INVESTIGATION OF NETWORK DEVELOPMENT AND PROPERTIES ...
INVESTIGATION OF NETWORK DEVELOPMENT AND PROPERTIES IN MULTIFUNCTIONAL EPOXY RESINS USING 3,3'-DIAMINODIPHENYLSULFONE Jeremy O. Swanson, Eric W. Fowler, Steven M. Wand, Monoj Pramanik and James W. Rawlins School of Polymers and High Performance Materials The University of Southern Mississippi 118 College Drive # 5217 Hattiesburg, MS 39406-0001 ABSTRACT Proper development of molecular architecture in polymers is vital to ensure long-term performance viability as it dramatically impacts both strength and toughness. In this study, epoxy resins cured with 3,3'-diaminodiphenylsulfone were evaluated for differences in molecular architecture and properties. The effect of conversion on network development, connectivity, morphology, and mechanical properties was investigated via thermal and thermo-mechanical techniques. 1. INTRODUCTION Epoxy resins are an important class of high performance materials characterized by good mechanical and thermal behavior, resistance to solvents and corrosive agents, low shrinkage upon cure, and easy processing.1 The network development that takes place in the curing of epoxies controls their long-term performance.2 Network development is affected by several factors including prepolymer composition, curing agent, degree of cure and level of connectivity.3 Higher crosslink density and network connectivity can be achieved by decreasing epoxy equivalent weight and/or increasing resin functionality. Epon® 828 is a diglycidyl ether of bisphenol-A (DGEBA) based epoxy resin. DGEBA epoxy resins are the traditional resin of choice due to their easy variance of molecular weights and acceptable performance in many applications. Since DGEBA is a bifunctional epoxide, multifunctional epoxides are preferred for high temperature materials.3-4 Araldite MY0510 is an epoxy resin composed of triglycidyl-p-aminophenol (TGAP) possessing two glycidyl-amines and one glycidyl-ether. Araldite MY721 is a N,N,N',N'-tetraglycidyl-4,4'diaminodiphenylmethane epoxy prepolymer with four glycidyl-amine groups. Epon 828, MY0510, and MY721 each have aromatic backbones, but provide differences in molecular network architecture development and connectivity due to their chemical makeup and functionality (Figure 1).3-8
Epon 828 O
O
N
O
O
MY0510
O
O
N
O
CH2
MY721
N
O
Figure 1. Chemical structure of di-, tri-, and tetra-functional epoxies Near infra-red spectroscopy (NIR) is a method capable of monitoring network development by following epoxy conversion along with the appearance and disappearance of amine moieties. NIR has advantages over alternative methodologies for studying the curing of epoxy resins. Mid-IR spectroscopy can be used for monitoring epoxy cure, but the overlapping bands in the mid-IR region increase the difficulty in analyzing spectra.5 However, the NIR spectra does not contain the same degree of overlapping bands and provides a better method to follow epoxy cure. Differential scanning calorimetry (DSC) can also be used to observe epoxy cure but the data only offers insight into the degree of cure and not the network development occurring during cure. Dynamic mechanical analysis (DMA) is a technique capable of identifying key parameters related to polymer network structure. 2,7,8 Conversion and heterogeneity of polymer networks can provide valuable insight into mechanical responses under a wide variety of environments and be readily determined from modulus, damping, and Tg characteristics.2,7,8 Additionally, DMA is known to provide accurate, reliable, and reproducible measurements of crosslink density and molecular weight between crosslinks. This work investigated epoxy conversion along with epoxy, primary, secondary, and tertiary amine concentrations during network development between multifunctional epoxies and a bifunctional epoxy. The effect of meta-substitution on mechanical properties was simultaneously explored using 3,3'-diaminodiphenylsulfone as the crosslinker (Figure 2). The degree of cure was correlated to the Tg and network heterogeneity determined by DMA.
