Toughening Epoxy Thermosets with Block Ionomers: The Role of

Oct 10, 2013 - But the toughening efficiency varies with the phase domain size, and ... Pingan Song , Zhiguang Xu , Matthew S. Dargusch , Zhi-Gang Che...
5 downloads 0 Views 4MB Size
Article pubs.acs.org/Macromolecules

Toughening Epoxy Thermosets with Block Ionomers: The Role of Phase Domain Size Shuying Wu,† Qipeng Guo,*,† Martin Kraska,‡ Bernd Stühn,‡ and Yiu-Wing Mai§ †

Polymers Research Group, Institute for Frontier Materials, Deakin University, Locked Bag 2000, Geelong, Victoria 3220, Australia Institut für Festkörperphysik, Technische Universität Darmstadt, Hochschulstraße 6-8, Darmstadt 64289, Germany § Centre for Advanced Materials Technology (CAMT), School of Aerospace, Mechanical and Mechatronic Engineering J07, The University of Sydney, Sydney, New South Wales 2006, Australia ‡

ABSTRACT: Herein we report a novel approach to toughen epoxy thermosets using a block ionomer, i.e., sulfonated polystyrene-block-poly(ethylene-cobutylene)-block-polystyrene (SSEBS). SSEBS was synthesized by sulfonation of SEBS with 67 wt % polystyrene (PS). Phase morphology of the epoxy/SSEBS blends can be controlled at either nanometer or micrometer scale by simply adjusting the sulfonation degree of SSEBS. It has been found that there exists a critical degree of sulfonation (10.8 mol %) forming nanostructures in these epoxy/SSEBS blends. Above this critical value, macrophase separation can be avoided and only microphase separation occurs, yielding transparent nanostructured blends. All epoxy/SSEBS blends display increased fracture toughness compared to neat epoxy. But the toughening efficiency varies with the phase domain size, and their correlation has been established over a broad range of length scales from nanometers to a few micrometers. In the nanostructured blends with SSEBS of high sulfonation degrees, the fracture toughness decreases with decreasing size of the phase domains. In the macrophase-separated blends, only a slight improvement in toughness can be obtained with SSEBS of low sulfonation degrees. The epoxy blend with submicrometer phase domains in the range 0.05−1.0 μm containing SSEBS of a moderate degree of sulfonation (5.8 mol %) displays the maximum toughness. This study has clearly clarified the role of phase domain size on toughening efficiency in epoxy thermosets.



INTRODUCTION Seeking effective toughening strategies for epoxy thermosets is of considerable importance and various methods have been developed in the past few decades.1−5 One of the most common and successful strategies is to incorporate a second phase, either elastomers3 or thermoplastics,4,5 into the matrix, forming multiphase morphology. Recently, nanostructured epoxy thermosets based on block copolymers have been widely reported.6−20 It has been shown that the nanoscale phase domains can significantly improve the toughness of polymer blends at low loadings.21,22 Sue and co-workers have reported direct experimental evidence about the mechanisms responsible for such toughening effect.23−25 In these toughened epoxy systems, the size of the phase domains plays a significant role since different toughening mechanisms and deformation processes may be induced.26−28 However, few systematic studies have been reported to obtain the correlation between the toughening efficiency and the phase domain size, especially over a broad range from the nanometer scale to the micrometer scale. It has been reported that, in diblock copolymer toughened epoxy thermosets,27 the largest increase in toughness is observed in blends containing 0.1−1.0 μm spherical phase domains. In previous work,29,30 we investigated the preparation of toughened nanostructured epoxy thermosets using block © 2013 American Chemical Society

