Novel Techniques for the Preparation of Different Epoxy/Thermoplastic Blends

  • Xiaole ChengEmail author
  • Jeffrey S. Wiggins
Living reference work entry


Epoxy/thermoplastic blends have received a high degree of attention owing to the fracture toughness improving without significantly compromising the thermal and mechanical properties. These materials are mostly prepared by mixing thermoplastic with epoxy monomers and curatives, and then undergoing a reaction-induced phase separation mechanism to form the final products. The mixing processes are very critical for forming the blends with the optimal properties and processabilities. A poor mixing can lead to localized property variation and deteriorate the morphological and mechanical properties of the final cured blends. Due to the importance of the mixing process, methods used for fabrication of epoxy/thermoplastic blends are described in detail in this chapter. Some novel techniques suitable for specific epoxy/thermoplastic blends are also discussed.


Epoxy/thermoplastic Blends Mixing Phase separation 


Epoxy resins are increasingly being used as coatings, adhesives, and resin matrices for composite materials such as advanced carbon-fiber-reinforced polymer composites. As the network forms during cure, the chemical, morphological, and rheological environments change substantially and the system transits from liquid to gel and to vitrified states. The versatile chemistries and the availability of commercial epoxies and curatives allow designing glassy resin networks with a broad range of properties. In general, glassy epoxy resins are known for their excellent mechanical properties, thermal stability, solvent resistance, manufacturability, and low shrinkage (May 1988). The factor that governs most of these characteristics is the order of the chemical cross-link density, as it ultimately determines the network architecture. However, those highly cross-linked structures of glassy epoxy networks result in inherent brittleness which limits their applications. Therefore, efforts to increase the toughness of glassy epoxy networks continue to be an important area of polymer scientific research.

Rubber elastomers such as carboxyl-terminated butadiene-acrylonitrile rubbers (CTBN) (Arias et al. 2003; Sultan and McGarry 1973) and amine-terminated butadiene-acrylonitrile rubbers (ATBN) (Butta et al. 1986) have traditionally been used as rubber tougheners to improve epoxy toughness. However, rubber tougheners are known to reduce the thermal and mechanical properties of glassy epoxy networks, which is undesirable for advanced application such as aerospace composite materials (Kim and Char 2000; Park and Kim 2001). More recently, linear high-molecular-weight thermoplastics have been used as the tougheners for epoxies with an advantage for preservation of thermal and mechanical properties. The first two toughening studies using thermoplastic additives were reported by Bucknall and Partridge (1983a, b). Commercial polyethersulfone (PES, Victrex 100P) was used in their studies to toughen epoxy resins, and effective enhancement was obtained when thermoplastic phase separated from epoxy matrix combined with good interfacial adhesion. Since then, various types of high-performance linear thermoplastics including polysulfone (PSF) (Kim et al. 1995), polyetherimide (PEI) (Hourston and Lane 1992), polyetherketone (PEK) (Bennett et al. 1991), etc. have been intensively explored in literatures. The selection of the optimum thermoplastic depends on its compatibility, heat resistance, and thermal stability.

Epoxy/thermoplastic blends generally formulated by mixing of a thermoplastic polymer with epoxy monomers and curatives and subsequently curing the mixtures that may undergo the following three possibilities: (a) Thermoplastic is fully miscible in epoxies before curing. Once cure reaction is initiated by heat or UV light, the increased molecular weight of epoxies reduces the miscibility between thermoplastic and epoxy matrix. At a certain extent of cure, system can cross the thermodynamic phase boundaries and enter the unstable region, resulting in a transition from homogeneously miscible state into phase-separated state to give heterogeneous structures such as droplet-dispersed morphology, co-continuous morphology, and phase-inverted morphology. This phase-separating process induced by cure reaction of epoxy thermosets is called cure-reaction-induced phase separation (CRIPS). (b) The initial homogeneous mixture of thermoplastic and epoxy remains miscible during cure process and shows no indication of phase separation after cured. This is most likely due to the chain-extended chemical reaction between thermoplastic reactive chain ends and epoxide groups. The chain-extended thermoplastic is more soluble than thermoplastic itself and suppresses the degree of demixing during cure, resulting a microscale homogenous morphology. The CRIPS can still occur but the domain size was found to be smaller than expected, often in the order of tens of nanometer. In this case, commonly used techniques such as scanning electron microscopy, optical microscopy, and rheological analysis are not able to produce a recognizable feature of phase separation. Epoxy/polycarbonate transesterification (Di Liello et al. 1994) and epoxy/phenolic-terminated PES chain-extended reaction (Blanco et al. 2006) are the best examples. (c) Thermoplastic is immiscible in epoxy matrix before cure and remains heterogeneous during cure, such as polyacrylonitrile (PAN) (Zhang et al. 2012) and poly(vinylidene fluoride) (PVDF) (Kim and Robertson 1992).

Most epoxy/thermoplastic blends of practical interest are prepared through CRIPS mechanism. CRIPS is a promising technique to control the morphologies of epoxy/thermoplastic blends within the range of nanoscale to microscale with tunable thermal and mechanical properties. Phase separation is a kinetic process and can be distinguished into two types of kinetic mechanism: spinodal decomposition (SD) and nucleation and growth (NG). For SD mechanism, a new phase is spontaneously generated from the parent phase at a thermodynamically unstable state. The new phase and parent phase form an interconnected co-continuous structure and can continue growing until they reach their equilibrium states. If the viscosity of the system were sufficiently low, density differences of both phases may lead to dynamic asymmetry between the two components and cause the phase transition from a co-continuous structure to spheres structure with a lower interfacial energy. SD occurs when the system enters the spinodal region of the phase diagram. A fast transition such as quench is typically required to allow the system moving from the stable region through the metastable region and into the unstable spinodal region. For NG mechanism, a series of new nuclei is generated from the parent phase so that the parent phase reduces its composition with a decreased free energy. Once the nuclei are formed, they can grow spherically and may coalesce with each other. In order for nuclei formation, it is required that the system maintain at the metastable state and composition fluctuations is large enough. NG differs from SD in that phase separation only occurs at nucleation sites instead of throughout the parent phase. However, both mechanisms may give the same phase-separated morphology at the end of phase separation if matrix viscosity is sufficiently low.

Since phase separation is a kinetic process and highly viscosity-dependent, the increased molecular weight and viscosity of epoxy matrices upon cure reaction could reduce the phase separation rate and completely suppress when system is vitrified. As a consequence, the final cured morphology of epoxy/thermoplastic binary system is affected by the competition between phase separation kinetics and cure reaction rate. Knowledge of phase separation mechanism and cross-linked reaction kinetics comprise the basis of morphology design for thermoplastic/thermoset systems. Numerous studies, including some reviews (Inoue 1995; Williams et al. 1997), have been devoted to investigate the morphologies of epoxy/thermoplastic blends from different points of view. It has been known that the morphologies can be modified in multiple ways including changing thermoplastic weight fraction and molecular weight (Yu et al. 2008), altering thermoplastic end-group functionality(Ma et al. 2009), and varying epoxy cure chemistry such as cure rate, cure temperature, epoxy-amine stoichiometric ratio, and cure agents (Girard-Reydet et al. 1997; Mimura et al. 2000). General trends have been found that decreasing miscibility of thermoplastic in epoxy before curing or reducing cure rate of epoxy cross-linking chemical reaction could facilitate phase separation process.

The mixing processes of epoxy/thermoplastic blendsare not well discussed in literatures, yet they are very critical for forming the blends with optimal properties and processabilities. An efficient modification of epoxy resin by thermoplastic polymers usually prefers to form a homogenous blend before cure and phase separate after cure. A poor mixing and heterogeneous dispersion of thermoplastic polymers in epoxies can lead to localized matrix property variation and deteriorate morphological and mechanical properties of the final cured blends. Due to the importance of mixing, this chapter will be focused on the preparation techniques for epoxy/thermoplastic blends which are divided into two major parts: mechanical methods and nonmechanical methods.