H2N
NH2 O S O
Figure 2. Chemical structure of 3,3’-diaminodiphenylsulfone (3,3'-DDS)
2. EXPERIMENTAL 2.1 Materials The multifunctional epoxy resins used in this work, i.e., triglycidyl-p-aminophenol (Araldite MY0510) and N, N, N',N'-tetraglycidyl-4,4'-diaminodiphenylmethane (Araldite MY721) were supplied by Huntsman Advanced Materials, LLC. The bifunctional epoxy resin used in this work was the diglycidyl ether of bisphenol-A (Epon® 828) supplied by Hexion Specialty Chemicals, Inc. 3,3'-diaminodiphenylsulfone (3,3'-DDS, 98%) was purchased from TCI, America. All chemicals were used as received. 2.2 Film Preparation Epoxy resins were mixed with stoichiometric amounts of 3,3'-DDS in an oil bath at 100-125 °C until the blend turned transparent. For NIR spectroscopic analysis, the epoxy resin-3,3'-DDS mixture was sandwiched between two glass slides with a polytetrafluoroethylene spacer to achieve a standard sample thickness of 0.16 mm. The resin mixture was transferred to heated silicone molds and cured at 125 °C for 5 hours followed by post-cure at 200 °C for up to 3.5 hours. 2.3 NIR IR Study NIR spectra were obtained using an Antaris II, FT-NIR Analyzer from Thermo Scientific. Absorption spectra were recorded in the NIR region of 4000-8000 cm-1 (32 scans, resolution 4 cm-1) with integrating sphere via OMNIC version 7.3 software. Epoxy, primary amine, and secondary amine concentrations were calculated from the areas under their respective peaks using Lambert-Beer’s law while tertiary amine concentration was back calculated by subtracting primary and secondary amine concentrations from initial total amine concentration. 2.4 Dynamic Mechanical Analysis DMA studies were conducted on a DMA Q800 (TA Instruments, Inc.) in tensile mode. The free films were approximately 1 mm thick x 5 mm wide and analyzed at a frequency of 1.0 Hz (heating rate of 2 °C/min from 25 °C to 300 °C under nitrogen) using strain control. Three samples of each formulation were evaluated to ensure reproducibility.
3. RESULTS 3.1 NIR Spectroscopy Figure 3 shows the NIR spectra (4400 – 5200 cm-1 and 7500 – 6300 cm-1) of Epon 828 cured with stoichiometric ratios of 3,3'-DDS. Each spectrum was normalized using the phenyl band at 4622 cm-1. Decreases in the primary amine band at 5066 cm-1, combined band of epoxy and primary amine groups at 4531 cm-1, and a combined band of primary and secondary amine groups at 6569-6678 cm-1 were observed over time while the hydroxyl band at 6991 cm-1 increased in intensity as the cure proceeded. In addition, these trends were evaluated using the same process with a stoichiometric ratio of 3,3'-DDS to MY0510 and MY721.
Figure 3. NIR spectra of Epon 828 cured with stoichiometric amounts of 3,3'-DDS Conversion data evaluated from the NIR spectra of Figure 3 is shown in Figure 4. Initial monomer concentration is recorded as time -1 h to represent pre-mixing. Epoxy conversion reached 80.82% within 3 hours at 125°C before the onset of vitrification limited conversion to 87.08% after 5 hours at 125 ºC. Post-curing at 200 °C for 1 hour increased the conversion to 96.52%. Since curing the samples for an additional two hours at 200 °C only enhanced the conversion to 96.68%, it was determined that a 1 hour post-cure was sufficient to maximize cure.
Figure 4. Conversion data for epoxy, primary and secondary amines for stoichiometric ratios of Epon 828 with 3,3'-DDS. To determine the connectivity and gain insight into network architecture, it is important to track amine species throughout the polymerization. Relative concentrations of epoxy, primary and secondary amines are shown in Figure 5. As expected, the primary amine is rapidly consumed early in the reaction and accompanied by a simultaneous increase in secondary amine concentration. Secondary amine formation yields linear segments and it is not until the concentration of primary amines is exhausted that the secondary amine species begin to react and form branches. After three hours of cure, vitrification can be noted upon conversion of ~50% of the secondary amine species. Post-curing allows the remaining functionalities to react until mobility restrictions prohibit further conversion.