ionomer complexes. It was originally inspired by using the concept of functionalization to promote compatibilization between thermosetting matrix and immiscible modifiers. That is, sulfonation of styrene block introduces sulfonic acid groups into the nonpolar polymer chains, which can promote favorable interactions with the epoxy network.31−33 However, neither the parent block copolymer polystyrene-block-poly(ethylene-ranbutylene)-block-polystyrene (SEBS, with 29 mol % polystyrene) nor the corresponding block ionomer sulfonated SEBS (SSEBS, with sulfonation degree up to ∼40 mol %) was found miscible with diglycidyl ether of bisphenol A (DGEBA) epoxy resin. Nevertheless, a block ionomer complex of SSEBS with 3(dimethylamino)propylamine-terminated poly(ε-caprolactone) were synthesized and used successfully to prepare nanostructured epoxy blends.29 The presence of epoxy-miscible poly(ε-caprolactone) side chains in the block ionomer complex avoided macrophase separation. Correlation of nanostructures with mechanical properties of the resultant nanostructured epoxy thermosets was obtained.30 Notwithstanding this important achievement, there remains a major gap to develop Received: July 13, 2013 Revised: September 28, 2013 Published: October 10, 2013 8190

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Small-Angle X-ray Scattering (SAXS). SAXS experiments were conducted at the Australian Synchrotron on the small/wide-angle X-ray scattering beamline utilizing an undulator source that allowed measurement at a very high flux to moderate scattering angles and a good flux at the minimum q limit (0.012 nm−1). The intensity profiles were interpreted as the plot of scattering intensity (I) versus scattering vector, q = (4π/λ) sin(θ/2) (where θ is scattering angle and the wavelength λ was 0.062 nm). Scanning Electron Microscopy (SEM). The multiphase structures of the epoxy blends were studied by SEM on the fracture surface of specimens broken under cryogenic conditions using liquid nitrogen. The fracture surface was immersed in THF at room temperature for 24 h. The block ionomer phase and/or the sol fraction of the epoxy network were preferentially etched by the solvent whereas the cured epoxy phase remained relatively unaffected. Then, the etched samples were dried in a vacuum oven to remove the solvent. The phase structure was observed using a field emission gun scanning electron microscope (Zeiss, Supra 55 VP). Fracture surfaces were coated with a thin gold layer to avoid charging of the samples. Images were obtained under an accelerating voltage of 5 kV with a working distance of 10 mm. Differential Scanning Calorimetry (DSC). TA Q200 differential scanning calorimeter was used for calorimetric measurements. Indium and tin standards were adopted to calibrate low- and high-temperature regions, respectively. ∼5 mg samples were used in the experiments. Except indicated otherwise, all samples were first heated to 100 °C from 0 °C at a rate of 20 °C/min (first heating scan) and kept at that temperature for 5 min; subsequently, they were cooled at a rate of −10 °C/min (cooling scan). Following the cooling scan, a second scan was conducted at the same heating rate. Tg values were taken as the midpoint of the transition in the second scan of the DSC curves. Dynamic Mechanical Thermal Analysis (DMTA). Dynamic mechanical tests were performed on a dynamic mechanical thermal analyzer (DMTA) (TA Q800) in a single-cantilever mode under liquid nitrogen. Five different frequencies were used (0.2, 0.4, 1, 2, and 4 Hz) at a heating rate of 3.0 °C/min. The specimen dimensions were 30 × 5.0 × 2.0 mm3. The storage modulus (G′), loss modulus (G″), and tan δ were measured from −100 to 250 °C. Tα was taken at the maximum of the tan δ curve in the glass transition region. Mechanical and Fracture Toughness Tests. The epoxy/SSEBS blends were subjected to tensile testing using an Instron 30 kN screw driven tensile tester equipped with a noncontact video extensometer. Dumbbell specimens with a 16.0 × 3.5 × 2.5 mm3 neck were used. Tests were performed at room temperature at a constant crosshead speed of 5 mm/min. At least five samples were tested to obtain the average values of tensile properties for all blends and neat epoxy. Samples typically failed abruptly in the gauge section. The plane-strain fracture toughness testing of the blends was conducted according to ASTM D5045 standard using a three-pointbend (SEN-3PB) single-edge-notch specimen.37 Plane strain energy release rate (GIC) was determined from the plane strain critical stress intensity factor (KIC) obtained. Specimens with a sharp notch were prepared by using a mold designed for this purpose. Then, a sharp crack tip was initiated by tapping a fresh razor blade. Care was taken to avoid forming a long crack or breaking the sample. Specimens with a long crack, i.e., a/W exceeding 0.55, were discarded. The specimen dimensions and calculation methods of GIC and KIC were the same as those given in our previous work.30

