SEARCH

SEARCH BY CITATION

Keywords:

  • block copolymers;
  • micelles;
  • thermosets;
  • morphology

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY
  7. Acknowledgements
  8. REFERENCES AND NOTES

It has been found that by the addition of low concentrations of an amphiphilic block copolymer to an epoxy resin, novel disordered morphologies can be formed and preserved through curing. This article will focus on characterizing the influence of the block copolymer and casting solvent on the templated morphology achieved in the thermoset sample. The ultimate goal of this work is to determine the parameters that would control the microphase morphology produced. Epoxy resins blended with a series of amphiphilic block copolymers based on hydrogenated polyisoprene (polyethylene-alt-propylene or PEP) and polyethylene oxide (PEO), specifically, were investigated. In this article, the cure-induced order–order phase transition from the spherical to wormlike micelle morphology will also be discussed. It is proposed that the formation of the wormlike micelle structure from the spherical micelle structure is similar to the phase transition behavior that occurs in dilute block copolymer solutions as a function of the influence of the solvent on micelle morphology. © 2007 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 45: 3338–3348, 2007


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY
  7. Acknowledgements
  8. REFERENCES AND NOTES

Thermoset polymers have found use in a variety of applications ranging from adhesives to bulk structural components, with extensive use in the aerospace, electronics, dental, and other medical fields. Epoxy resins are usually two component thermosets that chemically react to form a highly crosslinked network that is stiff, chemically and environmentally resistant, and thermally stable in high temperature applications. These benefits are balanced by the fact that crosslinked resins are typically brittle, and possess relatively poor fracture toughness with minimal ductility. To increase the range of commercial applications available to thermoset materials, methods for enhancing the fracture toughness with little or no loss of the desired stiffness and thermal stability must be developed.

A common toughening mechanism utilized for epoxy resins is the addition of phase separating additives. The phase separating media can take the form of preformed particulates or a dissolved polymeric phase that undergoes reaction induced phase separation during the epoxy curing process. The addition of rubber particles or other preformed thermoplastic polymers has been shown to provide effective toughening of epoxy resin materials.1–4 However, the addition of rubber particles to the epoxy resin can give rise to processing difficulties and the glass transition temperature, Tg, as well as the modulus of the thermoset sample can drop significantly.4, 5 The size and distribution of the rubber particles in the epoxy resin matrix are difficult to control, which in turn influence the degree of toughening enhancement.6–10 The same difficulties in processing and controlling the size and distribution of the additive are also encountered when other immiscible or preformed toughening agents are used.1, 11–13

In an effort to overcome the limitations in the use of rubber particles and other preformed additives as toughening agents, the addition of small amounts of block copolymers that contain both an epoxy miscible and an epoxy immiscible block segment have been investigated.10, 14, 15 It has been found that by the addition of low concentrations of these types of block copolymers, novel disordered morphologies can be formed. With the use of a block copolymer as the toughening agent, the size of the resulting features can be controlled and are typically on the order of tens of nanometers. Recent work at The Dow Chemical Company and the University of Minnesota has investigated the addition of small amounts of block copolymer to improve the fracture resistance of an epoxy resin without sacrificing the desirable high modulus and Tg of the neat epoxy.4, 14, 16–20 Discrete morphologies, such as spherical micelles, wormlike micelles, and vesicles were observed in the thermoset samples investigated.4, 17, 21

An understanding of how the molecular structural parameters of the block copolymer additive such as the chemical composition of the two blocks, the ratio of the two blocks, and the molecular weight of the polymer influence the morphology developed in the epoxy resin is not yet fully realized. Additionally, the effect of the casting solvent on the resultant block copolymer microphase morphology is not clearly understood. The work reported here focuses on epoxy resins blended with a series of amphiphilic block copolymers based on hydrogenated polyisoprene (polyethylene-alt-propylene or PEP) and polyethylene oxide (PEO). This article will focus on characterizing the influence of the block copolymer molecular structure and casting solvent on the microphase morphology achieved in the thermoset sample, supporting the ultimate goal of determining the parameters which produce the microphase morphology offering the greatest enhancement of fracture toughness. The dispersed phase microstructure developed by the block copolymer in the epoxy resin has been found to be an exceptional thermoset toughening agent.14, 16, 17, 22 Although all of the well-defined microstructures were observed to improve toughness without compromising modulus, materials containing spherical and wormlike micelles demonstrated truly exceptional toughening.22

A transition from the spherical micelle structure to the wormlike micelle structure, as a consequence of the curing process, has been observed and will be discussed in this article. The self-assembled block copolymer morphology that forms in the uncured epoxy matrix is generally retained into the fully crosslinked, cured material. However, because the PEO block is expelled from the curing resin as the level of cure increases, order–order phase transitions can occur. The cause of this order–order phase transition, as well as the mechanism by which the wormlike micelles are formed in the vitrifying epoxy matrix will be addressed.