Figure 5. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of Epon 828 with 3,3'-DDS. The NIR evaluated conversion profile for MY0510 – 3,3'-DDS is shown in Figure 6. As before the initial monomer concentration is taken to be at time -1 hour. Within 2 hours, the epoxy conversion surpassed 79.44%. Subsequent reaction was limited by vitrification as indicated by epoxy conversion remaining at 84.07% after 5 hours at 125 °C. Increasing the temperature to 200 °C for post-cure raised the epoxy conversion to 99.44% but an additional hour of cure at 200 °C only increased the conversion to 99.50%. The MY0510 system reached its vitrification level at 125 °C, sooner than the Epon 828 system due to its higher functionality and lower critical conversion for network development.
Figure 6. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of MY0510 with 3,3'-DDS. The appearance and disappearance of amine moieties in the MY0510 – 3,3'-DDS system is shown in Figure 7. Just as with the Epon 828 system, the loss of primary amine is followed by a rise in secondary amine concentration. Even though the secondary amine and tertiary amine moieties do not reach full conversion, it is noted that the epoxy and primary amine concentrations drop significantly. TGAP (MY0510) contains a tertiary amine diepoxide group that can enhance the reactivity of neighboring groups and favor etherification reactions. At low primary amine concentrations towards the end of the reaction, the tertiary amine diepoxide increases the etherification level, which accounts for 27% epoxy conversion without a decrease in secondary amine concentration.
Figure 7. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of MY0510 with 3,3'-DDS. The MY721 – 3,3-DDS conversion profile is shown in Figure 8. The epoxy conversion for the MY721 system reaches 71.28% percent after two hours but further progress is limited due to reduced chain mobility. During post-cure at 200°C, the conversion rises sharply to 99.77% after 1 hour but does not progress much further in the next hour (99.78%). Of the three resins investigated here, the MY721 resin has the highest functionality and thus the lowest critical conversion for network development. Due to the increased functionality, the MY721 system (like the MY0510 system) reached vitrification at 125 °C within two hours, compared to the Epon 828 system that took 4 hours. The structure of these epoxies also affected the extent of reaction occurring before vitrification. TGDDM (MY721) contains the most sterically hindered epoxy groups after partial reaction, followed by MY0510 with Epon 828 being the least hindered. It is seen that vitrification is related not only to reduce chain mobility from network formation, but also to steric hindrance.
Figure 8. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of MY721 with 3,3'-DDS. The change in amine concentration in the MY721– 3,3'-DDS system over time and temperature is shown in Figure 9. Again, the loss of primary amine precedes the rise in secondary amine concentration. As noted with the MY0510 system, the epoxy and primary amine concentrations drop significantly, though the secondary and tertiary amine moieties do not reach full conversion levels. The complete epoxy conversion without corresponding full conversion of secondary amines is attributed to etherification as seen with MY0510.
Figure 9. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of MY721 with 3,3'-DDS. 3.2 Dynamic Mechanical Analysis Dynamic mechanical analysis was used to study the moduli, glass transition temperature (Tg), and network development during cure of Epon 828, MY0510, and MY721 with 3,3'-DDS. The storage modulus and tan data for the Epon 828– 3,3'-DDS system are shown in Figure 10. Increases in the onset softening temperature are related to increased conversion and connectivity, and are also observed via increased rubbery modulus and Tg. They can be directly correlated to increased crosslink density. Between the second and third hours of cure, a large change was noted in the onset of softening and rubbery modulus. The positive change in the slope of the rubbery modulus of the under-cured samples is indicative of further reaction taking place as vitrification limits are surpassed. The peak of the loss factor (tan δ) indicates an increase in Tg from 128.0 °C to 179.5 °C as curing proceeds from two hours to post-cure. The reduction in peak height of the tan δ curve indicates increased crosslinking.
