Thermosetting Polymers Part 8 pot

8 Preparation of Modified Thermosets 8.1 INTRODUCTION Thermoset precursors are frequently formulated by adding other components than monomers: these components, which are generically called modifiers, include small molecules, oils, low- or high-molar mass rubbers or thermoplastics, etc. Depending on the required applications and desired properties, the amount of modifier may vary in a broad range, from about 2 to 50 wt% of the total mixture with monomers. When rubbery domains in the micrometer range are randomly dispersed in a rigid thermoset matrix, the fracture energy and the toughness can be greatly enhanced (Chapter 13). The use of modifiers is required not only to improve toughness but also for other specific needs such as the dimensional stability of a molded part or the increase of the elastic modulus during the cure of a preimpregnated composite (prepreg), or the generation of microporous structures for thermal or electrical insulation, etc. Again, as in many other fields covered in the book, modified epoxies are the most studied systems (toughened epoxies for adhesive coatings and composites). But also rubber-modified phenolics and low-profile unsaturated polyesters for sheet and bulk molding compounds have been exten- sively studied. There are two main procedures used to generate a second phase in a modified thermoset: 1. The initial mixture is a dispersion of preformed (crosslinked or not) particles in the thermoset precursors and remains heteroge- neous during the cure. 2. The initial mixture is homogeneous, and phase separation takes place during the cure of the thermoset. This second technique is called reaction-induced phase separation (Williams et al., 1997) and may lead to several types of morphologies: a dispersion of modifier-rich particles in a thermoset matrix; a dispersion of thermoset-rich particles in a modifier matrix (phase-inverted morphology), or two bicontinuous phases. 8.2 MODIFIERS INITIALLY MISCIBLE IN THERMOSET PRECURSORS 8.2.1 Solubility of Main Modifiers Used a. Requirements Modifier miscibility plays an important role in this preparation. On the one hand, modifiers must be miscible with the reactive system; on the other hand, they must phase-separate during cure. Final morphologies are influ- enced by the phase-separation conditions. b. Rubbers A family of carboxy-terminated poly(butadiene co-acrylonitrile) liquid rubbers, abbreviated as CTBN (Table 8.1), was pioneered by BF Goodrich in the late 1960s and early 1970s and introduced commercially as versatile epoxy-toughening agents. They are also called reactive liquid polymers, RLP. Due to their low glass transition temperatures and low molar masses, less than 4 kg mol À1 , they are viscous liquids. Miscibility between CTBN and epoxy monomers such as diglycidyl ether of bisphenol A, DGEBA, is matched by incorporating polar acrylonitrile units into a nonpolar immiscible polybutadiene backbone, and also by modifying end groups, e.g., by converting carboxy into epoxy end groups via adduct formation (Riew and Gillham, 1984; Riew, 1989; Chen et al., 1994). Examples of experimental phase diagrams are given in Fig. 8.1, for different liquid rubbers and the same liquid epoxy prepolymer (Verche ` re et al., 1989). Cloud-point measurements using light transmission or light scat- tering are generally used to plot the phase diagram and the cloud-point curve. Liquid rubber modifiers exhibit an upper critical solution tempera- Preparation of Modified Thermosets 227 228 Chapter 8 TABLE 8.1 Examples of some modifiers for thermosetting polymers Acronym Name Formula Some characteristics CTBN Poly(butadiene- co-acrylonitrile) M n $ 3500 g mol À1 T g $À608C f COOH $ 1.8–2.0 wt % AN: 26, 18, 10, 0 PES Polyether sulfone T g ¼ 2208C PEI Polyether imide T g ¼ 2108C PPE Poly(2,6 dimethyl 1,4-phenylene ether) T g ¼ 2108C HOOC ( CH 2 ) CH 3 C CH 3 [ R ] CH 3 C CH 3 (CH 2 )COOH 2 2 CH ( CH 2 - CH ) CH 2 ( CH 2 - CH ) CN R = ( CH 2 - CH = CH - CH 2 ) a b c with SO 2 O N C C O O O C CH 3 CH 3 O N C C O O HO CH 3 CH 3 H Preparation of Modified Thermosets 229 PMMA Poly(methyl methacrylate) T g ¼ 1108C SAN Poly(styrene-co-acrylonitrile) T g ¼ 1008C PVAc Poly(vinyl acetate) T g ¼ 408C CH 2 C CH 3 COOCH 3 CH 2 CH CH 2 CH CN CH 2 CH O-COCH 3 ture behavior (UCST), i.e., the miscibility increases with increasing temperature. As shown in Fig. 8.1, formulations containing an initial volume fraction of modifier, f M0 in the range 10–20 wt% are miscible at T i >658Cin the case of CTBN (18 wt% AN), and T i > 1558C in the case of CTBN (10 wt% AN), while CTBN (26 wt% AN) is totally miscible with DGEBA. The state of miscibility of any mixture is governed by the Gibbs free energy of mixing, ÁG, which may be described by the lattice theory of Flory–Huggins (Flory, 1953), as follows: ÁG RT ¼ f TS V TS ln f TS þ 1 V M  Æ f M;i i  ln f M;i þ w V ref f M f TS ð8:1Þ A unit cell is defined with a molar volume V ref . This reference volume may be selected as the molar volume of the thermoset precursor(s), V TS ,or as the molar volume of the constitutive repeating unit of the modifier, V M .R is the gas constant and f TS and f M,i are the volume fractions of the thermoset precursor(s) and of the modifier i-mer, respectively; f M ¼ Æf M;i : In principle, the interaction parameter  must be considered as a function of temperature, T, composition, and the average degree of polymerization of the modifier component. If  decreases with temperature, then 230 Chapter 8 FIGURE 8.1 Cloud-point temperatures versus volume fraction of modifier, for mixtures of diglycidyl ether of bisphenol A, DGEBA ( " nn ¼ 0:15) with two CTBN copolymers with different acrylonitrile content: 18 and 10 wt%. (Reprinted from Verche ` re et al., 1989, Copyright 2001, with permission from Elsevier Science) phase diagrams show a UCST behavior. When  increases with temperature, they show a lower critical solution temperature, (LCST) behavior. The experimental cloud-point curves are fitted by selecting an adequate function for . The interaction parameter  can also be estimated by the use of solubility parameters (d);  is proportional to (d M À d TS Þ 2 . But this approach has a considerable error (Ád in the range of Æ 0:4 MPa 1=2 Þ, and considers only the excess free enthalpy. For these reasons it is better to determine the miscibility window experimentally. Although CTBN and derivatives still constitute the most important group of modifiers used in rubber-modified epoxies, several other types of modifiers, such as vegetal oils (castor oil), have been proposed as well. CTBN and derivatives are also used in unsaturated polyester and vinylester resins but with less success. Studies have shown that the elastomer additive alone is very often immiscible in a polyester resin. c. Thermoplastic Modifiers In the 1980s a novel approach to toughen epoxies, consisting of replacing the rubber by an engineering thermoplastic, was developed. Table 8.1 gives the formula of some of the thermoplastics used and Fig. 8.2 gives examples of the phase diagrams. In this case in addition to the cloud-point curve, the Preparation of Modified Thermosets 231 FIGURE 8.2 Schematic phase diagrams for thermoplastic-epoxy monomer (diglycidyl ether of bisphenol A) blends, (CPC ¼ cloud point curve, and VC ¼ vitrification curve). (a) and (b) UCST (upper critical solution temperature) behaviour for PPE and PEI (respectively) – DGEBA n ¼ 0.15; (c) LCST (lower critical solution temperature) behaviour for PES-DGEBA g n ¼ 0.15. (Pascault and Williams, 2000 – Copyright 2001. Reprinted by permission of John Wiley & Sons Inc.) transformation diagrams have to be completed by plotting the vitrification curve. The vitrification curve for a one-phase system can be estimated by the Fox equation: 1 T g ¼ W TS T g;TS þ W M T g;M ð8:2Þ Figures 8.2a and b describe a UCST behavior, while Fig. 8.2c represents an LCST behavior. It is interesting to note that in the case of an UCST behavior the cloud-point curve will usually intersect the vitrification curve, while this may not be the case for an LCST behavior. Experimental cloud-point curves are fitted by Eq. (8.1), selecting an adequate function for . Depending on their structures, thermoplastics are more or less soluble in epoxy monomers: poly(methyl methacrylate), PMMA and poly(styrene-co-acrylonitrile), SAN are quite soluble in liquid DGEBA, but the other thermoplastics shown in Table 8.1 are only partially miscible (Pascault and Williams, 2000). In some cases semicrystalline thermoplastics have been used as modifiers (polyamides, polyesters). With these thermoplastics it is necessary to increase the temperature beyond the melting temperature to obtain a solution. Another important application is the introduction of a thermoplastic such as poly(vinyl acetate), PVAc, into an unsaturated polyester resin to improve dimensional stability of the mold part. Keeping in mind (sec. 2.3.2) that an unsaturated polyester is dissolved in styrene, the modifier is the third component and the behavior is described with the help of a ternary-phase diagram (Fig. 8.3, Suspe ` ne et al., 1991). 8.2.2 Description of the Reaction-Induced Phase Separation Process a) Case of Stepwise Polymerization During a step-growth reaction the molar mass of the product increases gradually and the molar mass distribution becomes continuously wider (Sec. 2.2.1). Figure 8.4 gives a scheme of the reaction-induced phase-separation process in a transformation diagram temperature, T, versus composition, f M , for a UCST behavior. The initial formulation has a volume fraction of modifier equal to f M0 and is kept at the cure temperature, T i , where it is a homogeneous solution. As conversion increases at T i , the solubility of the modifier decreases, due to the increase in the average size of the distribution of the thermosetting species and the corresponding decrease in the entropic contribution to the free energy of mixing. 