EXPERIMENTAL

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY
  7. Acknowledgements
  8. REFERENCES AND NOTES

In this study, blends of 75 wt % The Dow Chemical Company bisphenol A based epoxy [(DER 383), Trademark of Dow Chemical Company; DER stands for Dow Epoxy Resin)] with 25 wt % of a brominated bisphenol A epoxy [(DER 560), Trademark of Dow Chemical Company; DER stands for Dow Epoxy Resin)] were investigated. Brominated epoxy resins offer enhanced fire retardation but impart increased brittleness to the sample. These blends were prepared using an equivalent amount of Durite SD1731 (Trademark of Borden Chemical) phenol novolac curing agent. One weight percent of 2-ethyl-4-methyl imidazole (2E4MI) was added as a catalyst. With the exception of one block copolymer purchased commercially, all of the block copolymers examined in this work were made in-house at The Dow Chemical Company. In all of the toughened thermoset samples, 5 wt % block copolymer was added, the weight percentage value is based on the solvent free samples. For the samples containing poly(ethylene oxide)-block-poly(ethylene-alt-propylene) (PEO-PEP), the PEP is prepared by the catalytic hydrogenation of polyisoprene (PI). The chemical structures of the PEO-PEP diblock copolymer, bisphenol A type epoxies, and the phenol curing agent are shown in Figure 1. PEP-PI and a commercial polyethylene-block-poly(ethylene oxide) (PE-PEO) block copolymer, purchased from the Aldrich Chemical Co., were also investigated as potential toughening agents for this epoxy resin blend at the same concentrations. The PE-PEO sample utilized in this study had a Mn = 1400 with 50 wt % PEO and a PE melt transition temperature of 115 °C. Full morphological characterization of the purchased PE-PEO block copolymer was not performed. The PE-PEO block copolymer was unique in this study as it was the only one which contained a semicrystalline block. The use of a semicrystalline component in a block copolymer templated epoxy has not been widely investigated.23

thumbnail image

Figure 1. Chemical structure of the components in the epoxy systems examined in this investigation: DER 383, DER 560, and the phenol novolac curing agent. The structure of PEO-PEP diblock copolymer is also shown.17

Download figure to PowerPoint

The PEO-PEP copolymers were prepared in three steps. First, isoprene was polymerized anionically and terminated with ethylene oxide to afford a hydroxylated polyisoprene (PI-OH) block. Next, hydrogenation of the PI-OH block to make hydroxylated PEP was performed. Finally, ethylene oxide was polymerized from the hydroxyl end of the PEP block to form the desired PEO-PEP block copolymer. The molecular weight of the PEP block was determined from the base PI by gel permeation chromatography (GPC) with respect to narrow polydispersity PI standards. The weight fraction of PEO was determined by 1H nuclear magnetic resonance (NMR) spectroscopy and the total molecular weights were confirmed by GPC. A table containing compositional information for the block copolymers utilized in this study is presented in Table 1. The block copolymers used in this study are designated by w-x-y-z, where w-x represents the block copolymer components, y is the total molecular weight (in kilograms), and z is the weight percent PEO. Characterization of the morphology of the neat block copolymers was performed by small-angle X-ray scattering (SAXS). In situ temperature ramps were performed between 20 and 200 °C, with data collection in 10 °C increments.

