Abstract
High-performance bismaleimide resin systems (here coined LBMI resin series) based on 4,4-bismaleimidodiphenylmethane, 2,4-bismaleimidotoluene, bisphenol A bisallyl ether, 4,4′-bis[2-(1-propenyl)phenoxy]benzophenone, 2-allylphenol, and cumene hydroperoxide for resin transfer molding (RTM) process with low cure and post-cure temperatures (≦180°C) have been developed. Considering the optimum formulation conditions, the injection temperature is in the range of 70–160°C, and the pot life at 100°C is determined to be approximately 100 min. After curing at 180°C for 6 h, the resins exhibit high thermal resistances, excellent mechanical properties, and exceptionally low dielectric loss. Among others, these findings render the materials suitable for the use as high-performance resins for the production of advanced composites via RTM technique and with low cure temperatures.
Introduction
Advanced polymeric composites represent key materials for modern industrial applications, particularly in polymers used in aerospace industry. Bismaleimide (BMI), one of the most important matrices for advanced polymeric composites, features high mechanical properties, excellent thermal oxidation stabilities as well as good humidity and corrosion resistance. 1 In general, the generated polymer needs to be cured (including post-curing) at high temperatures, that is, 220°C or higher. 2 More importantly, high cure temperatures lead to exceedingly high energy consumption, and the cured products tend to exhibit higher stress characteristics, often with internal defects. As a consequence, poor stability features of resulting materials with insufficient service performances can be observed. 3 –5 Furthermore, reaching high cure temperatures may be impractical in some cases or even unsafe for certain temperature-sensitive components, 6 such as printed circuit boards and on-aircraft repairs. The upper cure temperature for these applications is often limited to 177°C to remain below the auto-ignition temperature for aviation fuel and in order to prevent high temperature transmission into aluminum substructure, 7,8 microelectromechanical system buildups, or complementary metal–oxide semiconductor memory, which could degenerate or be destroyed at temperatures exceeding 200°C. 9 The development of high-performance matrices with lower cure temperatures remains a critical, albeit difficult, goal for both materials scientists and engineers.
Resin transfer molding (RTM), as a processing technique for advanced composites in both civilian and military applications, is receiving increasing attention since the early 1990s most notably due to the low cost, high efficiency, good flexibility, and low environmental impact. RTM for advanced polymer composites has been one of the most important applications in the field of composite process engineering. 10,11 Due to the special processing characteristics (low viscosity, suitable pot life, and good reactivity) of the RTM technique, developing high-performance matrices proves to be crucial in advanced studies of the RTM technique and different from conventional techniques for fabricating composites. Most of the current high-performance matrices often exhibit high viscosity, which also tends to accelerate the reactivity and shorten the pot life of the resin, 12 even though a temperature increase has been shown to reduce the overall viscosity. In order to take full advantage of the RTM technique, present matrices may be modified or new matrix systems may be developed to meet the strict requirements of the RTM technique. Extensive efforts have been devoted to the development of a series of high-performance resins, including epoxy, cyanate esters, and BMI materials. 13 –15 However, using BMI resins with a low melt viscosity in the RTM process still results in a high cure temperature required in the BMI system. Therefore, the development of a BMI resin with combined advantages of low melt viscosities and lower cure temperatures may provide a critical benefit.
In this article, we developed a high-performance BMI resin system with low cure temperature for RTM and we investigated the thermal, mechanical, and dielectric properties of the cured resins in detail.
Experimental section
Raw materials
2-Allylphenol (AP) was purchased from TCI Co., Ltd. (Shanghai, China). Bisphenol A bisallyl ether (BBE) was purchased from Zhongsi Shiye Co., Ltd (Shanghai, China). Cumene hydroperoxide (CH) was purchased from Macklin Biochemical Co., Ltd (Shanghai, China). 4,4-Bismaleimidodiphenylmethane (BDM) and 2,4-bismaleimidotoluene (BMT) were purchased from Honghu Shuangma Advanced Materials Tech Co., Ltd. (Hubei, China). 4,4′-Bis[2-(1-propenyl)phenoxy]benzophenone (MT) was synthesized in our laboratory. 16,17 All reagents and chemicals were used without further purification.