more effective modifiers which require a simple preparation procedure. Herein, we report a new route to toughen epoxy thermosets with a block ionomer. The block ionomer sulfonated polystyreneblock-poly(ethylene-co-butylene)-block-polystyrene (SSEBS) was synthesized by sulfonation of polystyrene-block-poly(ethylene-co-butylene)-block-polystyrene (SEBS) with 67 wt % of polystyrene (PS). In this work, only a block ionomer rather than a block ionomer complex was used as a new type of epoxy modifier. Hitherto, there are no reports on the toughening of epoxy thermosets with block ionomers. Epoxy blends with 10 wt % SSEBS of different sulfonation degree were prepared, aiming to systematically investigate the effects of sulfonation degree on the scale of phase separation so as to establish the phase domain size - fracture mechanical property relationship. Macroscopic and microscopic phase-separated structures could be selectively obtained. Phase structure of the epoxy blends was thoroughly studied by electron microscopy and small-angle X-ray scattering (SAXS) and correlated with their toughness and mechanical properties, providing a basic understanding on the structure−property relationship over a wide range from the nanometer scale to the micrometer scale.



EXPERIMENTAL SECTION

Materials and Preparation of Samples. Polystyrene-blockpoly(ethylene-co-butylene)-block-polystyrene (SEBS, Tuftec H1043) consisting of 67 wt % polystyrene was supplied by Asahi Kasei Chemicals Corporation, Tokyo, Japan. The average molecular weight Mw of the SEBS block copolymer was 40 000, and Mw/Mn was 1.05 as measured by GPC in tetrahydrofuran (THF) relative to polystyrene standard. All other chemicals including acetic anhydride, concentrated sulfuric acid (96%), 1,2-dichloroethane (DCE), isopropyl alcohol (IPA), and THF were reagent grade. The epoxy precursor diglycidyl ether of bisphenol A (DGEBA) with an epoxide equivalent weight of 172−176 and the curing agent 4,4′-methylenedianiline (MDA) were purchased from Sigma-Aldrich Co. To prepare sulfonated SEBS, SEBS was subjected to reaction with acetyl sulfonate from acetic anhydride and concentrated sulfuric acid. The procedure of sulfonation was described in detail in our previous studies.29,30 The SSEBS obtained was characterized by FTIR. Characteristic bands at around 1003 and 1124 cm−1, 1035 and 1172 cm−1 can be observed in the FTIR spectra (not shown here), assignable to in-plane bending vibrations of the aromatic ring substituted with sulfonate groups and symmetric and asymmetric stretching vibrations of the sulfonate group (OSO), respectively.35,36 The sulfonation degree of SSEBS, that is, content of sulfonic group grafted onto SEBS, was evaluated by titration with standard sodium hydroxide solution (0.1 N) to a phenolphthalein end point. SSEBS with four different sulfonation degrees were prepared, and these were 0.9, 5.8, 10.8, and 21.9 mol %, respectively. Thus, they are denoted correspondingly as 0.9SSEBS, 5.8SSEBS, 10.8SSEBS, and 21.9SSEBS. To prepare epoxy thermosets containing the SSEBS block ionomer, DGEBA and SSEBS were separately dissolved in THF. Then, individual THF solutions of DGEBA and SSEBS were mixed and rigorously stirred. Meanwhile, a stoichiometric amount of curing agent MDA was added and stirred to form a homogeneous mixture. The solvent was evaporated at room temperature in the fume hood first and the vacuum oven next. The mixture was then poured into a preheated mold and cured at 120 °C for 17 h and postcured successively at 180 °C for 2 h. Transmission Electron Microscopy (TEM). A JEOL JEM-2100 transmission electron microscope was employed to conduct TEM experiments at an acceleration voltage of 100 kV. The samples were cut into ultrathin sections of ∼70 nm thick with a diamond knife using a Leica EM UC6 ultramicrotome, collected on 400 mesh copper grids and stained with ruthenium tetroxide (RuO4) for TEM examination.