Table 1. Block Copolymer Composition, Molecular Weight, and SAXS Based Morphological Information at 20 °C
CopolymerPEO wt %Mn totald-Spacing (Å)Morphology
PEO-PEP-10k-333310.4k413Cylinder
PEO-PEP 4.6k-33334.6k125Cylinder
PEO-PEP 11k-424210.6k284Lamella
PEO-PI 11k-444411k242Lamella
PEO-PE 1.4k-50501.4kNANA
PEO-PEP 16k-323216k275Gyroid

To facilitate mixing of the block copolymer, catalyst, and curing agent into the epoxy matrix, the samples were mixed at room temperature in selected solvents. All thermoset samples were cast using one of two different solvents, tetrahydrofuran (THF) or acetone. The effect of the solvent selection on the resultant templated morphology will be addressed as part of this work; a more detailed description of the solvent casting process can be found elsewhere.4 The uncured samples consisted of the blended epoxy resin that had not been subjected to the thermal curing process. Although some crosslinking can occur during mixing, these materials were considered to have remained unreacted. The standard curing process was performed in a programmable temperature oven under nitrogen utilizing the following temperature program: one hour at 100 °C, 1 h at 125 °C, and 2 h at 150 °C, with a ramp rate of ∼5 °C/min. After the thermal curing process, the templated thermoset samples were allowed to cool to ambient temperature in the oven overnight.

The templated thermoset samples investigated in this study are listed in Table 2. Full thermal curing of the thermoset samples was determined by Raman spectroscopy and confirmed by differential scanning calorimetry (DSC). DSC was also used to determine the Tg of the samples. Microphase characterization of the cured toughened thermoset samples was performed by transmission electron microscopy (TEM). Images of the epoxy samples were obtained using an FEI-Philips Tecnai 12 TEM operating at 120 KeV in bright field mode. The samples were microtomed using a diamond knife at room temperature with a cutting rate of 0.4 mm/s to obtain ∼60 nm thick sections for imaging. The cut sections were floated onto 400 mesh copper grids using a water bath technique, and subsequently, stained by the vapors of an aqueous solution of 0.5 wt % ruthenium tetroxide (RuO4) for 2 min. In the TEM images, the PEO domains appear the darkest, the PEP domains are the least readily stained component by the RuO4 stain and thus remain the lightest, and the epoxy matrix is an intermediate shade of gray. The RuO4 staining process was also used for the templated thermoset samples utilizing the PEO-PE block copolymer. For the templated thermoset samples which contained the PEO-PI block copolymer additive, a staining agent of 4 wt % aqueous solution of osmium tetroxide (OsO4) was used. The PEO domains were still the most readily stained component in the PEO-PE and PEO-PI epoxy blends, and thus have the darkest appearance in the TEM images, with the PE or PI components appearing the lightest.

Table 2. Templated Thermoset Samples Investigated, the Block Copolymer, Casting Solvent Used, and the Resulting Templated Morphology Created
Templated Thermoset SampleCasting SolventBlock CopolymerTemplated Thermoset Morphology
Sample 1THFNANA
Sample 2THFPEO-PEP 10k-33Spherical
Sample 3AcetonePEO-PEP 10k-33Wormlike
Sample 4AcetonePEO-PEP 4.6k-33Spherical
Sample 5AcetonePEO-PEP 11k-42Spherical
Sample 6AcetonePEO-PI 11k-44Spherical
Sample 7THFPEO-PE 1.4k-50Spherical w/Rods
Sample 8AcetonePEO-PEP 16k-32Wormlike

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY
  7. Acknowledgements
  8. REFERENCES AND NOTES

The influence of the chemical composition of the block copolymer, the relative weight fraction of the two polymer blocks, and the total molecular weight of the copolymer on the resulting morphology in the epoxy matrix is the focus of this article. Table 1 lists the block copolymers used in this study, and the neat block copolymer phase morphology. The morphologies the block copolymers produced in the cured epoxy system are listed in Table 2. The selection of block copolymer substrate ranges and compositions was influenced by the work of Dean,4 who mapped out block copolymer composition and molecular weight ranges that gave rise to the various observed microstructures. For the present work in particular, molecular weights were chosen to be large enough to induce microphase separation in the uncured epoxy material such that morphology could be observed before and after the cure.

The uncured and thermally cured templated thermoset samples were characterized by TEM. A representative neat epoxy resin, Sample 1, was subjected to TEM imaging to ensure there were no visible microsize defects present, that is, voids or contaminants. The TEM images (not shown) were featureless, implying these samples were homogeneous throughout. All of the samples presented in this work were cured under exactly the same conditions. The templated thermoset Sample 2, which used the PEO-PEP 10k-33 block copolymer as the toughening agent and cast in THF, was imaged in both the uncured and cured states, as shown in Figure 2(a,b), respectively. Sample 2 exhibited the same spherical micelle structure regardless of the cure state. The spherical micelles in the TEM images were measured to be ∼15 nm in diameter in both the uncured and cured samples.

thumbnail image

Figure 2. TEM images of the (a) uncured and (b) cured epoxy Sample 2 containing the PEO-PEP 10k-33 block copolymer cast in THF. The spherical micelle morphology is retained from the uncured to the cured state.