Mechanical Methods

The mechanical mixers used for blending epoxy and thermoplastic are either batch or continuous types. In batch mixers, all components are added into the mixer vessel together or in a predefined sequence and mixed until a homogenous solution is achieved. On the other hand, in continuous mixers, components are continuously fed into the mixer according to the formulation. The mixing takes place during the material transport from the feeding port to the discharge nozzle. While batch mixers are generally based upon relatively low-temperature mixing over long periods of time, continuous mixers are designed for higher-temperature mixing over short periods of time. Residence time distribution in a continuous mixer is typically in the range of 1–5 min. The selection of batch or continuous mixers depends on several factors such as material physical states, mixing quality and efficiency, handling equipment, energy consumption, and labor cost.

Batch Mixers

Batch mixers are the oldest type of mixers developed for polymer processing and still widely used in polymer society. A typical batch mixer consists of a tank with an agitator and integral heating/cooling system. They have versatile designs and can be modified to suit different polymer applications since the processing conditions can be varied during mixing, additives can be added at a flexible time sequence, and mixing temperature and time can be accurately controlled. Furthermore, they are available in a broad range of sizes from less than 1 L to more than 15,000 L that can be easily accommodated to laboratory-scale trials or industrial-scale production.

There is no standard engineering classification for batch mixer equipment. Different types of mixers may fulfill the same mixing task. Based on their application purposes, batch mixers can be divided into two broad categories: solid mixers and liquid mixers. They can also be classified according to their shear forces: high-shear mixers, medium-shear mixers, and low-shear mixers. Details about different types of batch mixers and their mixing mechanisms have been well discussed and published by J. L. White (White and Bumm 2011), Z. Tadmor (Tadmor and Gogos 2006), and L. A. Utracki (Utracki and Wilkie 2002). This section will be primarily focused on the introduction of batch mixers used for epoxy/thermoplastic blends and their mixing mechanism.

Low-Shear Liquid Batch Mixer

Epoxy/thermoplastic blends have been predominantly prepared using low-shear liquid batch mixers. In a typical batch mixing process, epoxy/thermoplastic blends are prepared in two steps. Epoxies are firstly heated at elevated temperature (usually above 120 °C) to mix with thermoplastics. Thermoplastics in powder form are preferred to facilitate thermoplastic dispersion and solubilization in epoxy resins. Once homogeneous mixing is reached, batch mixer temperature is adjusted and curatives are added with continuing mixing. Degassing is required for both steps to eliminate air bubbles trapped during mixing. Any trapped air will cause negative impacts on material properties. Once curatives are dissolved, the resulted epoxy/thermoplastic curable mixtures may be partially cured in the batch mixer for a period of time and quenched to maintain a target cure conversion for further processing or discharged immediately from the batch mixer with minimal extent of cure, molded to certain shapes, and then fully cured in programmable ovens to form the final products. The choice of cure profile depends on the epoxy curative chemistry.

Myers Engineering Mixer VL550/500A (Fig. 1) is a batch mixer designed specifically for producing epoxy matrix polymers. The system is a dual motor tri-shaft high-viscosity mixer equipped with a hot oil temperature control system and pneumatic ram discharge press. In addition, the system is equipped with a full vacuum assembly to remove reaction solvents, degas during mixing, and assure solubilized gasses are minimized in the matrix materials. Dispersive and distributive mixing are accomplished using the tri-shaft dual motor advanced mixing system which simultaneously incorporates various blade configurations: a flat-bar/low-profile sweep blade with risers and wipers which sweeps the base and sides of the mixer, a distributive mixing gate blade with action which keeps the reaction regulated throughout the mixture assuring homogenous distribution, and a Myers dispersion blade is ultimately responsible for shear motions that reduce aggregate size and distribute particles evenly throughout the matrix. Low-shear liquid batch mixers operating with multiple blades are versatile, making them ideal for handling epoxy liquids over a broad range of viscosities. They are typically capable of processing resin viscosities from water-like to around 50,000 centipoises.
Fig. 1

Myers tri-shaft VL550/500A batch mixer (Courtesy: Myers Mixers, LLC)

Incorporation of a significant amount of thermoplastic into epoxies inevitably raises the matrix viscosity, especially when thermoplastic with high-molecular weight is used. The highly viscous matrix is not efficient for handling by batch mixers and becomes extremely difficult to mix with curatives. As a consequence, the nonuniform dispersion of curatives may disrupt the local resin stoichiometry and affect the way that a cross-link network forms and other desirable properties of final cured blends. Although elevating temperature of batch mixers can reduce the epoxy/thermoplastic blend viscosity, it is impractical to control epoxy cure conversion and prevent runaway reaction due to the substantial heat released by exothermal cure at elevated temperature.

In order to solve this problem and prepare epoxy/thermoplastic blends with relatively high-loading thermoplastic (ca. 20–30 wt. %), diluents, both reactive as well as nonreactive, are often incorporated into epoxy resins prior to adding curatives to reduce epoxy/thermoplastic mixture viscosity and enhance its processability. Commonly used diluents include monofunctional epoxies such as butyl glycidyl ether, octylene oxide, and phenyl glycidyl ether and multifunctional epoxies such as diglycidyl ether bisphenol F (DGEBF) and triglycidyl para-aminophenol (TGAP). Addition of those diluents can significantly reduce resin viscosities, further increase the shelf and pot life and wetting ability, and enable to add a higher amount of thermoplastic. For example, low viscous TGAP (around 700 centipoises at 25 °C) is often used in aerospace tetraglycidyl 4, 4′-diaminodiphenylmethane (TGDDM) (around 5,000 centipoises at 25 °C) with thermoplastic content of up to 30 wt. % to enhance fiber impregnation and control prepreg tack (Zhang et al. 2009). However, it needs to be noticed that diluents adversely affect the cross-link density, thermal stability, and chemical resistance of cured epoxies.

Epoxy/thermoplastic blends with thermoplastic concentration of more than 30 wt. % are difficult to prepare using traditional batch mixer approach even with the aid of dilutes. Above this concentration, thermoplastic-dispersed phase starts to collapse and form a continuous phase where epoxies disperse as droplets. The continuous thermoplastic matrix becomes too viscous to flow in batch mixers and reduces the mixing efficiency. Furthermore, the resulted mixtures are too viscous for many commercially used shaping processes such as prepreg filming and vacuum-assisted resin transfer molding process. Shifting to a different style of batch mixer is recommended in this case.

Internal Mixer

Internal mixers are very efficient batch mixers designed for materials with medium- and high-viscosity with continuous improvement of mixer design and mixing efficiency since the 1930s, internal mixers have been exceedingly used in the rubber and thermoplastic industries. It is a melt mixing process which employs high temperature and intense shear force to obtain homogenous mixing of polymeric materials. As recognized, internal mixer is a versatile and cost-saving process which is capable of producing a variety of thermoplastic polymers and composites in large volume scales.

The internal mixer typically consists of two rotors enclosed in a heating chamber and a drive unit. The two rotors in the mixer serve as its heart for mixing. They may be tangential or intermeshing and with different directions of rotation (counterrotating and corotating) (Moribe 2012). The design of the rotors has a significant influence on its mixing performance. A tangential rotor provides a short mixing time because the clearance between the two rotors is large and the intake is generally fast. However, because the mixing is performed between the rotor and the chamber wall, there is difficulty in distribution and cooling. On the contrary, an intermeshing rotor typically requires a long mixing time since the rotor clearance is small. Due to the mixing that occurs not only between the rotor and the chamber wall but also between the two rotors, it has excellent distribution and cooling performance which is particularly suitable for temperature-sensitive compounds.

Mixing temperature of an internal mixer is dependent on the transfer of heat from chamber surfaces to the polymer matrix and often the limiting factor of mixing performance. The mix chamber is only heated from the outside and the rotors transfer the heat away from the molten polymers, causing temperature drop across the chamber. In order to minimize the temperature variation, the mixing chamber, rotors, and discharge door are all temperature controlled with steam or circulating water.

Filling factor should also be taken into consideration during internal mixing. It defines the proportion of the mixing chamber volume occupied by the finished mixture which can be calculated from the weights and densities of the materials to be mixed. Filling factors in the range of 0.65–0.85 are generally preferred to achieve good level of mixing. Very low filling factors are obviously uneconomic, and excessively high filling factors result in material remaining in the stagnant region of the chamber and not taking part into the mixing process.