Preparation of LBMI prepolymers
Appropriate quantities of AP, MT, BBE, BDM, and BMT (cf. Table 1) were mixed upon stirring at 130°C for 15 min and appropriate quantities of initiator (CH) were added at about 90°C to obtain a homogeneous liquid. The prepolymers were coined LBMI-1, LBMI-2, and LBMI-3, respectively.
Formulation of the LBMI system.
BDM: 4,4-bismaleimidodi-phenylmethyane; AP: 2-allylphenol; CH: cumene hydroperoxide; BMT: 2,4-bismaleimidotoluene; MT: 4,4′-bis[2-(1-propenyl)phenolxy]benzophe-none; BBE: bisphenol A bisallyl ether; BMI: bismaleimide.
Preparation of cured LBMI resins
LBMI prepolymers were thoroughly degassed at 100°C for 30 min and were poured into a preheated (130°C) metal mold, using a curing procedure of 130°C/2 h + 160°C/2 h and 180°C/4 h in an air oven, successively. After this process, the cured samples were post-cured in an air oven at 180°C for 6 h. Upon completion of the curing process and in order to prevent cracking, the cured samples were slowly cooled down to room temperature. The prepared samples were then subjected to characterization and property evaluation.
Characterization techniques
The rheological measurement has been carried out on a Gemini 200 strain-controlled rotational rheometer (Malvern Instruments Ltd., UK) using a test specimen disc with a diameter of 25 mm and a thickness of 1.0 mm. The top plate was oscillated at 1.0 Hz and 10 Pa stress at a heating rate of 4°C min−1. The lower plate was attached to a transducer that was used to record the resulting torque. Differential scanning calorimetry (DSC) analyses were performed on a SEIKO DSC 6220 (Italy) with sample masses of approximately 5–10 mg under a nitrogen (N2) atmosphere at a heating rate of 10°C min−1. Thermogravimetric (TG) analyses were performed on a Perkin–Elmer TG-DTA 6300 (Waltham, Massachusetts, USA) at a heating rate of 10°C min−1 under a N2 atmosphere and in air at temperatures ranging from 100°C to 800°C. Dynamic mechanical analysis (DMA) experiments were carried out on a SEIKO DMS 6100. The storage modulus (E’) and tan δ were measured at a heating rate of 5°C min−1 from ambient to 350°C at a frequency of 1 Hz and an amplitude of 10 µm in a three-point bending mode. Measurements of the dielectric constant and loss were performed with an Agilent Technologies N5230A (Santa Clara, California, USA) measurement system at 25°C using a two-parallel-plate mode at a frequency ranging from 7 GHz to 18 GHz. The dimensions of each sample were Ø = 50 mm and thickness <2 mm. All samples were dried under vacuum at 105°C for 1 h before performing the measurements.
Results and discussion
Resin formulations
The melting point of BDM or BMT is relatively high, while a mixture of BDM and BMT could result in deep eutectic points. Therefore, we select a BDM/BMT blend to produce a low-viscosity eutectic mixture for the RTM technique. The comonomer and reactive diluent are 4,4′-bis[2-(1-propenyl)phenox-y]benzophenone, BBE, and AP, respectively, which are copolymerized with a mixture of BDM and BMT to yield the low-viscosity prepolymers for the RTM technique. A CH compound is used as an accelerant for low cure temperature. The molecular structure and stoichiometric amounts of all the monomers used in the reactions are shown in Figure 1 and Table 1, respectively.