Download figure to PowerPoint

When Sample 3, containing the same PEO-PEP 10k-33 block copolymer as in Sample 2, was cast using acetone instead of THF, a very different microstructure developed. The TEM image of Sample 3 exhibits the wormlike micelle morphology (Fig. 3). These results show the dramatic effect solvent selection can have on the morphology evolved in a templated thermoset sample. Dean4 also demonstrated the use of solvent selection to influence the morphology of the block copolymer in a epoxy matrix. She proposed the transition from the spherical micelle morphology to the wormlike micelle morphology occurred because acetone is a poorer solvent for PEO than the original solvent used in her study, chloroform. The use of a solvent that solvates the PEO phase less-well can drive the system to a different morphology, as the chains in the corona collapse towards the interface and induces crowding, such that the system minimizes interfacial tension by adopting a cylindrical morphology. The corona is not being as well solvated, and thus retracts towards the micellar interface, crowding the surface and leading to a transition to cylinders. That is, the interfacial area per chain and the extent of chain stretching in the PEO block is decreased in the wormlike structure compared to the spherical structure.24 This volume change is due to changes in solvation between the PEO block and the selected solvent. In Dean's work, the initial solvent used was chloroform compared to THF utilized in this work. However, acetone is also a poorer solvent for PEO compared with THF, so Dean's hypothesis should apply; the solubility parameters for acetone, chloroform, and THF are 19.7, 18.7, and 18.5 (J/cm3)1/2, respectively.25 The influence of solvent selection on the morphology of the block copolymer in solution has been demonstrated for a variety of block copolymer/solvent combinations.26–28

thumbnail image

Figure 3. High magnification TEM image of the cured Sample 3, containing the PEO-PEP 10k-33 block copolymer cast in acetone, stained with RuO4. The wormlike micelle morphology develops as a consequence of using acetone as the casting solvent. This sample contained the same block copolymer as Sample 2, PEO-PEP 10k-33, which was cast in THF.

Download figure to PowerPoint

The templated thermoset Samples 4 and 5 also possessed the spherical micelle morphology exhibited by Sample 2. However, unlike Sample 3, the use of acetone as the casting solvent did not result in the formation of the wormlike micelle morphology. Sample 4 utilized the PEO-PEP 4.6k-33 block copolymer, which was comprised of approximately the same weight fraction of PEO to PEP as the block copolymer used in Sample 3, but with a much lower molecular weight. For the templated thermoset Sample 5, the PEO-PEP 11k-42 block copolymer was employed. The PEO-PEP 11k-42 block copolymer had a comparable total diblock molecular weight with the block copolymer used in Sample 3, PEO-PEP 10k-33; however, the PEO-PEP 11k-42 block copolymer was composed of roughly 10 wt % more PEO. The SAXS results indicated the PEO-PEP 11k-42 possessed a lamellar microstructure while the PEO-PEP 10k-33 block copolymer exhibited a cylindrical microstructure due to the increase in PEO content.

Templated thermoset Sample 6 used a PEO-PI 11k-44 block copolymer as the additive. This templated thermoset sample utilized PI instead of PEP in the block copolymer. The PI polymer is the precursor to the hydrogenated PEP polymer used in the previous samples. Sample 6, cast in acetone, resulted in a spherical micelle morphology (Fig. 4). Unlike the spherical micelles observed in the templated thermosets utilizing a PEO-PEP block copolymer (Fig. 2), the micellar interface of the PEO-PI spherical micelles was very diffuse. Even in the cured state, a well-defined interface between the spherical micelles and the epoxy matrix did not develop and a distinct PI micelle core/PEO shell morphology was not apparent. A relative estimate of polymer miscibility can be made by comparing solubility parameters for polymer structures using repeat unit group contribution methods. The solubility parameter of PEP and PI are estimated to be 7.9 and 11.6 (J/cm3)1/2, respectively.29 The epoxy is estimated to have a solubility parameter of 20–22 (J/cm3)1/2, whereas the PEO has a solubility parameter of ∼20 (J/cm3)1/2.30, 31 Based on the estimated solubility parameters for PI compared to PEP, it is more likely that in the sample containing PI more phase mixing between the PI block and the epoxy-solvated PEO block would occur. Because more phase mixing could occur, a defined core between the PEO and PI blocks is less likely to develop to the same extent as was observed with the PEO-PEP block copolymers.

thumbnail image

Figure 4. TEM image of the cured Sample 6, made with the addition of a PEO-PI 11K-44 block copolymer cast in acetone. A diffuse spherical micelle morphology is visible.