Figure 10. Dynamic mechanical analysis of various stages of cure for stoichiometric ratio of Epon 828 with 3,3'-DDS. DMA results of the MY0510 system are shown in Figure 11. After one hour of cure, the Tg as determined from primary tan maxima is ~178.0 °C, which correlates to the initial vitrification point. As the network forms, the onset temperature of the primary transition increases with cure time, with the initial conversion limited by the onset of vitrification at ~175 °C. The Epon 828 system does not exhibit a low temperature transition because the epoxide reactivity is independent of each other and network formation proceeds by the formation of linear segments
prior to building crosslink density and developing its highest Tg at later stages of cure. Consequently, vitrification occurs after significant conversion late in the reaction. In the case of multifunctional epoxides, network formation and subsequent vitrification occurs rapidly at lower conversion due to the high system functionality.
Figure 11. Dynamic mechanical analysis of various stages of cure for stoichiometric ratio of MY0510 with 3,3'-DDS.
Results from the MY721 DMA tests are shown in Figure 12. The data is similar to that obtained with the MY0510 system, as both show two transitions and reach vitrification at ~175 °C. The primary transition is reduced as areas of low crosslink density continue to convert and expand into the connected molecular level network. Since MY721 has a symmetrical structure, each epoxide has equal reactivity and results in uniform network development. As the reaction progresses, the network becomes homogenous and the first transition is decreased.
Figure 12. Dynamic mechanical analysis of various stages of cure for stoichiometric ratio of MY721 with 3,3'-DDS.
Figure 13 compares the storage modulus and tan plots of the post-cured Epon 828, MY0510, and MY721 systems. The Epon 828 system exhibited the lowest Tg and crosslink density due to its high equivalent weight and low functionality. Even though Epon 828 network approached an ideal network as represented by the tan peak shape, the MY0510 and MY721 networks exhibited higher Tg and crosslink density because of their increased functionality and lower equivalent weight.
Figure 13. Dynamic mechanical comparison of post-cured epoxy resins
4. CONCLUSIONS Epoxides based on DGEBA, TGAP and TGDDM were cured with stoichiometric amounts of 3,3'-diaminodiphenylsulfone at 125 °C with additional post-cure at 200 °C. NIR spectroscopy indicated that Epon 828 proceeds through the anticipated cure process of loss of primary amine concentration, increasing secondary amine concentration and finally full conversion to tertiary amine. Both MY0510 and MY721 systems exhibited considerable divergence from this process due to their early onset of vitrification, chemical structure, and etherification. The difference in network development was also evident via DMA where changes in mechanism, reactivity and vitrification were discernible. The Epon 828 system exhibited a primary relaxation in tan with continued conversion being evident only by an increase in rubbery modulus. The increased slope correlates to a rate roughly that of the testing protocol. The MY0510 and MY721 systems exhibited vitrification at ~175 °C that exceeds the cure temperature due to the catalytic tertiary amines within their monomer structure. Despite variability in network architecture during development, the post-cure conditions homogenize the ultimate network architecture. The modulus values and Tgs achieved upon full conversion corresponded well with monomer rigidity. 5. 1.
2. 3. 4. 5. 6. 7. 8.