Figure 2 presents the example of a laboratory-scaled internal mixer head with non-intermeshing counterrotating mixing rotors produced by C.W. Brabender Instruments, Inc., one of the leading companies for the development of internal mixers. The mixer requires a small amount of raw materials, short residence time, little effort and operational expense, and is often used for the evaluation of compounds on a small scale. It is a three-piece design with a capacity of 350–420 ml depending on mixer blade configurations. The blade configurations are varied including roller, Banbury, cam, and sigma to achieve optimum shape for mixing. The selection of blade shapes is typically dependent on the materials to be mixed, and different blades may fulfill the same mixing task. Recently, Salahudeen et al. (2011) systematically evaluated the distributive mixing performance of various blade configurations of the internal mixer using simulation and verified by tracer experiment. The overall efficiency of distributive mixing in the roller type of mixer was found to be substantially better than cam and Banbury types of mixers due to better length of stretching and reorientation.
Fig. 2

C.W. Brabender Prep-Mixer internal mixer head with blades: roller, Banbury, cam, and sigma (from left to right) (Courtesy: C.W. Brabender® Instruments, Inc.)

Internal mixers are able to handle materials with a very broad viscosity range and therefore particularly useful for preparing epoxy/thermoplastic blends with higher thermoplastic content (above 30 wt. %). The material adding sequence is different from liquid batch mixer. As a general guide, thermoplastic is firstly melted in the chamber and mixed with epoxies. The mixing temperature is set at the thermoplastic-melting temperature to allow it to be quickly converted to melting state in which it will accept liquid additives. After adequate mixing is achieved, the cure agents are added through the feeding port of chamber for further mixing. Since the cure reaction takes place immediately during the continuation of the mixing process, the mixing time for cure agents must be carefully controlled to prevent system gelation and vitrification especially when stoichiometric amounts of cure agents are used. For example, Y. Ishii (Ishii and Ryan 2000) described a process wherein epoxy and thermoplastic polyphenylene ether (PPE) blends were prepared in a Brabender kneader at 185 °C by kneading for about one hour. The mixture was cooled to 150 °C and a stoichiometric mass of curatives was added using a syringe. After mixing for another 2 min, the mixture was quickly removed from the kneader and stored in a freezer.

One of the requirements for epoxy/thermoplastic blend preparation using internal mixers is that the used thermoplastic should have a relatively low melting temperature such as poly (methyl methacrylate), polypropylene, polyethylene, etc. For those high-performance thermoplastics such as PES, PEI, PEEK, etc. with much higher melting temperatures (above 250 °C), the required melt mixing temperature is too high to mix with epoxies and cure agents due to the difficulties in controlling the following cure reaction and preventing side reactions. If gelation or cross-link of epoxy matrix occurs during mixing, the torque increases dramatically which may cause potential damages to the rotors and leave the produced mixtures very difficult to clean after mixing.

Rheo Mixing (RMX)

Another interesting method for preparing epoxy/thermoplastic blends with high thermoplastic concentration is to use Rheo mixing (RMX) device which is recently designed by Muller et al. (Bouquey et al. 2011). Unlike the traditional batch mixers in which shear flows are predominant for mixing, RMX relies on the elongational flows to enhance the efficiency of mixing. As schematically represented in Fig. 3, RMX consists of several major components: two cylindrical chambers with two reciprocally moving pistons to alternately push the materials from one chamber to the other, the central channel connecting the cylindrical chambers, a mixing element with modulated geometry and L/D ratio, feeding units which allow the melting of polymer pellets and liquids entering the central channel, and an outlet port of the mixed materials. The volume of materials inside the mixer can be adjusted by the initial positions of the pistons with a maximum mixing volume of 10–100 cm3. When the materials are forced to pass through the narrow mixing elements, the high shear stress generated by convergent and divergent elongation flows induces the breakup of the agglomerates and results in a strong dispersive mixing. The outlet port of RMX device can be directly connected to a mold. As a result, no additional step (e.g., reheating, placing it into a mold, and molding) is required to mold specimens of specified shapes after mixing.
Fig. 3

Three-dimensional view of the RMX device: (a) chamber, (b) piston with seal, (c) mixing element, (d) feeding unit for melt, (e) feeding channel for liquids, and (f) mold (Reproduced with permission from John Wiley and Sons)

Chandran et al. (2015) was firstly reported using RMX device to prepare epoxy/thermoplastic blends and demonstrated its capability to address the difficulty for mixing high loadings of thermoplastic in conventional batch mixers. In his study, diglycidyl ether bisphenol A (DGEBA)/poly(trimethylene terephthalate) (PTT) blends with PTT content of up to 30 wt. % were mixed in RMX at 245 °C for 10 min. A stoichiometric amount of 4, 4′-diaminodiphenylsulfone (44DDS) was then introduced and mixed for another 10 min at the same temperature. The mixture was directly transferred to a preheated mold at 220 °C and cured for 4 h. The DGEBA/PTT blends prepared by RMX device exhibited significantly improved properties. The cured samples were optically transparent and no phase separation was observed by TEM. The author claimed molecular level dispersion of thermoplastic in epoxy matrix was achieved because of the excellent dispersive mixing of RMX device. The convergent and divergent field near the mixing element induced the breakup of the initially formed PPT-rich phases and pinned further phase separation. In addition, the blends prepared by RMX device showed a lower glass transition temperature and much higher elastic modulus as compared to the conventional batch mixing samples. Surprisingly no transporting issue was mentioned about mixing DGEBA/PTT/44DDS blends at 245 °C for 10 min in RMX whereas the system can be vitrified in less than 5 min in conventional batch reactors. This is likely caused by the inefficiently distributive mixing of RMX device which is being investigated and not reported yet.

Continuous Mixers

Despite being widely used for epoxy/thermoplastic blends, batch mixer approaches are still facing several challenges in industry. Firstly, the batch mixing process is energy intensive. Epoxides, thermoplastics, and curatives are mixed and reacted in the batch mixers which may exceed 1,000 L volume through slow heating. Once homogeneously dispersed, the mixtures are discharged from the batch mixers and stored at low temperature for indefinite periods of time to retard cure and preserve processability. Subsequently, the curable epoxy/thermoplastic mixtures are removed from cold storage, heated above their glass transition temperatures, converted into different shapes, and fully cured in ovens. The total energy consumption throughout this entire process is substantial. Furthermore, products prepared by batch mixers often suffer batch-to-batch variations. This behavior counts for the inherent inconsistencies in quality and leads to variation in the final cured blend structures and properties. Therefore, there is an increasing demand from manufacturers to reduce the energy consumption required to produce epoxy/thermoplastic blends and to minimize batch-to-batch variation in quality.

Continuous mixers carry material as a flowing stream where components are fed into the mixer, mixed, and discharged continuously. They typically consist of several feeding systems for handling materials with different physical states, one or more rotating screws that are capable of transporting and mixing, and a discharge end (die). Continuous mixers are operated under steady-state conditions. The power consumption is usually lower than in batch operations. Although continuous mixers require high capital investment, they are easy to automate and robotize, have high output, and can be run with a statistical quality loop control (Moad 1999). Due to these advantages of continuous mixers over batch mixers, more and more attention has been drawn into preparation of epoxy/thermoplastic blends using continuous mixers.


Extruder is one of the most commonly used continuous mixers for polymer processing. It has drawn much attention in both academia and industry since the 1960s as it is a continuous and economic method for industrial scale manufacturing. Literature reviews about twin-screw extruder mixing process have been published by Brown and Orlando (1988) and Xanthos (1992).

The primary advantage of extruders compared to other mixers, such as batch mixers, is the capability of transporting materials over a broad range of viscosities. Additionally, the absence of solvent combined with simultaneous transport of low-molecular-weight monomers and high-molecular-weight polymers improves energy consumption making the mixer environmentally favorable. Extruders also provide controlled shear energy, excellent heat transfer, precision feeding, mixing, devolatilization, and insensitivity to viscosity changes. They can divide into three categories according to their number of screw shafts including single-screw extruder, twin-screw extruder, and multiple-screw extruder.