Download figure to PowerPoint

The templated thermoset Sample 7 contained a commercial PE-PEO block copolymer as the additive and was solvent cast with THF. TEM images of this sample, Figure 5, show that a mix of morphologies developed in the epoxy matrix. The dominant morphology was spherical micelles which were relatively evenly distributed throughout the epoxy matrix. However, clusters of relatively large-scale rod-like structures were also visible. A similar structure has been observed in layered block copolymer-silicate nanocomposites.32–34 The diameter of the rod-like structures was relatively consistent at ∼30 nm. The PEO-PE block copolymer has a low molecular weight, MnPEO-PE = 1400 g/mol, with an estimated end-to-end distance of 156 Å, based on exclusion volume and monomer unit lengths given by Almdal et al.35 Thus, the rod-like structures were not simply produced by extended PEO blocks at the outer shell and the PE blocks in the core; a more complex structure must have formed. Because of the semicrystalline nature of the PE block in the PEO-PE additive, the rods could be a consequence of crystallite formation in the PE domains. DSC scans of this cured epoxy system also suggest the presence of PE crystallinity.

thumbnail image

Figure 5. TEM images of the cured Sample 7, made with the addition of a PEO-PE block copolymer cast in THF. A diffuse spherical micelle morphology is visible, with clusters of relatively large-scale rod-like structures present throughout.

Download figure to PowerPoint

The results presented in this article demonstrate that the morphologies that develop in templated thermoset samples are influenced by both the molecular characteristics of the block copolymer additive and the casting solvent. For the series of materials examined in this study, when the same templated thermoset sample was cast using acetone, as opposed to THF, the wormlike micelle morphology could be created from the former compared to the spherical micelle morphology in the latter. This is seen in the comparison of Sample 2, cast in THF, and Sample 3, cast in acetone [Figs. 2(b) and 3, respectively]. The initial results presented in this work indicate, that acetone, as opposed to THF, should be used as the casting solvent to produce the wormlike micelle morphology in this epoxy/block copolymer mixture. In addition, the PEO-PEP block copolymer needs to have a total number average molecular weight greater than Mn = 5000 g/mol and a ratio of the two blocks in the range of 20 < PEO wt % < 40, that is, within the cylindrical neat block copolymer microstructure phase window. It should be noted that a full understanding of the exact range of total molecular weight and block ratio required to produce the wormlike micelle morphology has not been established for these systems. The results presented in this section demonstrate the potential to customize the morphology in a templated thermoset sample by proper selection of the block copolymer additive and the solvent used in the casting process.

The use of a nano phase separating block copolymer as a thermoset toughening agent yielded dramatic toughening enhancement with no or little decrease in the Tg of the epoxy. It has been reported that at ≤ 5 wt % loading of PEO-PEP, in an epoxy matrix, the fracture toughness of the sample can increase threefold over that of the neat epoxy with less than a 10 °C drop in the glass transition temperature.17 One additional advantage of using a block copolymer as the toughening agent is that most of the samples remained optically clear after casting, primarily due to the nanometer size-scale of the block copolymer phases. Thermoset samples loaded with conventional rubber particles are often opaque as a result of the large size of the rubber particles. The block copolymer additives also have an advantage over their rubber particle counterparts because of the relative ease in controlling particle size.

For a set of unique block copolymer/epoxy/solvent systems, the development of the wormlike morphology during curing was observed. TEM images of Sample 3, cast from acetone, in the uncured and cured states, are shown in Figure 6. It can be observed that although some wormlike micelles are visible in the uncured sample, the spherical micelle morphology dominates. Alternatively, the wormlike morphology becomes the prominent morphology in the cured sample. Comparison of the TEM images of Sample 3 in the cured and uncured states indicate that the wormlike morphology developed from the spherical micelle morphology during the thermal curing process. This order–order phase transition during curing was also observed in Sample 8, cast from acetone as shown in Figure 7. It should be noted that even prior to the application of heat, the curing process was advancing in the epoxy matrix and the sample had become solidified. The driving force behind the formation of the wormlike micelle morphology and the mechanism by which they occur during the curing process are important to consider.

thumbnail image

Figure 6. TEM images of the (a) uncured and (b) cured Sample 3, containing the PEO-PEP 10k-33 block copolymer cast in acetone, showing the development of the wormlike micelle morphology during the thermal curing process. The fuzzy interface of the micelles in the uncured sample occurs because PEO is still miscible in the epoxy. As the epoxy progresses in the curing process the PEO polymer is expelled, indicated by the defined interface between the PEO and epoxy in the cured sample.