REFERENCES
Musto, Pellegrino, Abbate, M., Ragosta, G., Scarinzi, G. “A Study by Raman, NearInfrared and Dynamic-mechanical spectroscopies on the Curing Behavior, Molecular Structure and Viscoelastic Properties of Epoxy/anhydride Networks.” Polymer 48 (2007): 3703. Sperling, L. H. Introduction to Physical Polymer Science 4th Ed. New Jersey: John Wiley & Sons, Inc. 2006. Guerrero, P., De la Caba, K., Valea, A., Corcuera, M. A., Mondragon, I. “Influence of Cure Schedule and Stoichiometry on the Dynamic Mechanical Behavior of Tetrafunctional Epoxy Resins Cured with Anhydrides.” Polymer 37 (1996): 2195. Foreman, J.P., Porter, D. Behzadi, P.T., Jones, F.R. “Predicting the thermomechanical properties of an epoxy resin blend as a function of temperature and strain rate.” Composites: Part A (2009) Varley, R. J., Heath, G. R. Hawthorne, G., Hodgkin, J.H. “Toughening of Trifunctional Epoxy System: 1. Near Infra-red Spectroscopy Study of Homopolymer Cure.” Polymer 36 (1995): 1347. Xu, L., Schlup, J.R. “Etherification Versus Amine Addition During Epoxy Resin/Amine Cure: An in Situ Study Using Near-Infrared Spectroscopy.” Journal of Applied Polymer Science 67 (1998): 895. Dusek, K., Plestil, J., Lednicky, F. & Lunak, S. “Are Cured Epoxy Resins Inhomogeneous?” Polymer 19 (1978): 393. Ikkai, F. & Shibayama, M. “Inhomogeneity Control in Polymer Gels.” Journal of Polymer Science Part B Polymer Physics 43 (2005): 617.
Epon 828 O
O
N
O
O
MY0510
O
O
N
O
CH2
MY721
N
O
Figure 1. Chemical structure of di-, tri-, and tetra-functional epoxies Near infra-red spectroscopy (NIR) is a method capable of monitoring network development by following epoxy conversion along with the appearance and disappearance of amine moieties. NIR has advantages over alternative methodologies for studying the curing of epoxy resins. Mid-IR spectroscopy can be used for monitoring epoxy cure, but the overlapping bands in the mid-IR region increase the difficulty in analyzing spectra.5 However, the NIR spectra does not contain the same degree of overlapping bands and provides a better method to follow epoxy cure. Differential scanning calorimetry (DSC) can also be used to observe epoxy cure but the data only offers insight into the degree of cure and not the network development occurring during cure. Dynamic mechanical analysis (DMA) is a technique capable of identifying key parameters related to polymer network structure. 2,7,8 Conversion and heterogeneity of polymer networks can provide valuable insight into mechanical responses under a wide variety of environments and be readily determined from modulus, damping, and Tg characteristics.2,7,8 Additionally, DMA is known to provide accurate, reliable, and reproducible measurements of crosslink density and molecular weight between crosslinks. This work investigated epoxy conversion along with epoxy, primary, secondary, and tertiary amine concentrations during network development between multifunctional epoxies and a bifunctional epoxy. The effect of meta-substitution on mechanical properties was simultaneously explored using 3,3'-diaminodiphenylsulfone as the crosslinker (Figure 2). The degree of cure was correlated to the Tg and network heterogeneity determined by DMA.
H2N
NH2 O S O
Figure 2. Chemical structure of 3,3’-diaminodiphenylsulfone (3,3'-DDS)
2. EXPERIMENTAL 2.1 Materials The multifunctional epoxy resins used in this work, i.e., triglycidyl-p-aminophenol (Araldite MY0510) and N, N, N',N'-tetraglycidyl-4,4'-diaminodiphenylmethane (Araldite MY721) were supplied by Huntsman Advanced Materials, LLC. The bifunctional epoxy resin used in this work was the diglycidyl ether of bisphenol-A (Epon® 828) supplied by Hexion Specialty Chemicals, Inc. 3,3'-diaminodiphenylsulfone (3,3'-DDS, 98%) was purchased from TCI, America. All chemicals were used as received. 2.2 Film Preparation Epoxy resins were mixed with stoichiometric amounts of 3,3'-DDS in an oil bath at 100-125 °C until the blend turned transparent. For NIR spectroscopic analysis, the epoxy resin-3,3'-DDS mixture was sandwiched between two glass slides with a polytetrafluoroethylene spacer to achieve a standard sample thickness of 0.16 mm. The resin mixture was transferred to heated silicone molds and cured at 125 °C for 5 hours followed by post-cure at 200 °C for up to 3.5 hours. 2.3 NIR IR Study NIR spectra were obtained using an Antaris II, FT-NIR Analyzer from Thermo Scientific. Absorption spectra were recorded in the NIR region of 4000-8000 cm-1 (32 scans, resolution 4 cm-1) with integrating sphere via OMNIC version 7.3 software. Epoxy, primary amine, and secondary amine concentrations were calculated from the areas under their respective peaks using Lambert-Beer’s law while tertiary amine concentration was back calculated by subtracting primary and secondary amine concentrations from initial total amine concentration. 2.4 Dynamic Mechanical Analysis DMA studies were conducted on a DMA Q800 (TA Instruments, Inc.) in tensile mode. The free films were approximately 1 mm thick x 5 mm wide and analyzed at a frequency of 1.0 Hz (heating rate of 2 °C/min from 25 °C to 300 °C under nitrogen) using strain control. Three samples of each formulation were evaluated to ensure reproducibility.