BMI and comonomer building blocks:
Processing properties of LBMI prepolymers
The processing properties of a resin represent one of the most important key factors for the RTM technique. The properties include the melt viscosity, process window, and pot life. Generally, the initial viscosity of a resin should be lower than 500 mPa·s to avoid a high injection pressure, potentially leading to fiber displacements in the mold cavity. The final viscosity should be less than 1000 mPa·s in applications for advanced composites, with fiber volumes above 50%. 12 The dependency of the melt viscosity on the temperature for LBMI resins is shown in Figure 2. From inspection of Figure 2, it becomes clear that LBMI resins retain a rather low viscosity level (below 500 mPa·s) and a significantly wider process window becomes available at temperatures ranging from 70°C to 160°C. Furthermore, we discovered an interesting phenomenon by comparing the LBMI-1, LBMI-2, and LBMI-3 resins: with the increase in BBE content, the value of viscosity among LBMI resins decreased in the following order: LBMI-1 > LBMI-2 > LBMI-3 below 70°C. Conversely, above 70°C, the value of viscosity of the three kinds of resins decreased in the following order: LBMI-3 > LBMI-2 > LBMI-1. However, at room temperature and compared to MT (solid), the BBE compound presents itself as a low-viscosity liquid, and the viscosity of the LBMI resins decreased with an increasing BBE content at low temperatures (below 70°C). Nevertheless, MT can retain a lower viscosity level (below 500 mPa·s) over 75°C leading to rather contrary results. Due to the limits of the initial viscosity value (below 500 mPa·s) by RTM, the recommended injection temperature range to maintain for LBMI resins should be between 70°C and 160°C. Therefore, a stable injection temperature should be determined in view of the pot life.
LBMI system viscosity as a function of temperature.
The term pot life of a RTM resin refers to the length of time that the resin keeps its viscosity lower than 1000 mPa·s, while the resin injection temperature is maintained. 2 The pot life can be obtained from the viscosity time plot at the injection temperature (100°C) as shown in Figure 3. From the viscosity profiles shown in Figure 3, it can be determined that the pot life of all the LBMI resins is more than 100 min, meeting the pot life requirement for fabricating large products with complex shapes as well as small or simple products. Furthermore, the resins can be injected at classical RTM processing temperatures with high efficiency.
LBMI system viscosity as a function of time tested at 100°C.
In order to further investigate the processing properties, the curing reactions of the RTM resins were investigated via DSC curves (cf. Figure 4). All LBMI resins exhibit a main peak at low temperatures of around 190°C and exothermic peaks at around 259°C, indicating that the system can be cured at moderate temperatures. Furthermore, it can be found that the curing peaks at high temperature are largely dependent on the BBE content. Various reports in the literature point out that BBE can translate into diallylbisphenol A (DP) followed by a thermal Claisen rearrangement upon heating at 180°C. 18,19 Moreover, it is known that the curing peak of DP/BDM system occurs at approximately 260°C, so the curing peak of LBMI system at 259°C is stronger upon increasing the BBE content. Furthermore, the LBMI system shows a low reactivity under 100°C from DSC curves and in combination with the viscosity time plot (cf. Figure 3); so we suggest that the injection temperature could be set 100°C (under 500 mPa·s lasts for over 1 h), which proves to be suitable for a RTM process.
DSC curves of the LBMI system. DSC: differential scanning calorimetry.
Properties of the cured LBMI resins
The ability of a polymer to withstand load at elevated temperatures is one of the key criteria in assessing the thermo-mechanical behavior required for studies involving high-performance applications. DMA is a method that measures the stiffness and mechanical damping of a cyclically deformed material as a function of the temperature. Figure 5 shows the E’ and the tan δ data of the cross-linked LBMI resins. At room temperature, the E’ value of the cured resins is approximately 3.8, 4.9, and 5.4 GPa for LBMI-1, LBMI-2, and LBMI-3, respectively, with all remaining at around 1 GPa up to 230°C. The latter finding provides evidence for the notion that highly cross-linked LBMI resins with excellent heat resistance properties were formed. The storage modulus gradually increases with an increasing BBE content. Because of higher molecular weight of MT, the proportion of double bonds in the same amounts of these comonomers taken by weight, BBE/BMI features more double bonds in the structure compared with MT/BMI, leading to a higher cross-linking density after the cured process. The loss tangent (tan δ) is a sensitive indicator of cross-linking, and the value of tan δ decreased with an increase in cross-link density. 20 It becomes clear that the value of tan δ decreases in the following order: LBMI-1 > LBMI-2 > LBMI-3. This further supports the hypothesis that the cross-link density of LBMI system increased in the same order. The glass transition temperature (T g), identified as the temperature corresponding to the maximum of the tan δ peak, was determined to be approximately 250°C, which is 70°C higher than the cure temperature (180°C). This large discrepancy between T g and T cure of LBMI resins, caused by the high cross-link density, may offer a significant processing advantage in the production of high-performance materials without the need of an autoclave.
E′ and tan δ curves, as a function of temperature for the cured LBMI resins. E′: storage modulus.