Download figure to PowerPoint

thumbnail image

Figure 7. TEM images of the uncured Sample 8, containing the PEO-PEP 16k-32 block copolymer cast in acetone, stained with RuO4. The images demonstrate the developing wormlike micelle morphology by coalescence of spherical micelle structures.

Download figure to PowerPoint

Although this work addresses the behavior of dilute concentrations of block copolymer in an epoxy matrix, theories developed for the behavior of dilute block copolymer solutions in a solvent can be applicable. The issue of the order–order phase transition of spherical micelles to rod (or wormlike) micelles in dilute diblock copolymers solutions has been addressed in the literature.26, 27, 36–41 Pendersen and Schurtenberger36 investigated morphological changes in dilute solutions of block copolymers as a function of temperature. They proposed that, in systems that began as spherical micelles, as the temperature increased and the solvent quality became increasingly poorer the spherical micelles became unstable. This instability was proposed to result from the core of the micelle becoming enlarged as the thickness of the corona was reduced in an effort to minimize contact with the poor solvent. The increase in the core radius thus drove the system to instability, with the formation of a rod or wormlike micelle morphology occurring to reduce the core volume and thus improve stability. This mechanism of the formation of wormlike micelles from spherical micelles has also been suggested by Termonia and others.26, 27, 40, 42

The expulsion of the PEO blocks from the epoxy matrix as a result of the curing process can be elucidated by comparing the block copolymer micelle-epoxy interface in the TEM images of the uncured and cured nanocomposites, as shown in Figure 6. The diffuse interface between the micelles and the epoxy matrix observed in the uncured Sample 3 is an indication that the PEO blocks were still miscible in the epoxy. As the crosslink density of the epoxy matrix increased the PEO blocks were expelled, creating a sharper interface between the PEO and epoxy, as indicated by the more well-defined interface between the micelles and matrix in the TEM image of the cured Sample 3. The expulsion of the epoxy miscible block from the epoxy matrix, as a result of the curing process, has also been observed in other templated thermoset systems utilizing a nonreactive block copolymer additive.19, 30, 43, 44

This mechanism for the morphological order–order phase transition described above can be used to describe the cure induced spherical to wormlike micelle transition observed in this work. In these systems, the epoxy matrix itself acts as an increasingly poor solvent for the micelle corona as the curing process progresses. As curing progresses, the spherical micelles swell and become unstable as the PEP cores of the spherical micelles become enlarged due to the expulsion of the PEO chains from the epoxy matrix. To stabilize the block copolymer structure the wormlike morphology forms, reducing the total core volume, and resulting in a smaller average diameter compared to the swelled spherical micelles. This spherical-to-wormlike order–order phase transition has also been described as a “wet” to “dry” brush transition.19 In support of this idea, a representative high magnification TEM image of the ends of wormlike micelles in the cured Sample 3 are shown in Figure 8. The diameter of the end caps of the wormlike micelles appear enlarged compared to the average diameter of the body of the wormlike micelles.

thumbnail image

Figure 8. High magnification TEM images of the cured Sample 3, containing the PEO-PEP 10k-33 block copolymer cast in acetone. The end of the wormlike micelles can be seen. The end caps of the wormlike micelles have a larger diameter than the diameter along the length of the worm.

Download figure to PowerPoint

Although there is significant data in the literature to support the driving forces behind the spherical to wormlike morphology, less is understood about how the formation of the wormlike micelles from the spherical micelles actually occurs. In a study by Zhang and Eisenberg26 of polystyrene-b-poly(acrylic acid) diblock copolymers, in a dimethyl-formamide (DMF)/water mixture, it was proposed that the wormlike micelles form by adhesive collision and fusion of the spherical micelles. Lam and Goldbeck-Wood38 simulated the morphological behavior of PEO-PBO block copolymers in an aqueous solution using dynamic mean field density theory, and predicted that the wormlike micelle structure would grow through coalescence of neighboring spherical micelles. Similarly, in the study of a dilute poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) (PEO-PPO-PEO) triblock copolymer, in an aqueous solution, Mortensen and Pedersen39 observed the elongation and distortion of the spherical micelles into ellipsoids in the presence of a poor solvent, resulting from increasing temperature.