RESULTS AND DISCUSSION Morphology of Epoxy/21.9SSEBS Blends. TEM images of 97.5/2.5, 95/5, 90/10, 85/15, and 80/20 epoxy/21.9SSEBS blends are displayed in Figure 1. It was observed that all epoxy blends were transparent at 120 °C before curing, which indicated the mixtures were macroscopically homogeneous. From TEM images of the cured samples, the heterogeneous morphology at nanoscale is seen in all cases showing no 8191

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Figure 1. TEM images of epoxy/21.9SSEBS blends: (a) 97.5/2.5, (b) 95/5, (c) 90/10, (d) 85/15, and (e) 80/20.

reaction-induced macroscopic separation had taken place in the 21.9SSEBS modified epoxy. It is noted that the epoxy/21.9SSEBS blends display similar morphology as the epoxy system modified with block ionomer complex (as revealed in our previous studies29,30), viz., spherical morphology consisting of a bright spherical core and a dark thin shell. These spherical microdomains are dispersed in the gray continuous areas. Owing to the electron density difference of different groups and the diversity in the preference of RuO4 staining,38,39 the gray continuous areas can be ascribed to the cross-linked epoxy matrix. Likewise, the bright spherical cores can be attributed to the rubbery EB (poly(ethylene-ranbutylene)) blocks, whereas the thin dark shells surrounding the spherical cores can be assigned to the microphase-separated SPS (sulfonated polystyrene) blocks. The average size of the spherical microdomains in the 97.5/2.5 epoxy/21.9SSEBS blend is ∼15 nm. As the 21.9SSEBS content increases, the spherical morphology consisting of a bright core and a dark shell remains unchanged, but smaller microdomains are seen. Also, the number of spherical domains increases and the distance between them decreases. According to the miscibility of the subchains of SSEBS with epoxy matrix, the formation of this spherical core−shell

nanostructure is suggested to follow similar mechanism to the previous epoxy system modified with block ionomer complex. The rubbery EB blocks are immiscible with epoxy precursor and therefore first excluded, forming spherical cores prior to curing. As for the SPS blocks, it is not easy to ascertain that phase separation of SPS blocks takes place before curing or is induced by the curing merely based on the solubility parameter owing to their interactions with the epoxy network.29,30 The microphase structure of MDA-cured epoxy blends with block ionomer was further investigated by SAXS whose profiles are shown in Figure 2. Well-defined scattering peaks are observed for all the blends except the 97.5/2.5 epoxy/21.9SSEBS blend, indicating that the blends exhibit nanoscale structures and may contain long-range ordered nanostructures. No pronounced scattering peaks in the SAXS profile of the 97.5/2.5 epoxy/21.9SSEBS blend are displayed. The average distance between neighboring microdomains (d = 2π/q*) in this blend is ∼100 nm (see Figure 1a), which should generate a first-order scattering maximum at q* ≈ 0.063 nm−1. This position is at the edge of the experimental q-window, and it is hence not detected. In addition, it is difficult to obtain quantitative information due to the very low content of the epoxy immiscible block at this 8192

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Figure 3. Rc, Ts, Rhs, and σ obtained from fitting SAXS data versus the block ionomer content.