Single-screw extruder is a simple and economic continuous mixer type. The main sections of the extruder include the barrels, a single screw that fits inside the barrel, a motor-drive unit, a control system for the barrel temperature and screw speed, and a die to discharge the molten materials. A hopper is attached to the barrel at upstream of the extruder and the materials are fed into the barrel either by gravity feeding or starve feeding using screw flights. Once materials are fed into the barrel, they are compressed and transported by drag flow, which generated from the contact between materials, barrels, and the moving screws. The screw within the barrel can be divided into three zones: solids conveying, melting, and metering zone. The solid-conveying zone begins at the feed port and extends to a point where the solid starts to melt. The screw design in this zone is characterized by screw elements with deeper flight between the root and the tip of the screw. The melting zone follows the solid-conveying zone and is used to compress the solid materials. Within this zone the screw channel depth decreases and the solids coexist with its melt. The metering zone begins at the point where all the solids are melted and it extends to the discharge of the extruder. It acts as a pump, transferring and homogenizing the molten materials and building up the pressure to force the material through the die. When mixing is important, dispersing screw elements with narrow slits can be incorporated in the metering zone to promote dispersion. To enter the slit, the materials are exposed to an increased shear force caused by the elongational flow. But compared to other mixers especially twin-screw extruder, the mixing efficiency provided by single-screw extruder is still very limited.

There are many types of twin-screw extruders. The main geometrical features that distinguish them are the sense of rotation and the degree of intermeshing. Twin-screw extruders with screws that rotate in the same direction are called corotating extruders. When the screws rotate in the opposite direction, it is called counterrotating extruder. The degree of intermeshing can vary from fully intermeshing, partially intermeshing, to non-intermeshing.

Among all types of extruders, fully intermeshing corotating twin-screw extruders offer the highest level of mixing, dispersion, and shear control, making them the primary technology used as continuous mixers (Rauwendaal 1981, 1998). Figure 4 shows a typical twin-screw extruder based on ZSK 26MC Compounder with L/D = 40. The processing section is the core technology of twin-screw extruders. Barrels along the extruder are the points where various liquids, solid reactants, catalysts, modifiers, vacuums, etc. are introduced. Each barrel is equipped with independent temperature control by adjusting heating and cooling. Mixer shafts and screw elements are precisely fit within the series of barrels. It offers a broad array of screw element configurations that provides necessary transport, mixing, and shear abilities. Modular screws are designed by placing appropriate screw elements in their proper positions according to the type of action favorable to accomplish specific reactions or activities within specified regions of the mixers.
Fig. 4

Corotating intermeshing twin-screw extruder based on ZSK 26MC Compounder (Courtesy: Coperion, Inc.)

Typically, three types of screw elements are used including conveying elements, kneading elements, and reversing elements. Examples of modular screw elements used for intermeshing corotating twin-screw extruders are illustrated in Fig. 5. Classical conveying elements are used to convey materials away from feed opening or discharge processed materials at the end of the extruder. They also serve as drivers to provide forwarding pressure that supplies material to kneading and mixing elements. Conveying elements can have various flights with different pitches. Single-flight elements have a deep channel and are useful when good conveying is required. Double-flight screws have a medium channel depth and moderate conveying ability. Triple-flight elements have shallow channels and low conveying capability. They are useful when considerable shear is necessary. The double-flight elements are most commonly used in reactive extrusion, as they have the largest reactive volume combined with the minimum shear work input. Kneading blocks are the dominant elements in determining the mixing efficiency and the degree of fill as well as residence time distribution. They are usually used when materials have to be sheared and dispersively mixed. Kneading blocks are staggered at an angle, which is called advance angle. These advance angles determine the conveying ability of the elements ranging from forwarding (right) to neutral to reversing (left). Neutral elements push material neither forward nor backward. Reverse kneading blocks have a retaining character and are usually utilized when large mechanical stress needs to be built up. Mixing elements (ZME) meet the challenge of conveying and mixing simultaneously. They provide more space for distributive mixing without or nearly no loss in forwarding properties.
Fig. 5

Screw elements for corotating twin-screw extruder

During continuous processing, mixing mechanism is generally categorized into dispersive mixing and distributive mixing. For dispersive mixing, a critical stress is applied to the dispersant through laminar shear stress generated along the screw elements which overcomes cohesive forces of particulates so phase sizes are reduced. High shear rates are a requirement for successful dispersive mixing. In contrast, distributive mixing is more effectively carried out by shear stress that generates large strains as there is no critical stress threshold. Distributive mixing is facilitated by splitting and reorienting the flow streams. Figure 6 illustrates the dispersive and distributive mixing in a particle/liquid system. When considering screw element geometries, wide kneading blocks with reverse pitch facilitate dispersive mixing while narrower kneading blocks with forward pitch, gear, and tooth elements provide distributive mixing.
Fig. 6

Schematic dispersive mixing and distributive mixing mechanism

The advantages of twin-screw extruders are related to the fact that some thermoplastic polymers are difficult to mix with epoxies and curatives using the traditional batch mixers in view of their high melting temperature and high loading level. The short mixing time, reduced mixing volume, and broad viscosity range of twin-screw extruder make it ideal for melt mixing thermoplastic with epoxy resins at elevated temperatures.

Continuous preparation of thermoplastic epoxies based on twin-screw extruders can be preceded with two approaches: (a) thermoplastic and epoxies are extruded at the melting temperature of thermoplastic without curatives followed by quickly mixing with curatives to ensure no cross-link reaction occurs and (b) thermoplastic, epoxies, and curatives are slurry mixed in one step at temperature below the cure temperature of the system. Both methods require carefully controlling the mixing temperature, mixing time, and curing kinetics to minimize the cure reaction during the continuous mixer process in order to avoid chemical cross-link reaction and prevent potential damage to the screw elements.

For the first approach, S. Izawa (Izawa and Toyama 1973) described a continuous process to prepare glass-fiber-reinforced polyphenylene ether (PPE)/epoxy resin blends using corotating intermeshing twin-screw extruder. Uncured epoxy resins were added in an amount of 45 wt. % or less of the total weight to improve the molding processability of PPE without losing the advantage of glass-fiber reinforcement. For example, PPE (90 parts) and solid epoxy resin (10 parts) having an average molecular weight of 8,000 g/mol (trademark: Dow D.E.R. 669) were mixed at barrel temperatures of 310 °C. The resinous composition was then mixed with glass fibers by being fed at the vent of extruder whereby the barrel temperature was adjusted to 330 °C. The produced glass-fiber-reinforced PPE/uncured epoxy blends were shaped by injection molding with no cure agents. F. Constantin et al. (2003) prepared poly (hydroxyl-amino ether) (PHAE)/diglycidyl ether of bisphenol A (DGEBA) blends in a corotating twin-screw extruder. Thermoplastic PHAE was grounded into fine powder to increase surface contact area with DGEBA. The initial epoxy monomer and PHAE (up to 70 wt. %) were extruded using a classical thermoplastic/liquid twin screw profile, and the extrudates were quenched in water and dried. Subsequently a curing agent was added to the blends and mixed for 1 min at 135 °C using an internal mixer. In both examples, since curatives are either not used or added after mixing thermoplastic with epoxies, it is not a real sense of continuous mixing method for preparing epoxy thermosets/thermoplastic blends.

The continuous mixer method was further developed by E. A. van den Berg (Berg et al. 1997). He proposed a simple but promising one-step preparation of PPE/epoxy resin blends with curatives using a modified twin-screw extrusion device, as shown in Fig. 7. PPE and epoxies were added through the feed port of the extruder. The barrel temperatures and the screw design were modulated in such way that PPE and epoxy resin were melted and mixed in the upstream portion of the extruder. Then at the downstream barrels, the molten blends were further mixed with a suitable curative by adding the curative through a side stuffer. Mixing elements such as kneading blocks, turbine mixing elements, and gear mixing elements are preferably used to facilitate the homogenous mixing of curatives. Adding the curing agent in the downstream barrel zone of the extruder shortly before the melting blends exiting the extruder is important to minimize the cure reaction and prevent the cross-link. The blends comprising PPE, epoxy, and curatives were then pumped through a pelletizing die to a water bath and pelletizer, or through a sheet die to nip roll pairs for sheet/film extrusion, or through a sheet die, then combined with carbon fiber or glass fiber in nip roll pairs or double belt lamination device for fiber reinforced composite manufacture.
Fig. 7

One-step methods for preparing PPE/epoxy resin blends using twin-screw extruder: (1) main feeders, (2) side feeder, (3) rolls, (4) nip rolls

The second approach for preparing epoxy/thermoplastic blends using continuous mixer is to slurry mix the thermoplastic, epoxies, and curatives at a temperature below the cure temperature of the system. In this case, a miscible thermoplastic and curatives are mixed in liquid epoxies to form solid discontinuous phase with a particle size less than 20 μm (Repecka 1988). During the mixing process, only a very small amount of thermoplastic and curatives were dissolved without causing significant changes in the matrix viscosity. Those well-dispersed thermoplastic and curative particles can further dissolve in the liquid epoxy monomers at elevated temperatures and undergo reaction-induced phase separation to form the final blends. Thermoplastics such as the polyimides, polyetherimides (PEI), polysulfones (PSF), polyethersulfones (PES), and polyetheretherketones (PEEK) with glass transition temperatures greater than 150 °C are ideal candidates for this method due to their excellent compatibility with epoxies. This method is particularly useful for prepregs and film adhesive applications where tack and drape are the primary concerns of resin properties at their intended operating temperature.