3. RESULTS 3.1 NIR Spectroscopy Figure 3 shows the NIR spectra (4400 – 5200 cm-1 and 7500 – 6300 cm-1) of Epon 828 cured with stoichiometric ratios of 3,3'-DDS. Each spectrum was normalized using the phenyl band at 4622 cm-1. Decreases in the primary amine band at 5066 cm-1, combined band of epoxy and primary amine groups at 4531 cm-1, and a combined band of primary and secondary amine groups at 6569-6678 cm-1 were observed over time while the hydroxyl band at 6991 cm-1 increased in intensity as the cure proceeded. In addition, these trends were evaluated using the same process with a stoichiometric ratio of 3,3'-DDS to MY0510 and MY721.
Figure 3. NIR spectra of Epon 828 cured with stoichiometric amounts of 3,3'-DDS Conversion data evaluated from the NIR spectra of Figure 3 is shown in Figure 4. Initial monomer concentration is recorded as time -1 h to represent pre-mixing. Epoxy conversion reached 80.82% within 3 hours at 125°C before the onset of vitrification limited conversion to 87.08% after 5 hours at 125 ºC. Post-curing at 200 °C for 1 hour increased the conversion to 96.52%. Since curing the samples for an additional two hours at 200 °C only enhanced the conversion to 96.68%, it was determined that a 1 hour post-cure was sufficient to maximize cure.
Figure 4. Conversion data for epoxy, primary and secondary amines for stoichiometric ratios of Epon 828 with 3,3'-DDS. To determine the connectivity and gain insight into network architecture, it is important to track amine species throughout the polymerization. Relative concentrations of epoxy, primary and secondary amines are shown in Figure 5. As expected, the primary amine is rapidly consumed early in the reaction and accompanied by a simultaneous increase in secondary amine concentration. Secondary amine formation yields linear segments and it is not until the concentration of primary amines is exhausted that the secondary amine species begin to react and form branches. After three hours of cure, vitrification can be noted upon conversion of ~50% of the secondary amine species. Post-curing allows the remaining functionalities to react until mobility restrictions prohibit further conversion.
Figure 5. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of Epon 828 with 3,3'-DDS. The NIR evaluated conversion profile for MY0510 – 3,3'-DDS is shown in Figure 6. As before the initial monomer concentration is taken to be at time -1 hour. Within 2 hours, the epoxy conversion surpassed 79.44%. Subsequent reaction was limited by vitrification as indicated by epoxy conversion remaining at 84.07% after 5 hours at 125 °C. Increasing the temperature to 200 °C for post-cure raised the epoxy conversion to 99.44% but an additional hour of cure at 200 °C only increased the conversion to 99.50%. The MY0510 system reached its vitrification level at 125 °C, sooner than the Epon 828 system due to its higher functionality and lower critical conversion for network development.