Figure 2. SAXS profiles of epoxy/21.9SSEBS blends of different composition. Scattering profiles were shifted vertically for clarity. The full lines represent the fitting curves.

domains become much more uniformly distributed in the epoxies. The variation of the structural parameters with block ionomer content can be interpreted from the formation mechanism of the spherical core−shell nanostructures. First, the core consists of epoxy immiscible EB block which increases in content with increasing amount of 21.9SSEBS in the blends. Therefore, larger cores are formed due to the increase in volume fraction of the EB block. The shell, however, consists of microphase-separated SPS blocks. The shell thickness is thus determined by the extent of phase separation of SPS blocks. In the present system, the epoxy/21.9SSEBS blend with the least amount (5 wt %) of block ionomer presents the thickest shell. This indicates more SPS blocks are phase separated even though this blend contains the least amount of SPS blocks. Finally, the mean hard sphere interaction radius Rhs, defined as half the distance between the hard sphere centers at their closest contact,43 decreases with increasing content of 21.9SSEBS, presumably caused by the dramatic decrease of the average distance between the spherical phase domains. Thermomechanical Properties of Epoxy/21.9SSEBS Blends. MDA-cured neat epoxy, epoxy/21.9SSEBS blends, parent SEBS, and 21.9SSEBS were measured by DSC. Their DSC thermograms of the second scan are shown in Figure 4. The parent SEBS exhibits a glass transition at ∼81 °C, which can be ascribed to the glass transition of the PS block.44 After the sulfonation, this glass transition temperature increased to ∼107 °C, possibly caused by the hydrogen-bonding interactions between the sulfonic acid groups in the SPS subchains.44 The neat epoxy displays a glass transition temperature at 176 °C which becomes broad and ambiguous in the blends. Moreover, Tg generally shifts down to a lower temperature with increasing content of 21.9SSEBS. For the epoxy blends containing 20 wt % 21.9SSEBS, Tg decreases to ∼154 °C, which is probably due to the penetration of soft chains into the epoxy network (plasticization effect).45,46 There is an exception

composition. For the epoxy blends with higher block ionomer content (≥5 wt %), a first-order scattering peak can be clearly seen and its position shifts to a higher q value corresponding to a shorter average distance. More specifically, a first-order scattering peak situated at q* = 0.17 nm−1 is observed for the 95/5 epoxy/21.9SSEBS blend corresponding to a long spacing at 36.7 nm which is decreased gradually to 24.6 nm for the epoxy blend with 20 wt % 21.9SSEBS. Further, a second-order scattering peak is clearly visible at 2q*. In addition, there is another discernible secondary scattering peak which can be identified as a form factor scattering from the spherical microdomains in liquid-like order (see ref 30). To extract quantitative structural information on the spherical nanostructures, we fitted the SAXS data using a model of core− shell type domains with liquid-like order. The overall intensity I(q) of scattered X-rays in the SAXS experiment can be described in terms of a form factor P(q;Rc;Ts), corresponding to a sphere with one shell40,41 and a hard sphere structure factor S(q;Rhs)42 describing interparticle correlations. Moreover, a normalized Schultz distribution for the core radii was assumed in this core−shell model. Details about this model were given and written as eqs 4 and 5 in our previous paper.30 The fitting curves are given as full lines in Figure 2. It can be seen that this model fit the data very well. Figure 3 shows the structural parameters, the average core radius Rc, the shell thickness Ts, the hard sphere radius Rhs, and the core polydispersity σ derived from the core−shell model plotted as a function of block ionomer content. For the 95/5 epoxy/ 21.9SSEBS blend, Rc, Ts, and Rhs are 2.23, 3.24, and 16.27 nm, respectively. With increasing content of 21.9SSEBS, Rc increases, reaching 4.17 nm for the 80/20 epoxy/21.9SSEBS blend. Opposite to this trend, Ts and Rhs decrease with increasing wt % of 21.9SSEBS. It is also noted that the polydispersity of the core radius decreases remarkably with increasing block ionomer content, indicating the spherical 8193