232 Chapter 8 Preparation of Modified Thermosets 233 FIGURE 8.3 Phase diagram for styrene (S)–unsaturated polyester (UP) prepolymer–PVAc ternary blends at T ¼ 238C,–&– UP: M n ¼ 1690 g.mol À1 and M w / M n ¼ 7.5; –o– UP: M n ¼ 1480 g.mol À1 and M w /M n ¼ 3.1. The dashed triangle represents the formulation range of typical industrial UP resins. (Reprinted from Suspe ` ne et al., 1991, Copyright 2001, with permission from Elsevier Science) FIGURE 8.4 Temperature vs. composition transformation diagram for a modified thermoset with an upper critical solution temperature (UCST) behavior (Crit ¼ critical point; for a and b see text). Looking back to Eq. (8.1), the absolute value of the first term decreases with increasing conversion. The interaction parameter  may also vary along the reaction, due to the continuous change in the chemical structure of the thermoset. This secondary effect may favor mixing or demixing, depending on whether  decreases or increases, respectively, with conversion. Phase separation begins when the cloud-point curve reaches the point a at (f M0 ,T i ). This is defined as the cloud-point conversion, x CP , which is usually lower than the gel conversion, x gel . As a result of polydispersity effects, the composition of the incipient b- phase segregated at the cloud point is located on a shadow curve, outside the cloud-point curve (point b in Fig. 8.4). The effects of polydispersity on phase diagrams and phase compositions may be found in specialized reviews (Tompa, 1956; Kamide, 1990; Williams et al., 1997). Because f M0 < (f M,crit (x CP ), the incipient b-phase, which is richer in the modifier, will be dispersed in the a-phase, which is richer in the growing thermosetting polymer. The opposite occurs when f M0 > f M,crit (x CP ). It has been shown both theoretically (Riccardi et al., 1994 and 1996; Williams et al., 1997), and experimentally (Bonnet et al., 1999) that 1. The b-phase rich in modifier, contains the higher-molar mass species of the modifier as well as a significant amount of thermosetting species, mainly monomers and low-molar-mass species. Therefore the conversion in the b-phase is lower than in the a- phase. On the contrary, the a-phase is richer in low-molar-mass species of the modifier and the high-molar mass species of the thermosetting polymer. 2. As conversion increases beyond x CP , the a and b-phases become richer in their main components. 3. Most of the primary morphology development is arrested at gelation of the a-phase. 4. A secondary phase separation may take place inside the b-phase, and due to diffusion effects it leads to a sub-structure of thermoset-rich domains. Secondary phase separation may be arrested by gelation or vitrification of the b-phase. 5. When the modifier is a high-T g thermoplastic, a postcure step at T>T g,M is necessary to complete the reaction in the b-phase and to develop the final morphology of the blend. In Fig. 8.4 an UCST behavior is represented. The corresponding situation for an LCST behavior is a shift of the cloud-point curve to lower temperatures as conversion increases. A similar description of the phase- separation process is valid also for the LCST case. 234 Chapter 8 The reaction-induced phase separation may also be described using conversion, x, versus composition, f M , transformation diagrams at a constant cure temperature (Fig. 8.5). Cloud point and spinodal curves bound stable, metastable, and unstable regions. Experimental studies of phase separation (Chen et al., 1993; Girard-Reydet et al., 1998), revealed that compositions located close to the critical point (e.g., trajectory 2 in Fig. 8.5) undergo spinodal demixing, while off-critical compositions (e.g., trajectories 1 and 3) exhibit phase separation by a nucleation-growth mechanism. Independently of the mechanism by which phase separation is produced, final morphologies depend primarily on the location of the trajectory. Trajectory 1 leads to a random dispersion of modifier-rich particles in the thermoset-rich matrix, and trajectory 3 leads to the opposite situation. In a composition region located close to the critical point (trajectory 2), bicontinuous structures may be obtained. Preparation of Modified Thermosets 235 FIGURE 8.5 Conversion vs composition transformation diagram at a constant cure temperature, showing cloud-point curves and spinodal curves that bound stable, metastable, and unstable regions; * 1 , *2 , and *3 represent the three trajectories, starting from different initial thermoplastic concentrations and leading to different morphologies. (Pascault and Williams, 2000 – Copyright 2001. Reprinted by permission of John Wiley & Sons Inc.) [...]