The University of Southern Mississippi (Mississippi, USA) recently developed a method to prepare heat-curable PEEK-reinforced epoxy blends using twin-screw extruder (Kingsley et al. 2012). ZSK 26 MC Compounder (L/D = 40) equipped with a liquid resin feeding system, a solid curative feeder, and a PEEK feeder was used as shown in Fig. 8. Epoxies, curatives, and PEEK were slurry mixed at the barrel temperatures below 100 °C to ensure no cure reaction occurs during continuous mixing. The slurry products are solid-like materials at room temperature and can be compression molded and cured into the final shape at elevated temperature.
Fig. 8

Twin-screw extruder preparation of heat-curable PEEK-reinforced epoxy blends using slurry mixing

Three-Roller Mills

Three-roller mills, commonly known as calendaring mills, are another ideal continuous mixer for blending high viscous materials such as epoxies. It employs both shear flow and extensional flow created by rotating rollers with different speeds to mix, disperse, and homogenize viscous matrices with additives or particles. In addition, it is a solvent-free process which is environmentally favorable. Theoretically three-roller mills are capable of handling materials with viscosity ranging from 200 to 2,000,000 centipoises. It can be easily combined with other resin processes such as filming to create a continuous production.

The basic design of three-roller mills consists of three counterrotating cylindrical rollers that are adjacent to each other with a circulating oil temperature control system (Fig. 9). The first and third roller, called the feeding and apron rollers, rotate in the same direction while the center roller rotates in the opposite direction. The three-roller mills offer different rotating speeds which are varied by drive units. The speed ratio for the three rollers (known as friction ratio) is usually about 1:3:9.
Fig. 9

(a) EXAKT three-roll mills and (b) corresponding schematic showing the general configuration and its working mechanism (Courtesy: EXAKT Technologies, Inc.)

In a typical mixing cycle, the materials to be mixed are either pre-blended or fed directly into the hopper. After being pre-dispersed in the first gap between the feeding roller and central roller, materials are transported into the second gap by the central roller. In the second gap, the material is subjected to even higher shear force due to the higher speed of apron roller and dispersed into desired dispersing level. The resulted materials are then discharged and collected for further processing. This milling cycle can be repeated several times until uniform mixing is achieved. Friction, speed ratio, sizes, and gap of the rollers impact the dispersion and intensity of mixing.

Application of three-roller mills has been primarily focused on dispersing inorganic particles in epoxy matrices. It has been proved to be a highly efficient dispersion method for nanoparticles containing fiber-like fillers such as glass fiber, carbon nanotubes, carbon fiber, etc. (Chang et al. 2009; Gojny et al. 2004). Because of the strong shear force between the three rollers, particles can achieve a high degree of intercalation/exfoliation within a short period of time. Another major advantage of using this method is the gap between rollers can be mechanically adjusted within micro length so that the particle sizes are controllable to obtain a narrow size distribution of nanoparticles.

Literatures about mixing epoxies with thermoplastic using three-roller mills are in fact limited. One of the possible reasons is that the three-roll millers that are currently being used are designed with limited heating capability (up to 60 °C for EXAKT). The feeding materials should be in viscous state when mixing with other materials. Thus this mixing technique may not be applied for mixing thermoplastic which typically has much higher melting temperature. However, owing to the broad viscosity processing window of three-roll mixers, it can be used for dispersing curative into the pre-dissolved epoxies/thermoplastic mixtures without the need of elevated temperature. Choe et al. (2003) prepared polyamide copolymer and polyetherimide-modified DGEBA blends by mixing a stoichiometrically balanced amount of curing agent with pre-dissolved DGEBA/thermoplastic mixture in a three-roll mill. Curable mixtures were obtained after mixing for 1 h at room temperature. This approach is very promising for room temperature curing epoxy systems for which elevated temperature mixing is not practical.

Continuous Polymerization Reactor

The term “continuous polymerization reactor” is used to describe a polymer process that involves chemical reactions. It is a continuous, flexible process offering the technical and economic advantages compared to other reactors such as batch reactors. Corotating twin-screw extruder is an excellent chemical reactor and can be used to efficiently control polymerization reactions. Through simultaneously applying kneading, shearing, pressuring, and heating, continuous reaction results in highly homogenous mixing products with low reaction duration and high consistency, which are the major manufacturing cost-saving factors. Examples for twin-screw extruders being used as chemical reactors include grafting reactions (Cai et al. 2008), post-reaction processes including bulk polymer reactions (Finnigan et al. 2004), reactive compatibilization of particulates and reinforcements (Shokoohi et al. 2011), and reactive blending (Oyama 2009).

While batch reactors are generally based upon relatively low-temperature reactions over long periods of time, continuous reactors are designed for higher-temperature reactions over short periods of time. Reaction time of continuous reactor is described as residence time. It is controlled through screw configuration, screw speed, feeding rate, and reactor length and typically in the range of 1–10 min. Since residence time is short, reaction quantities are small and heat-transfer efficiencies are high; continuous reactors based on twin-screw extruders are preferred to conduct at elevated temperature. Catalyst and intensive mixing, as well as pressure, are often used to accelerate the reaction rates.

Within numerous applications for polymer reactions, continuous reactors have been mainly used for high-molecular-weight thermoplastic matrices. They are nontraditional for conducting cure reaction when preparing cured glassy epoxies. As the gelation occurs and the network forms during cure, the system transits from liquid to gel and to vitrified states which will result in conveying issues in continuous reactor. This is the major reason that curatives are added after mixing thermoplastics and epoxies with reduced mixing time or slurry mixed with epoxies in twin-screw extruders as discussed in the previous section.

For certain applications such as aerospace prepreg, partially reacted epoxy oligomers, also called B-stage epoxy prepolymers, are required for the purpose of controlled flow and self-adhesion. This brings the possibility of using continuous reactor to prepare partially cured epoxy prepolymers. Traditionally the epoxy prepolymers are prepared using liquid batch mixers. Epoxies are firstly mixed with additives such as thermoplastic tougheners and curatives at elevated temperature. After mixed, cure reaction continues at appropriate temperature over long periods of time to advance to a prescribed molecular weight with desired viscosity. The produced epoxy prepolymers are discharged from the batch mixers and stored at low temperature for indefinite periods of time to retard cure and preserve viscosity.

Attempts to conduct epoxy prepolymer reaction in continuous reactors are rare. Previous research reported by Titier (Titier et al. 1995, 1996) described a continuous reactor method using corotating twin-screw extruder to synthesize epoxy-amine multiacrylate prepolymers. The system is based on a competitive reaction of a difunctional amine with a difunctional epoxy, DGEBA, acting as a chain extender and a mono-epoxide acrylate, acting as a chain termination agent, avoiding the gelation of the system and introducing double bonds in the epoxy oligomer formed. The formed epoxy-amine multiacrylate prepolymers can further lead to cross-linked networks via double-bond polymerizations.