Figure 6. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of MY0510 with 3,3'-DDS. The appearance and disappearance of amine moieties in the MY0510 – 3,3'-DDS system is shown in Figure 7. Just as with the Epon 828 system, the loss of primary amine is followed by a rise in secondary amine concentration. Even though the secondary amine and tertiary amine moieties do not reach full conversion, it is noted that the epoxy and primary amine concentrations drop significantly. TGAP (MY0510) contains a tertiary amine diepoxide group that can enhance the reactivity of neighboring groups and favor etherification reactions. At low primary amine concentrations towards the end of the reaction, the tertiary amine diepoxide increases the etherification level, which accounts for 27% epoxy conversion without a decrease in secondary amine concentration.
Figure 7. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of MY0510 with 3,3'-DDS. The MY721 – 3,3-DDS conversion profile is shown in Figure 8. The epoxy conversion for the MY721 system reaches 71.28% percent after two hours but further progress is limited due to reduced chain mobility. During post-cure at 200°C, the conversion rises sharply to 99.77% after 1 hour but does not progress much further in the next hour (99.78%). Of the three resins investigated here, the MY721 resin has the highest functionality and thus the lowest critical conversion for network development. Due to the increased functionality, the MY721 system (like the MY0510 system) reached vitrification at 125 °C within two hours, compared to the Epon 828 system that took 4 hours. The structure of these epoxies also affected the extent of reaction occurring before vitrification. TGDDM (MY721) contains the most sterically hindered epoxy groups after partial reaction, followed by MY0510 with Epon 828 being the least hindered. It is seen that vitrification is related not only to reduce chain mobility from network formation, but also to steric hindrance.
Figure 8. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of MY721 with 3,3'-DDS. The change in amine concentration in the MY721– 3,3'-DDS system over time and temperature is shown in Figure 9. Again, the loss of primary amine precedes the rise in secondary amine concentration. As noted with the MY0510 system, the epoxy and primary amine concentrations drop significantly, though the secondary and tertiary amine moieties do not reach full conversion levels. The complete epoxy conversion without corresponding full conversion of secondary amines is attributed to etherification as seen with MY0510.
Figure 9. Conversion data for epoxy, primary and secondary amines for stoichiometric ratio of MY721 with 3,3'-DDS. 3.2 Dynamic Mechanical Analysis Dynamic mechanical analysis was used to study the moduli, glass transition temperature (Tg), and network development during cure of Epon 828, MY0510, and MY721 with 3,3'-DDS. The storage modulus and tan data for the Epon 828– 3,3'-DDS system are shown in Figure 10. Increases in the onset softening temperature are related to increased conversion and connectivity, and are also observed via increased rubbery modulus and Tg. They can be directly correlated to increased crosslink density. Between the second and third hours of cure, a large change was noted in the onset of softening and rubbery modulus. The positive change in the slope of the rubbery modulus of the under-cured samples is indicative of further reaction taking place as vitrification limits are surpassed. The peak of the loss factor (tan δ) indicates an increase in Tg from 128.0 °C to 179.5 °C as curing proceeds from two hours to post-cure. The reduction in peak height of the tan δ curve indicates increased crosslinking.
Figure 10. Dynamic mechanical analysis of various stages of cure for stoichiometric ratio of Epon 828 with 3,3'-DDS. DMA results of the MY0510 system are shown in Figure 11. After one hour of cure, the Tg as determined from primary tan maxima is ~178.0 °C, which correlates to the initial vitrification point. As the network forms, the onset temperature of the primary transition increases with cure time, with the initial conversion limited by the onset of vitrification at ~175 °C. The Epon 828 system does not exhibit a low temperature transition because the epoxide reactivity is independent of each other and network formation proceeds by the formation of linear segments
prior to building crosslink density and developing its highest Tg at later stages of cure. Consequently, vitrification occurs after significant conversion late in the reaction. In the case of multifunctional epoxides, network formation and subsequent vitrification occurs rapidly at lower conversion due to the high system functionality.
Figure 11. Dynamic mechanical analysis of various stages of cure for stoichiometric ratio of MY0510 with 3,3'-DDS.