... is introduced in DGEBA-4,40 diamino diphenyl sulfone, DDS, system precured at 1358C (time > tgel) and then postcured at 2308C Rubber-rich particles are spherical, D $ 2 .8 Æ 0.5 mm, and well dispersed (From LMM Library.) 2 38 Chapter 8 FIGURE 8. 8 SEM photograph of a fully-cured rubber-modified epoxy network The rubber CTBN ( 18 wt% AN) was also pre-reacted with a large excess of DGEBA and then introduced... concentration (a) 10 wt% PEI precured at 80 8C and postcured at 1908C ; (b) 20 wt% PEI, same cure schedule; (c) 10 wt% PEI precured at 1608C and postcured at 1908C; (d) 20 wt% PEI same cure schedule as (c) Phase inversion is around 20 wt% (From LMM Library.) Preparation of Modified Thermosets 241 8. 3 DISPERSION OF AN ORGANIC SECOND PHASE IN THE THERMOSET PRECURSORS 8. 3.1 Introduction Another procedure for... 1358C (T < Tg,PPE) and (b) at 2408C (T > Tg,PPE) (a) Epoxy-rich substructures are visible inside PPErich particles; (b) the particle sizes decrease with an increase in cure temperature and a binodal distribution is observed (From LMM Library.) In some cases, due to coalescence mechanisms (Oswald ripening), bimodal particle-size distributions are observed The interfacial adhesion between dispersed particles... and participate in the crosslinking reaction through the amide groups (Lennon et al., 2000) For this reason, the cure cycle must be selected in order to keep the polyamide particles below their melting point (in the range 1708C or 2208C, depending on the type of polyamide used), and thus keep their initial shape and size But in some cases a partial dissolution of the powder surface can improve the particle–polymer... poly(methyl methacrylate), or polyethylene oxide As for core-shell particles, the quality of the dispersion depends on the degree of miscibility of the stabilizing blocks (Fig 8. 12) For high block copolymer concentrations, nanostructured networks can be prepared (Lipic et al., 19 98) 244 Chapter 8 FIGURE 8. 12 TEM photographs of triblock copolymers dispersed in a DGEBA– diamine epoxy network The triblock... preformed particles in the initial formulation This technique is also well documented for modified thermoplastics (Paul and Bucknall, 2000) In Chapter 7 different macromolecular architectures such block copolymers, crosslinked microparticles, hyperbranched polymers, and dendrimers, were presented (Fig 7.11) All these compact molecules can be used as thermoset modifiers Thermoplastic powders and core-shell polymers. .. exactly the same morphologies (Williams et al., 1997) Emulsifiers, such as block copolymers, also modify both particle size and interfacial adhesion (Girard-Reydet et al., 1999) 240 Chapter 8 (a) (b) (c) (d) FIGURE 8. 10 TEM photographs of thermoplastic-modified epoxy networks The epoxy system is DGEBA–MCDEA (same as in Fig 8. 9), and the thermoplastic is a polyetherimide, PEI The figure illustrates both... particle–polymer network interactions Amorphous polyimide powders prepared by dissolution/precipitation processes, can be used to toughen thermosetting polymers Polyethylene powders are frequently used in low-shrink unsaturated polyester formulations 8. 3.3 Core-Shell Particles Core-shell polymers were commercially introduced as impact modifiers for poly(vinyl chloride) PVC, in the 1960s They are produced by a two-stage... crosslinked poly(butadiene), random copolymers of styrene and butadiene, 242 Chapter 8 poly(butyl acrylate), and copolymers The cores keep their size during the thermoset cure Core diameters can range from about 50 nm to 5 mm For toughening applications, a high amount of elastomeric core relative to the total particle is desirable; e.g., from about 50 to 90 wt% of the particles The shell component is grafted... usually miscible with epoxy monomers but may phase-separate during reaction If this happens, a partial aggregation of particles may take place The interfacial adhesion between dispersed particles and the matrix can be improved by functionalizing the core-shell particles with any chemical group that can react with the thermosetting polymer For example, glycidyl methacrylate can be introduced in the shell composition . 227 2 28 Chapter 8 TABLE 8. 1 Examples of some modifiers for thermosetting polymers Acronym Name Formula Some characteristics CTBN Poly(butadiene- co-acrylonitrile) M n $ 3500 g mol À1 T g $À608C f COOH $. of dispersed phase and the average size of particles is expected. 2 38 Chapter 8 FIGURE 8. 8 SEM photograph of a fully-cured rubber-modified epoxy network. The rubber CTBN ( 18 wt% AN) was also pre-reacted with. mol À1 T g $À608C f COOH $ 1 .8 2.0 wt % AN: 26, 18, 10, 0 PES Polyether sulfone T g ¼ 2208C PEI Polyether imide T g ¼ 2108C PPE Poly(2,6 dimethyl 1,4-phenylene ether) T g ¼ 2108C HOOC ( CH 2 ) CH 3 C CH 3 [

Định dạng
Số trang	22
Dung lượng	1,15 MB