Very recently, X. Cheng (Cheng and Wiggins 2014) designed a continuous reactor to prepare thermoplastic-modified epoxy prepolymer blends with a thermoplastic content of up to 20 wt. % of total mass weight. Thermo Prism 16 mm corotating twin-screw extruder (L/D = 25) was used in his study. The reactor consists of a feed zone, five electrically heated and liquid-cooled zones, and an electrically heated die zone. The screw configuration shown in Fig. 10 was designed to balance shear mixing and residence time with a combination of various conveying, kneading, and reversing elements. Thermoplastic PES and a stoichiometric amount of 44DDS was fully dissolved in TGDDM after the continuous reaction process at the barrel temperatures of 180–220 °C. The produced prepolymers were optically transparent. In addition, concurrent epoxy-amine chain extension reaction advanced the TGDDM prepolymer to targeted molecular weights and viscosities. The level of cure was kinetically controlled by adjusting barrel temperatures and residence times of the continuous reactor. Gelation and cross-link were eliminated through a systematic analysis of epoxy cure reaction kinetics, processing conditions, and continuous reactor design. The produced epoxy prepolymers can be directly used for prepreg filming operations.
Fig. 10

Screw configuration of the continuous polymerization reactor for epoxy/thermoplastic prepolymer blends (Reproduced with permission from John Wiley and Sons)

Compared to the batch reactor, the continuous reactor efficiency in mixing and dissolution is attributed to the high shear, excellent heat transfer, and small reaction volume. Although temperature can be increased to accelerate thermoplastic mixing in a batch reactor, it is impractical to control curative dissolution and reaction conversion at high temperatures considering the large volume of batch reactor. Small reaction volumes and excellent heat transfer in continuous reactors allow to prepare epoxy prepolymers at elevated temperature within reduced reaction time. But the challenge for avoiding gelation remains a critical problem. It was found that 220 °C was the maximum processing temperature for TGDDM/44DDS system using continuous reactor method.

The advancement of continuous reactors in this field is very promising and will lead to new avenues for mixing and blending a broad array of co-reactants, thermoplastics, nanoparticles with epoxy matrices due to better control over rheology, shear states, and reactor design advantages.

Nonmechanical Methods

There are many nonmechanical methods used for material mixing while not all of them are designed for epoxy/thermoplastic blending. One of the examples is the sonication which is known for effective nanoparticle dispersion. It applies ultrasound energy through ultrasonic batch or ultrasonic probe to agitate particles in solution or polymer melt and result in the separation of individualized nanoparticles from the bundles. However, it is too powerful to blend epoxy and thermoplastic due to the degradation and therefore is out of the scope in this chapter.

Solvent Casting

Solvent casting has been the main technique to fabricate thin-layered polymer films since the early nineteenth century. It consists of dissolution of the ingredients in a volatile solvent and then evaporation of the solvent through suitable drying devices. Compared to melt mixing process, solvent casting exhibits higher mixing quality, thinner film, lack of residual stresses, and higher purity and clarity (Ricklin 1983). In addition, the solution or polymer film is exposed to relatively low thermal or mechanical stress throughout the entire production process. As a result, degradation or side reactions are insignificant.

There are numerous literatures concerned with the mixing of epoxy and thermoplastics using solvent-casting techniques (Bucknall et al. 1994; Mustata and Bicu 2006; Hourston et al. 1997). In a typical process, the epoxy resins and curatives are dissolved in a solvent such as dichloromethane with low boiling temperature of 39.6 °C. Thermoplastics are then added to the resin/curative solution under stirring. After homogenously mixed, the solution is heated to the boiling temperature of solvent to remove most of the solvent and then poured into a preheated mold set at a temperature with a reasonable viscosity for casting. The residual solvent is further removed by degassing under vacuum. The solvent-free mixture is fully cured into final products at elevated temperature. The unique advantage of solvent casting for preparing epoxy/thermoplastic blends is to produce an epoxy solution with greatly reduced viscosity which aids thermoplastic mixing and improves the resin processability.

Although adding the thermoplastic to epoxy resins is made easier by the use of a solvent, it does cause problems. The evaporation and degassing of the solvent prior to the use of the resin is time consuming. Furthermore, solvent is hard to be fully removed from the mixture. Any residual solvent acts as a plasticizer for the polymer and has noticeable effects on the modulus of the final resin matrix. Last but not least, extensive use in solvent makes the process less commercially attractive and less environmentally friendly. As a consequence, solvent-casting method is often limited in laboratory-scale experiments instead of industrial scale manufacturing.

Resonant Acoustic Mixer

Resonant acoustic mixer (RAM) is a relatively new non-contact mixing approach to disperse and distribute materials within a broad range of viscosities. It relies upon the application of low-frequency and high-intensity acoustic field throughout the entire mixing vessel to facilitate fast mixing. This differs from traditional mixers such as batch mixers or continuous mixers where the mixing only occurs at the localized region near the tips of mixing blades. Figure 11 shows the laboratory-scale resonant acoustic mixer and its schematic mixing mechanism. The mixer device includes three major components: the mixing vessel, the plate, and the spring-mass system which is composed of multiple springs and masses. During its operation, the springs and masses move simultaneously by applying an external force and initiate oscillation of the system. At a particular oscillating frequency, the inertia force of the masses is offset by the stored force in the springs, allowing transferring of the mechanical energy created by the mixer to the materials in the vessel via the propagation of acoustic pressure wave. This in turn causes the micro-mixing cells throughout the vessel with a mixing cell length around 50 μm as indicated in Fig. 11. The particular oscillating frequency is adjusted according to the properties of the materials to be mixed and normally set at 60 Hz. The only controllable parameter of the RAM mixer is the mixing intensity which determines the amplitude of the mechanical oscillation.
Fig. 11

Laboratory-scale resonant acoustic mixer and schematic showing the resonant acoustic mixing process (Courtesy: Resodyn Acoustic Mixers, Inc.)

There has been no peer-reviewed publications about the application of RAM technology. According to the limited information provided by the Resodyn Acoustic Mixers, Inc., RAM mixer is capable of mixing various types of materials, which include liquid-liquid, liquid-solid, gas-liquid, and solid-solid systems with viscosities ranging from 100 to 100,000,000 centipoises. The temperature increase during mixing caused by the acoustic resonance is insignificant. Vacuum can be applied to eliminate the bubbles being trapped during mixing. RAM mixer also tends to be cost and time saving through the substantial reduction in mixing times and the elimination of time and costs associated with cleanup. All these characteristics of RAM mixer are of extreme interest for a broad range of industries that involves material mixing, in particular for mixing epoxy/thermoplastic blends. It has the potential to be used for scale-up manufacturing. More and more attention is expected to be devoted to this interesting technology.

In Situ Polymerization

Approaches for preparing epoxy/thermoplastic blends described so far have been primarily focused on homogeneously dissolving thermoplastic polymers into epoxy resins with or without the aid of solvent. This is a semi-interpenetrating polymer network (semi-IPN) technique that blends two polymers by entangling the molecular chain of thermoplastic polymers to an epoxy network without chemical bonding between them. The formation of the semi-IPN structure allows to homogeneously blend the epoxy matrix and thermoplastic polymer in a molecular level or to microscopically disperse the thermoplastic in epoxy matrix. However, all of those approaches face the same problem: the increased viscosity caused by incorporation of thermoplastic reduces epoxy resin processing window, especially when epoxies are used for applications such as wet-lay-up laminating process and resin infusion process. The resin viscosities for those processes need to be sufficiently low to provide controlled resin flow and self-adhesion and is typically in the range between 1,000 and 10,000 centipoises.

Recently, attempt was made to address this problem by using in situ polymerization technique. The in situ polymerization method is very similar to the simultaneous interoperating polymer network technique (SIPN) except one of the polymer architectures is not a cross-linking network. The thermoplastic monomers, polymerization activators, and curatives are firstly mixed with the epoxy monomers. The polymerization reactions of thermoplastics and cure reactions of epoxies are carried out simultaneously but by noninterfering reactions. Thermoplastic polymer is formed during the cure process of the epoxy resin by radical polymerization. K. Mimura (Mimura et al. 2001) prepared epoxy/thermoplastic blends using the in situ polymerization method. Thermoplastic polymer was prepared by homogenously dissolved N-phenylmaleimide (PMI), benzyl methacrylate (BzMA), and styrene (St) monomers in DGEBA resin and phenol curatives at their cure temperature of 100 °C. As soon as the accelerator 1-isobutyl-2-methylimidazol (IBMI) and initiator 2, 5-dimethyl-2, 5-bis (benzoyl peroxy) hexane were added and dissolved, the mixture was rapidly poured into glass molds and fully cured in the oven. During cure process, thermoplastic polymers were quickly formed through radical polymerization of PMI/BzMA/St monomers and separated from epoxy matrix via reaction-induced phase separation.