Results from the MY721 DMA tests are shown in Figure 12. The data is similar to that obtained with the MY0510 system, as both show two transitions and reach vitrification at ~175 °C. The primary transition is reduced as areas of low crosslink density continue to convert and expand into the connected molecular level network. Since MY721 has a symmetrical structure, each epoxide has equal reactivity and results in uniform network development. As the reaction progresses, the network becomes homogenous and the first transition is decreased.
Figure 12. Dynamic mechanical analysis of various stages of cure for stoichiometric ratio of MY721 with 3,3'-DDS.
Figure 13 compares the storage modulus and tan plots of the post-cured Epon 828, MY0510, and MY721 systems. The Epon 828 system exhibited the lowest Tg and crosslink density due to its high equivalent weight and low functionality. Even though Epon 828 network approached an ideal network as represented by the tan peak shape, the MY0510 and MY721 networks exhibited higher Tg and crosslink density because of their increased functionality and lower equivalent weight.
Figure 13. Dynamic mechanical comparison of post-cured epoxy resins
4. CONCLUSIONS Epoxides based on DGEBA, TGAP and TGDDM were cured with stoichiometric amounts of 3,3'-diaminodiphenylsulfone at 125 °C with additional post-cure at 200 °C. NIR spectroscopy indicated that Epon 828 proceeds through the anticipated cure process of loss of primary amine concentration, increasing secondary amine concentration and finally full conversion to tertiary amine. Both MY0510 and MY721 systems exhibited considerable divergence from this process due to their early onset of vitrification, chemical structure, and etherification. The difference in network development was also evident via DMA where changes in mechanism, reactivity and vitrification were discernible. The Epon 828 system exhibited a primary relaxation in tan with continued conversion being evident only by an increase in rubbery modulus. The increased slope correlates to a rate roughly that of the testing protocol. The MY0510 and MY721 systems exhibited vitrification at ~175 °C that exceeds the cure temperature due to the catalytic tertiary amines within their monomer structure. Despite variability in network architecture during development, the post-cure conditions homogenize the ultimate network architecture. The modulus values and Tgs achieved upon full conversion corresponded well with monomer rigidity. 5. 1.
2. 3. 4. 5. 6. 7. 8.
REFERENCES
Musto, Pellegrino, Abbate, M., Ragosta, G., Scarinzi, G. “A Study by Raman, NearInfrared and Dynamic-mechanical spectroscopies on the Curing Behavior, Molecular Structure and Viscoelastic Properties of Epoxy/anhydride Networks.” Polymer 48 (2007): 3703. Sperling, L. H. Introduction to Physical Polymer Science 4th Ed. New Jersey: John Wiley & Sons, Inc. 2006. Guerrero, P., De la Caba, K., Valea, A., Corcuera, M. A., Mondragon, I. “Influence of Cure Schedule and Stoichiometry on the Dynamic Mechanical Behavior of Tetrafunctional Epoxy Resins Cured with Anhydrides.” Polymer 37 (1996): 2195. Foreman, J.P., Porter, D. Behzadi, P.T., Jones, F.R. “Predicting the thermomechanical properties of an epoxy resin blend as a function of temperature and strain rate.” Composites: Part A (2009) Varley, R. J., Heath, G. R. Hawthorne, G., Hodgkin, J.H. “Toughening of Trifunctional Epoxy System: 1. Near Infra-red Spectroscopy Study of Homopolymer Cure.” Polymer 36 (1995): 1347. Xu, L., Schlup, J.R. “Etherification Versus Amine Addition During Epoxy Resin/Amine Cure: An in Situ Study Using Near-Infrared Spectroscopy.” Journal of Applied Polymer Science 67 (1998): 895. Dusek, K., Plestil, J., Lednicky, F. & Lunak, S. “Are Cured Epoxy Resins Inhomogeneous?” Polymer 19 (1978): 393. Ikkai, F. & Shibayama, M. “Inhomogeneity Control in Polymer Gels.” Journal of Polymer Science Part B Polymer Physics 43 (2005): 617.