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

density. However, at the same time, the phase-separated domains occupied positions in between the reaction sites and thereby hindering the curing reaction at some particular sites.52 These two opposite effects compete with each other and account for the variations of Tg and storage modulus. Mechanical Properties of Epoxy/21.9SSEBS Blends. The tensile mechanical properties of epoxy/21.9SSEBS blends of different composition, that is, the average values of Young’s modulus, tensile strength, and elongation at break, are listed in Table 1. The addition of block ionomer to the neat epoxy causes reductions in Young’s modulus and tensile strength but improvement in elongation at break. These trends are more remarkable for epoxy blends with higher content of block ionomer. Hence, when the block ionomer content is increased from 0 to 10 wt %, Young’s modulus and tensile strength decrease from 2.48 GPa and 62.78 MPa (for neat epoxy) to 2.39 GPa and 47.36 MPa, respectively, whereas the elongation at break increases from 3.69% (for neat epoxy) to 6.85%. Nevertheless, the fracture toughness of the epoxy blends with block ionomer is found to be improved dramatically. KIC and GIC of neat epoxy are 0.77 MPa·m1/2 and 0.21 kJ/m2, which increase to 1.09 MPa·m1/2 and 0.43 kJ/m2, respectively, for the 97.5/2.5 epoxy/21.9SSEBS blend. With increasing 21.9SSEBS content, the toughness increases gradually, and the most remarkable improvement comes with the blend with 10 wt % 21.9SSEBS where KIC and GIC are 1.39 MPa·m1/2 and 0.74 kJ/m2, respectively. These large increases can be correlated to the structural characteristics of the epoxy blends. Based on a previous report,53 epoxies modified with block copolymers selfassembled into a spherical geometry (that is, vesicles or spherical micelles); the toughness increases with decreasing ratio of interparticle distance to average diameter of spherical micelles. From the aforementioned SAXS analysis, Rc increases with increasing 21.9SSEBS content while the d-spacing between neighboring domains decreases dramatically from 36.7 to 25.6 nm (based on the position of the first-order scattering maximum). Hence, a more remarkable improvement in toughness can be observed, since the ratio is significantly reduced for the 90/10 epoxy/21.9SSEBS blend. There are two common explanations: (a) possible stress field overlapping of adjacent particles54 and (b) transition from plane strain to plane stress experienced by the ligament between neighboring particles when a critical interparticle distance is reached.55 Further, the SPS shell thickness decreases with increasing block ionomer content according to the SAXS study, which indicates more SPS blocks penetrate into the epoxy matrix. This results in stronger adhesion between the spherical microdomains and epoxy matrix, leading to more noticeable toughening efficiency. Miscibility of Epoxy with SSEBS of Different Degree of Sulfonation. It is noted from the above discussions that the sulfonation of PS blocks plays a key role in forming the nanoscale structures in epoxy blends. The interactions between sulfonic acid groups of SSEBS and epoxy are the main driving force to effectively avoid macroscopic phase separation. Thus, the sulfonation degree of SEBS should have deep impacts on the microphase structures and thereby the mechanical properties. In this study, we have examined the critical sulfonation degree required to form nanoscale structures in the epoxy blends without compromising the optical transparency. That is, to ensure no macroscopic phase separation occurs in the cured epoxy blends. For comparison, epoxy blends with the same amount (10 wt %) of SSEBS but different sulfonation degree of 0 (SEBS), 0.9 (0.9SSEBS), 5.8 (5.8SSEBS), and 10.8 mol % (10.8SSEBS) were

Figure 4. DSC curves of the second scan of epoxy/21.9SSEBS blends at a heating rate of 20 °C/min.