Since the thermoplastic was formed during the cure process of epoxy resin, it avoids any increase in viscosity at the time of mixing and molding. Figure 12 shows the viscosities of the neat resin composed of epoxy and phenol curative, the resin containing PMI/BzMA/St monomers, and the resin modified with a thermoplastic polymer which had been previously polymerized with approximately the same molecular weight as the one obtained by the in situ polymerization. The viscosity of the neat resin composed only of the epoxy and the curatives was 550 centipoises at 100 °C. When thermoplastic polymer was added to the resin using conventional mixing method, the viscosity became very high and was about 28,000 centipoises. In contrast, when the monomers were added, the viscosity of resin containing these monomers was only 65 centipoises at 100 °C, decreasing to about one-eighth of the neat resin and several orders of magnitude lower than the one containing thermoplastic polymer prepared by conventional technique.
Fig. 12

Effect of addition of vinyl monomers or thermoplastic polymer on the viscosity of epoxy resins before curing. ○ unmodified resin. □ resin containing vinyl monomers (PMI/BzMA/St = 5/5/3 in molar ratio; content: 23.6 wt. %); Δ resin modified with the thermoplastic polymer 23.6 wt. % (Reproduced with permission from Elsevier)

In situ polymerization method, however, has a certain limitation. During thermoplastic polymerization, the growth of thermoplastic molecular weights can be hindered by the increased viscosity of epoxy matrix upon cure and completely suppressed when system is vitrified. As a result fast-cured epoxy systems are impractical to use considering its short gelation and vitrification time.

In Situ Dissolution

In order to overcome viscosity increase in the traditional thermoplastic-toughening strategy used for prepreg-based composite, a novel in situ dissolution method is recently proposed by M. Naffakh et al. (2006). In this method, polyetherimide (PEI) thin films were firstly prepared using solvent casting. Alternative layers of glass fibers and thermoplastic PEI films were then put into a mold. The epoxy reactive mixture was injected into the heated mold using resin transfer molding process (RTM) to form the prepreg structure.

During the cure process, several physic-chemical events take place in a sequential way including epoxy impregnation, in situ dissolution of the thermoplastic film, epoxy cure reaction, and subsequent phase separation as schematically illustrated in Fig. 13. The important factors allowing these event sequences are the cure reaction kinetics of epoxy, initial solubility of thermoplastics, evolution of rheology, and the onset of reaction-induced phase separation. The PEI thermoplastic films need to be readily dissolved by the reactive resin at the early stage of cure where no phase separation occurs. Studies of dissolution time versus film thickness have proved that thermoplastic film with a thickness below 20 μm is necessary to ensure thermoplastic layers readily dissolving into the resin matrices (Cicala et al. 2009). Subsequently, the reaction-induced phase separation is initiated by epoxy cure reaction and yields an epoxy/thermoplastic thin layer in the interply regions with desired phase-separated morphology to prevent crack prorogation. The formed epoxy/thermoplastic thin layer, so-called interleaf, can lead to similar morphology and mechanical properties compared to the system prepared via the traditional method. In addition, the viscosity increase caused by adding thermoplastic is appropriately addressed without changing any of the processing aspects of the RTM method.
Fig. 13

RTM processing based on the in situ dissolution of epoxy/PEI blends: (a) insertion of a PEI film in between fiber plies, (b) dissolution of the PEI film and polymerization processes, and (c) generation of a PEI-dispersed phase from a reaction-induced phase separation and final curing of the epoxy matrix (Reproduced with permission from Elsevier)


In this chapter, an overview about the techniques used for preparation of epoxy/thermoplastic blends was provided. As discussed, the majority of epoxy/thermoplastics blends are mechanically mixed in batch mixers and continuous mixers. The selection of batch or continuous mixers depends on material physical states, mixing quality and efficiency, handling equipment, energy consumption, and labor cost.

It is understood that mixing thermoplastic with epoxy resins inevitably increased resin viscosity. Liquid low-shear batch mixers are preferred to use when preparing epoxy/thermoplastic blends with a relatively low thermoplastic content. For the blends with a high amount of thermoplastic, especially when thermoplastics form a continuous matrix phase, highly viscous mixers such as internal mixers or twin-screw extruders are often used. In addition, compared to batch mixers, continuous mixers based on twin-screw extruders offer more modularity to handle different types of components and chemistries. Because of this unique advantage, they can be used as continuous reactors to produce curable epoxy/thermoplastic blends in a one-step approach.

Nonmechanical methods are also an important routine to prepare epoxy/thermoplastic blends. Solvent casting is a simple and effective method for mixing thermoplastic with epoxy but is often limited in laboratory-scale experiments due to cost, safety, and environment reasons. Some other novel techniques such as RAM technology, in situ polymerization, and in situ dissolution are very attractive since the processing issues caused by adding thermoplastic are appropriate addressed without significantly changing the manufacturing methods. However, it must be stressed that each method mentioned here has a strict requirement for chemistry and compatibility of selected epoxy and thermoplastic.