Figure 6 shows the TG curves of LBMI resins. Further experimental data are also presented in Table 2. The onset temperatures, corresponding to 5% weight loss, are found to be beyond 400°C both in air and N2 and show a char value of approximately 30% at 800°C. Compared with the MT monomer, a weak link with poor thermal resistance, (i.e. –C(CH3)2–), can be found in the molecular structure of BBE, potentially reducing the thermal stability of the formed LBMI system. However, for the same content of MT and BBE in the cured blends, BBE features more double bonds in the structures, leading to a higher cross-linking density after the cure process and ultimately increasing the thermal stability of LBMI. Owing to the high structure density of the benzene rings in MT, the char yields increased upon increasing the MT content. From the foregoing factors influencing the thermal stabilities, LBMI-2 shows a high 5% weight loss (429°C) and char yields (Y c) at 800°C (31%) compared to the other two samples, LBMI-1 and LBMI-3. As we know, the high-temperature curing epoxy resin always needs to be cured at about 160°C–180°C, so the LBMI system can be cocured with epoxy resin to further enhance its heat resistance.
TG curves of the cured LBMI resins in air (a) and under N2 (b). TG: thermogravimetry; N2: nitrogen.
Thermal and flexural properties of the LBMI system.
T d5: decomposition temperature at 5% weight loss; T d30: decomposition temperature at 30% weight loss; Y c: residual weight retention at 800°C under nitrogen; T g: peak temperature in the tan δ curve by dynamic mechanical analysis; N2: nitrogen.
Table 2 lists typical mechanical properties of LBMI resins. It becomes obvious that the flexural strength of LBMI resins increased at room temperature upon increasing the BBE content. As stated above, with the increasing BBE content, the value of cross-link density increased in the following order: LBMI-1 < LBMI-2 < LBMI-3, which enhances the flexural strength of the corresponding LBMI resins. Especially, the LBMI-3 resin exhibits preferable balance between flexural strength and modulus (92 MPa and 4.4 GPa). Even though it cured at 180°C, the cured system also exhibits desirable mechanical properties, which could potentially meet the requirements of normal structural strength for polymer matrices.
Figure 7 highlights the dependency of the dielectric constant (ε) and dielectric loss tangent (tan δ) on the frequency of the cured LBMI resins. It has been reported previously that the dielectric constant can be reduced by increasing the cross-linking density of the structure and by decreasing the polarization. 21 –23 The greater the cross-linking density, the more difficult the chain segment will move, leading to reduced dipole relaxation as well as polarization and ultimately reducing the dielectric constant. As shown above, the storage modulus, flexural strength, and flexural modulus all increased in the following order: LBMI-1 > LBMI-2 > LBMI-3. This finding indicates that the cross-linking density increased in the same order, with the dielectric constant decreasing in the following order: LBMI-3 < LBMI-2 < LBMI-1. Interestingly, all LBMI resins feature a lower dielectric loss (0.007–0.010) compared to the conventional DP/BMI system (0.012). Compared to traditional DP, the chemical structure of MT proves to be highly symmetrical and features almost no polar groups. Presumably, it suggests that these features contribute to the overall reduction of the dielectric loss for the cured LBMI resins. The high density of the rigid benzene ring on MT improves the rigidity and stability of the cured LBMI resins and, potentially, further reduces the dielectric loss. While BBE forms more polar DP, the dielectric loss decreases in the following order: LBMI-1 < LBMI-2 < LBMI-3.
Dielectric constant (a) and dielectric loss (b) at room temperature for the cured LBMI resins.
Conclusions
An RTM BMI resin system based on MT and BBE with low cure temperature (post-cured: 180°C) that can be injected at temperatures as low as 90°C was developed. The thermal, mechanical, and dielectric properties show that the modified BMI resin system not only exhibits excellent processing characteristics, for example, low viscosity, suitable pot life, and long storage life, but also features good thermal properties, high mechanical properties, and low dielectric loss, particularly true for the LBMI-3 resin. Such resin could potentially be used as a matrix for advanced composites using the RTM process with a low cure temperature.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Youth Science Foundation of Heilongjiang Province, China (Grant no. QC2014C008) and the Science Foundation of Heilongjiang Academy of Sciences (no. 2015-YQ-01).