that the 97.5/2.5 epoxy/21.9SSEBS blend shows a slight increase compared to neat epoxy. Figures 5a−f show the storage modulus and tan δ vs temperature curves of neat epoxy resin and nanostructured epoxy thermosets containing up to 20 wt % of 21.9SSEBS. The tests were conducted at frequencies of 0.2, 0.4, 1.0, 2.0, and 4.0 Hz. The MDA-cured neat epoxy (Figure 5a) shows three typical relaxations of the epoxy network, in agreement with previous study.47 They are at around 180, 57, and −50 °C, denoted by α, β′, and β relaxations, respectively. The α-relaxation in the spectra can be associated with the glass transition while the βrelaxation at the lower temperature (∼−50 °C) is known to arise from the motion of the hydroxyl ether structural units [CH2−CH(OH)−CH2−O−] and other parts in the polymer chains.47 The β′ relaxation can be attributed to motions of the less cross-linked zones in the network according to Pogany48 and Arridge.49 However, it was also suggested that this relaxation is likely to be associated with motions of p-phenylene.50 For the MDA-cured epoxy/21.9SSEBS blends, α, β′, and β relaxation peaks are clearly observed in all the spectra. Adding 2.5 wt % 21.9SSEBS increases Tα of the epoxy blend to 185 °C, which is consistent with the DSC results described above. By increasing the content of 21.9SSEBS further, Tα of the epoxy thermosets decreases gradually to 153 °C for the 80/20 epoxy/ 21.9SSEBS blend. In addition, both β and β′ relaxation peaks shift down slightly to lower temperatures. Moreover, for a given epoxy blend, Tα generally moves toward a higher temperature with increasing test frequency. The storage moduli at both glassy region and rubber plateau were found to increase slightly at low loading (2.5 wt %), which then decrease gradually with further increasing content of 21.9SSEBS. The presence of the soft rubbery phase will reduce the stiffness of the relatively rigid epoxy matrix, thus decreasing the modulus. From the theory of rubber elasticity,51 the plateau storage modulus at temperature above Tg is proportional to the cross-link density. Therefore, it can be inferred that the cross-link density decreases gradually with increasing 21.9SSEBS content. The interactions between 21.9SSEBS and epoxy network may increase the cross-link 8194

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Figure 5. Variation of storage modulus and tan δ with temperature at different frequencies (0.1, 0.4, 1, 2, and 4 Hz) of (a) neat epoxy and epoxy/ 21.9SSEBS blends: (b) 97.5/2.5, (c) 95/5, (d) 90/10, (e) 85/15, and (f) 80/20.

was translucent. By contrast, the epoxy blend containing the same amount of 10.8SSEBS is transparent with no discernible macroscopic separation, indicating the formation of nanostructured

prepared. It should be noted that the cured epoxy blends containing 10 wt % of SEBS and 0.9SSEBS were opaque, showing macroscopic phase separation, while the blend containing 5.8SSEBS 8195

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Table 1. Mechanical Properties of Nanostructured Epoxy/21.9SSEBS Blends epoxy/21.9SSEBS neat epoxy 97.5/2.5 95/5 90/10 a

Young’s modulus (GPa) 2.48 2.44 2.41 2.39

± ± ± ±

0.05 0.01 0.08 0.03

tensile strength (MPa) 62.78 56.29 52.19 47.36

± ± ± ±

elongation at break (%)

1.61 2.89 3.01 2.81

3.69 5.34 6.80 6.85

± ± ± ±

0.24 0.15 0.20 0.19

KIC (MPa·m1/2) 0.77 1.09 1.34 1.39

± ± ± ±

0.02 0.01 0.03 0.04

GIC (kJ/m2)

GIC/GIC,ma

± ± ± ±

1.00 2.05 3.19 3.52

0.21 0.43 0.67 0.74

0.03 0.01 0.03 0.02

GIC,m is plane strain energy release rate for neat epoxy.

