behavior. The mechanical properties of the neat and modified resins are given in table 2. The liquid polymer addition led to a decrease in maximum stress that ranged from 12 to 71%, as well as a decrease in flexural modulus that ranged from 11 to 69%, and the strain at break was significantly enhanced, as anticipated, from 4 to 179%, compared with that for the neat resin. This sharp change in behavior would be in agreement with the morphological variations, as has also been observed by other authors [32, 33]. As stated above, at percentages higher than 5 wt%, the sample becomes macroscopically translucent, and opaque in the case of 15 and 20 wt% POPTA modified ones.
4. SUMMARY In the present study, the influence of modifier content on kinetics, dynamic mechanic thermal and mechanical properties of vinyl ester resins was investigated. The reaction kinetics and morphology in VE-based networks are affected by the modifier content. Early
A VINYL ESTER RESIN MODIFIED WITH A LIQUID POLYMER 615
in the reaction, polymerization passes from chemical to diffusion control so that the final conversion, the polymerization rate, and the final properties depend only on the network mobility and hence on the network structure. The results obtained identify the influence of the cure schedule in improving conversion. Infrared spectroscopy allowed us to obtain the conversion profiles of both styrene and vinyl ester unsaturations that were present in the system. Post-curing increased the conversions of both VE and styrene groups, but complete conversion was not reached due to diffusion limitations. The different modifier content added to the mixtures did not significantly modify the glass transition temperature, showing a good correlation with the global conversion reached which remained practically constant. The addition of the liquid modifier produced a reduction in the mechanical properties. There wais a continuous decrease of flexural modulus and maximum stress of the mixtures as more POPTA was added. There was a dramatic drop of properties for amounts higher than 10 wt% POPTA.
NOTE 1. Author to whom correspondence should be addressed: e-mail:
[email protected]
REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]
Malik M, Choudhary V and Varma IK 2000 J.M.S. Rev. Macromol. Chem. Phys. C40(2&3) 139 Li L, Sun X and Lee L J 1999 Polym. Eng. Sci. 39 646 Dusek K and Spevacek J 1979 Polymer 21 750 Dusek K, Galina H and Mikes J 1980 Polym. Bull 3 19 Yang YS and Suspene L 1991 Polym. Eng. Sci. 31 321 Hsu CP and Lee L J 1993 Polymer 34 4506 Chiu Y Y and Lee L J 1995 J. Polym. Sci., Part A: Polym. Chem. 33 269 Wang G Y, Zhu M Q and Hu C P 2000 J. Polym. Sci., Polym. Chem. 38 136 Huang Y J, Lee S C and Dong J P 2000 J. Appl. Polym. Sci. 78 558 Vilas J L, Laza J M, Garay M T, Rodriguez M and Leon I M 2001 J. Polym. Sci., Polym. Phys. 39 146 Dreerman E, Narkis M, Siegmann A, Joseph R, Dodiuk H and Dibenedetto A T 1999 J. Appl. Polym. Sci. 72 647 Ullet J S and Chartoff R P 1995 Polym. Eng. Sci. 35 1086 de la Caba K, Guerrero P, Franco M and Mondragón I 1997 Macromol. Symp. 114 271 de la Caba K, Guerrero P, Gavaldá J and Mondragón I 1999 J. Polym. Sci.,Polym. Phys. 37 1677 Auad M L, Aranguren M I and Borrajo J 1997 J. Appl. Polym. Sci. 74 1059 Auad M L, Proia M, Borrajo J and Aranguren M I 2002 J. Mat. Sci. 37 4117 Rey L, Galy J and Sautereau H 2000 Macromolecules 33 6780 Gryshchuk O, Jost N and Karger-Kocsis J 2002 J. Appl. Polym. Sci. 84 672 Marroyo L M, Ramis X and Salla J M 2003 J. Appl. Polym. Sci. 89 3618 Cook W D, Simon G P, Burchill P J, Lau M and Fitch T J 1997 J. Appl. Polym. Sci. 64, 769
616 K. DE LA CABA ET AL.
[21] Cook W D 1992 Polymer 33 2159 [22] Verchere D, Sautereau H, Pascault J P, Riccardi C C, Moschiar S M and Williams R J J 1990 Macromolecules 23 725 [23] Auad M L, Aranguren M I, Eliçabe G and Borrajo J 1999 J. Appl. Polym. Sci. 74 1044 [24] Brill R P and Palmese G P 2000 J. Appl. Polym. Sci. 76 1572 [25] de la Caba K, Guerrero P, Eceiza A and Mondragón I 1997 Eur. Polym. J. 33 19 [26] Ganem N, Mortaigne B, Bellenger V and Verdu J 1993 J. Macromol. Sci. A30 11 [27] Dua S, McCullough R L and Palmese G R 1999 Polym. Comp. 20 379 [28] Huang Y J and Chen C J 1992 J. Appl. Polym. Sci. 46 1573 [29] Huang Y J and Chen C J 1993 J. Appl. Polym. Sci. 48 151 [30] Huang Y J and Chen C J 1993 J. Appl. Polym. Sci. 47 1533 [31] Ziaee S and Palmese G R 1999 J. Polym. Sci.1 Part B: Polym. Phys. 37 725 [32] Gryshchuk O, Jost N and Karger-Kocsis J 2002 Polymer 43 4763 [33] Dreeman E, Narkis M, Siegmann A, Joseph R, Dodiuk H and DiBenedetto A T 1999 J. Appl. Polym. Sci. 72 647