It is proposed that the formation of the wormlike micelle structure from the spherical micelle structure occurred via a combination of the methods described above. As the crosslink density of the epoxy resin matrix increased, the PEO blocks are increasingly expelled and the micelles initially swell, becoming distorted and elongated. As the curing of the resin progresses, the unstable, distorted spherical micelles adhesively collide with neighboring micelles and form a structure similar to a string of interconnected spheres reported by Discher and others.39, 45, 46 From this collection of interconnected spheres, the more stable wormlike micelle structure forms. The high magnification TEM images of the uncured Sample 8, shown in Figure 7, support this method of wormlike micelle formation. Several series of spherical micelles aligned in rows, distorting as they are drawn together, are also noted.

An interesting point to note is that this proposed morphological transition occurred in a solidified thermoset. Because the templated thermosets viscosity increases dramatically with increasing levels of cure, the transition to the wormlike morphology should be impeded due to the repressed mobility of the block copolymer in the epoxy matrix as the epoxy crosslink density increases. For this reason, it is likely that the morphological transition occurs early in the curing process while localized mobility is possible. Lipic et al.19 reported a cure-induced sphere to cylinder phase transition in a templated thermoset sample containing a 35 wt % PEO-PEP block copolymer. As with our material, the phase transition was observed in systems that were characterized after gelation had taken place. However, it was noted that gelation is the point where a macroscopic, not microscopic, epoxy network is formed, so some localized mobility of the block polymer microstructure may remain well into the curing process.19 The rate of curing was not considered as a variable in this study, although it is recognized that the potentially differing rates of micellar reorganization and epoxy curing might well lead to unique results depending on the thermal profile used to bring about the epoxy matrix cure.

SUMMARY

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY
  7. Acknowledgements
  8. REFERENCES AND NOTES

The results presented in this article demonstrate that the morphology which develops in a templated thermoset can be influenced by the molecular parameters of the block copolymer additive. The chemical composition of the block copolymer, the relative weight fraction of the two polymer blocks, and the total molecular weight of the copolymer has been shown to influence the resulting morphology in the epoxy matrix. This work has demonstrated that the block copolymer morphology in the epoxy matrix can also be controlled by solvent selection. That is, a templated thermoset which exhibited a spherical micelle morphology when cast in THF could yield a wormlike micelle morphology when cast using acetone. The initial results presented in this work indicate that to produce the wormlike micelle morphology acetone, as opposed to THF, must be used as the casting solvent. In addition, the PEO-PEP block copolymer needs to have a total number average molecular weight greater than Mn = 5000 g/mol and a ratio of the two blocks in the range of 20 < PEO wt % < 40, that is, within the cylindrical block copolymer microstructure phase window. It should be noted that the exact range of the total Mn and block ratio that produces the wormlike micelle morphology has not been established. Additionally, it is unknown what other casting solvents, if any, could produce the worm-like micelle morphology or how the composition of epoxy resin influences the resultant templated morphology. The results presented in this article demonstrate the potential to customize the morphology in a templated thermoset sample by the selection of the block copolymer additive and solvent used in the casting process.

In this article, the cure-induced order–order phase transition from the spherical to wormlike micelle morphology was also discussed. Even in a solidified matrix, local reorganization of the PEP-PEO block copolymer structure was able to occur. It is believed that the phase transition occurred as a consequence of the PEO block being expelled from the epoxy matrix during curing as the crosslink density increases. Although this work addresses the phase behavior of a block copolymer in an epoxy matrix, theories developed for the behavior of dilute block copolymer solutions in a solvent were shown to be applicable. Although the TEM images seem to agree with the proposed method of wormlike micelle growth from spherical micelles, a more detailed investigation of these mechanisms is recommended.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY
  7. Acknowledgements
  8. REFERENCES AND NOTES

The authors would like to thank Frank Bates, University of Minnesota and Nikhil Verghese and Ha Pham, The Dow Chemical Company for their guidance and support during this project.

REFERENCES AND NOTES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY
  7. Acknowledgements
  8. REFERENCES AND NOTES