  1. Arias ML, Frontini PM, Williams RJJ (2003) Analysis of the damage zone around the crack tip for two rubber-modified epoxy matrices exhibiting different toughenability. Polymer 44:1537–1546CrossRefGoogle Scholar
  2. Bennett GS, Farris RJ, Thompson SA (1991) Amine-terminated poly(aryl ether ketone)-epoxy/amine resin systems as tough high performance materials. Polymer 32:1633–1641CrossRefGoogle Scholar
  3. Berg EA et al (1997) Melt-mixing thermoplastic and epoxy resin above Tg or Tm of thermoplastic with curing agent. US 6,197,898Google Scholar
  4. Blanco I, Cicala G, Costa M, Recca AJ (2006) Development of an epoxy system characterized by low water absorption and high thermomechanical performances. Appl Polym Sci 100:4880–4887CrossRefGoogle Scholar
  5. Bouquey M, Loux C, Muller R, Bouchet G (2011) Morphological study of two-phase polymer blends during compounding in a novel compounder on the basis of elongational flows. J Appl Polym Sci 119:482–490CrossRefGoogle Scholar
  6. Brown SB, Orlando CM (1988) Encyclopedia of polymer science and engineering. Wiley, New YorkGoogle Scholar
  7. Bucknall CB, Partridge IK (1983a) Addition of polyethersulfone to epoxy resins. Br Polym J 15:71–75CrossRefGoogle Scholar
  8. Bucknall CB, Partridge IK (1983b) Phase separation in epoxy resins containing polyethersulphone. Polymer 24:639–644CrossRefGoogle Scholar
  9. Bucknall CB, Gomez CM, Quintard I (1994) Phase separation from solutions of poly (ether sulphone) in epoxy resins. Polymer 35:353–359CrossRefGoogle Scholar
  10. Butta E, Levita G, Marchetti A, Lazeri A (1986) Morphology and mechanical properties of amine-terminated butadiene-acrylonitrile/epoxy blends. Polym Eng Sci 26:63–73CrossRefGoogle Scholar
  11. Cai C, Shi Q, Li L, Zhu L, Yin J (2008) Grafting acrylic acid onto polypropylene by reactive extrusion with pre-irradiated PP as initiator. Radiat Phys Chem 77:370–372CrossRefGoogle Scholar
  12. Chandran CS, Muller R, Bouquey M, Serra C, Thomas S (2015) Effect of blend ratio and elongation on the morphology and properties of epoxy resin- poly(trimethylene terephthalate) blends. Polym Eng Sci 55:1679–1688CrossRefGoogle Scholar
  13. Chang L, Friedrich K, Ye L, Toro P (2009) Evaluation and visualization of the percolating networks in multi-wall carbon nanotube/epoxy composites. J Mater Sci 44:4003–4012CrossRefGoogle Scholar
  14. Cheng X, Wiggins JS (2014) Continuous reactor preparation of thermoplastic modified epoxy-amine prepolymers. Polym Int 63:1777–1784CrossRefGoogle Scholar
  15. Choe Y, Kim M, Kim W (2003) In situ detection of the onset of phase separation and gelation in epoxy/anhydride/thermoplastic blends. Macromol Res 11:267–272CrossRefGoogle Scholar
  16. Cicala G, Recca G, Ziegmann G, Niemeyer S (2009) Development of thermoplastic interleaf films which build up a controlled morphology for toughening of rtm resins. In: 17th international conference on composite materials, Edinburgh, pp 51–57Google Scholar
  17. Constantin F, Fenouillot F, Guillaume JL, Koeing R, Pascault JP (2003) Blends of a new thermoplastic in a thermoset epoxy matrix. Macromol Symp 198:335–344CrossRefGoogle Scholar
  18. Di Liello V, Martuscelli E, Musto P, Ragosta G, Scarinzi GJ (1994) Toughening of highly crosslinked epoxy resins by reactive blending with bisphenol A polycarbonate. II. Yield and fracture behavior. J Appl Polym Sci 32:409–419CrossRefGoogle Scholar
  19. Finnigan B, Martin D, Halley P, Truss R, Campbell K (2004) Morphology and properties of thermoplastic polyurethane nanocomposites incorporating hydrophilic layered silicates. Polymer 45:2249–2260CrossRefGoogle Scholar
  20. Girard-Reydet E, Vicard V, Pascault JP, Sautereau HJ (1997) Polyetherimide-modified epoxy networks: influence of cure conditions on morphology and mechanical properties. J Appl Polym Sci 65:2433–2445CrossRefGoogle Scholar
  21. Gojny FH, Wichmann MHG, Köpke U, Fiedler B, Schulte K (2004) Carbon nanotube-reinforced epoxy-composites: enhanced stiffness and fracture toughness at low nanotube content. Compos Sci Technol 64:2363–2371CrossRefGoogle Scholar
  22. Hourston DJ, Lane JM (1992) The toughening of epoxy resins with thermoplastics: 1. Trifunctional epoxy resin-polyetherimide blends. Polymer 33:1379–1383CrossRefGoogle Scholar
  23. Hourston DJ, Lane JM, Zhang HX (1997) Toughening of epoxy resins with thermoplastics: 3. An investigation into the effects of composition on the properties of epoxy resin blends. Polym Int 42:349–355CrossRefGoogle Scholar
  24. Inoue T (1995) Reaction-induced phase decomposition in polymer blends. Prog Polym Sci 20:119–153CrossRefGoogle Scholar
  25. Ishii Y, Ryan AJ (2000) Processing of poly(2,6-dimethyl-1,4-phenylene ether) with epoxy resin. 1. Reaction-induced phase separation. Macromolecules 33:158–166CrossRefGoogle Scholar
  26. Izawa S, Toyama K (1973) Glass fiber reinforced thermoplastic resinous composition. US 3,763,088Google Scholar
  27. Kim HC, Char KH (2000) Effect of phase separation on rheological properties during the isothermal curing of epoxy toughened with thermoplastic polymer. Ind Eng Chem Res 39:955–959CrossRefGoogle Scholar
  28. Kim JK, Robertson RE (1992) Toughening of thermoset polymers by rigid crystalline particles. J Mater Sci 27:161–174CrossRefGoogle Scholar
  29. Kim BS, Chiba T, Inoue T (1995) Morphology development via reaction-induced phase separation in epoxy/poly(ether sulfone) blends: morphology control using poly(ether sulfone) with functional end-groups. Polymer 1:43–47CrossRefGoogle Scholar
  30. Kingsley D, Cheng X, Yang J, Wiggins JS (2012) Continuous polymer reactor design for prepreg epoxy matrix prepolymers. In: International SAMPE technical conference, CharlestonGoogle Scholar
  31. Ma Z, Stanford JL, Dutta BK (2009) Reaction kinetics of epoxy resin modified with reactive and nonreactive thermoplastic copolymers. J Appl Polym Sci 112:2391–2400CrossRefGoogle Scholar
  32. May C (1988) Epoxy resins: chemistry and technology. Marcel Dekker, New YorkGoogle Scholar
  33. Mimura K, Ito H, Fujioka H (2000) Improvement of thermal and mechanical properties by control of morphologies in PES-modified epoxy resins. Polymer 41:4451–4459CrossRefGoogle Scholar
  34. Mimura K, Ito H, Fujioka H (2001) Toughening of epoxy resin modified with in-situ polymerized thermoplastic polymers. Polymer 42:9223–9233CrossRefGoogle Scholar
  35. Moad G (1999) The synthesis of polyolefin graft copolymers by reactive extrusion. Prog Polym Sci 24:81–142CrossRefGoogle Scholar
  36. Moribe T (2012) Advanced intermeshing mixers for energy saving and reduction of environmental impact. Mitsubishi Heavy Ind Technol Rev 49:38–43Google Scholar
  37. Mustata F, Bicu I (2006) Synthesis, characterization, and properties of multifunctional epoxy maleimide resins. Macromol Mater Eng 29:732–741CrossRefGoogle Scholar
  38. Naffakh M, Dumon M, Gerard JF (2006) Study of a reactive epoxy–amine resin enabling in-situ dissolution of thermoplastic films during resin transfer moulding for toughening composites. Compos Sci Technol 66:1376–1384CrossRefGoogle Scholar
  39. Oyama HT (2009) Super-tough poly(lactic acid) materials: reactive blending with ethylene copolymer. Polymer 50:747–751CrossRefGoogle Scholar
  40. Park SJ, Kim HC (2001) Thermal stability and toughening of epoxy resin with polysulfone resin. J Polym Sci B Polym Phys 39:121–128CrossRefGoogle Scholar
  41. Rauwendaal C (1981) Analysis and experimental evaluation of twin screw extruders. Polym Eng Sci 21:1092–1100CrossRefGoogle Scholar
  42. Rauwendaal C (1998) Polymer mixing, a self-study guide. Carl Hanser Verlag, MunichGoogle Scholar
  43. Repecka LN (1988) Slurry-mixed heat-curable resin systems having superior tack and drape. US 5,747,615Google Scholar
  44. Ricklin S (1983) Solvent cast films. Plast Eng 39:29–33Google Scholar
  45. Salahudeen SA, Elleithy RH, Alothman O, Alzahrani SM (2011) Comparative study of internal batch mixer such as cam, banbury and roller: numerical simulation and experimental verification. Chem Eng Sci 66:2502–2511CrossRefGoogle Scholar
  46. Shokoohi S, Arefazer A, Naderi G (2011) Compatibilized polypropylene/ethylene–propylene–diene-monomer/polyamide6 ternary blends: effect of twin screw extruder processing parameters. Mater Des 32:1697–1703CrossRefGoogle Scholar
  47. Sultan JN, McGarry FJ (1973) Effect of rubber particle size on deformation mechanisms in glassy epoxy. Polym Eng Sci 13:29–34CrossRefGoogle Scholar
  48. Tadmor Z, Gogos CG (2006) Principles of polymer processing. Wiley, New YorkGoogle Scholar
  49. Titier C, Pascault JP, Taha M, Rozenberg BJ (1995) Epoxy-amine multimethacrylic prepolymers, kinetic and structural studies. J Polym Sci A 33:175–184CrossRefGoogle Scholar
  50. Titier C, Pascault JP, Taha M (1996) Synthesis of epoxy–amine multiacrylic prepolymers by reactive extrusion. J Polym Sci A 59:415–422CrossRefGoogle Scholar
  51. Utracki LA, Wilkie C (2002) Polymer blends handbook. Springer, New YorkGoogle Scholar
  52. White JL, Bumm SH (2011) Encyclopedia of polymer blends, volume 2, processing. Wiley, New YorkGoogle Scholar
  53. Williams RJJ, Rozenberg BA, Pascault JP (1997) Reaction-induced phase separation in modified thermosetting polymers. Adv Polym Sci 128:95–156CrossRefGoogle Scholar
  54. Xanthos M (1992) Reactive extrusion: principles and practice. Hanser, MunichGoogle Scholar
  55. Yu Y, Wang M, Foix D, Li S (2008) Rheological study of epoxy systems blended with poly(ether sulfone) of different molecular weights. Ind Eng Chem Res 47:9361–9369CrossRefGoogle Scholar
  56. Zhang J, Guo Q, Fox BL (2009) Study on thermoplastic-modified multifunctional epoxies: influence of heating rate on cure behavior and phase separation. Compos Sci Technol 69:1172–1179CrossRefGoogle Scholar
  57. Zhang J, Yang T, Lin T, Wang CH (2012) Phase morphology of nanofibre interlayers: critical factor for toughening carbon/epoxy composites. Compos Sci Technol 72:256–262CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  1. 1.School of Polymers and High Performance Materials, Polymer Science and EngineeringUniversity of Southern MississippiHattiesburgUSA

Personalised recommendations