Figure 6. DSC curves of the second scan of (a) SEBS and SSEBS with different sulfonation degree and (b) the corresponding epoxy blends at a heating rate of 20 °C/min.

lower temperatures compared with neat epoxy. For epoxy blends containing SEBS, 0.9SSEBS, and 5.8SSEBS, the presence of two glass transitions assignable to the glass transition of the epoxy-rich phase indicates that there are probably compositional gradients in the epoxy matrix.57 Moreover, both glass transition temperatures decrease when SSEBS with higher sulfonation degrees are used as the modifier, which is probably due to the fact that SSEBS with higher sulfonation degrees show better miscibility with epoxy matrix leading to more pronounced plasticization effect. For the same reason, the epoxy blend containing 10.8SSEBS displays a higher Tg than that with 21.9SSEBS. The DMTA spectra of the epoxy blends with 10 wt % SEBS and SSEBS of different sulfonation degree are illustrated in Figure 7. These spectra provide further evidence of the miscibility of the epoxy blends. For blends containing 10 wt % 10.8SSEBS, there are three typical relaxations, α, β′, and β relaxations, in agreement with the epoxy/21.9SSEBS blends discussed above. They occur at around 162, 50, and −50 °C, respectively. Compared with the epoxy blends with 10 wt % 21.9SSEBS, all of these three relaxation temperatures increase slightly, probably due to the less pronounced plasticization effect originating from the smaller amount of sulfonic acid groups in 10.8SSEBS. Apart from these typical relaxations, a relaxation peak at ∼85 °C can be clearly observed for the epoxy blends containing SEBS, 0.9SSEBS, and 5.8SSEBS. This relaxation could be attributed to the α relaxation of SPS microdomains, indicating that SPS blocks are demixed from the epoxy matrix consistent with the opaque appearance of these blends. Moreover, it is obvious that there are two major transitions

epoxy blends. To study the miscibility of SSEBS with epoxy, samples were subjected to DSC and DMTA tests. Figure 6 shows the second scan DSC thermograms. The DSC curve of SEBS (Figure 6a) displays a broad endothermic peak between −20 and 20 °C, which is probably due to the melting of small crystallites formed by long sequences of ethylene.56 Another obvious endothermic transition at 81 °C can be ascribed to the Tg of the PS block. According to a previous report,56 the rubbery block (EB block) should have a Tg at around −40 °C. However, there is no discernible Tg observed at this temperature which might be due to the insensitivity of DSC for detecting Tg. Otherwise, the different composition of the present SEBS (67 wt % PS) might result in a higher Tg which could overlap with the melting peak between −20 and 20 °C. For the SSEBS, Tg of PS block generally increases with increasing sulfonation degree, in agreement with a previous study.44 21.9SSEBS with the highest sulfonation degree exhibits a Tg of SPS block at 107 °C. Figure 6b shows the DSC curves of epoxy blends with SSEBS of different sulfonation degree. The neat epoxy exhibits a glass transition (Tg) at 176 °C. For the epoxy blends, there are no discernible endothermic transitions which can be assigned to the PS (or SPS) and EB blocks. Two possible reasons account for this, namely, the components of the blends are miscible or DSC has a relatively low sensitivity for analyzing the glass transition of polymer components. However, it can be clearly seen that there are two glass transitions at high temperature (around 150 °C) for the blends containing SEBS, 0.9SSEBS, and 5.8SSEBS whereas only one for those containing 10.8SSEBS and 21.9SSEBS. Epoxy/10.8SSEBS and epoxy/21.9SSEBS blends display Tg at 161 and 156 °C, respectively, shifting to 8196

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Figure 7. Variation of storage modulus and tan δ with temperature at different frequency (0.2, 0.4, 1, 2, and 4 Hz) of epoxy blends with 10 wt % of (a) SEBS, (b) 0.9SSEBS, (c) 5.8SSEBS, and (d) 10.8SSEBS.

10.8SSEBS, 5.8SSEBS, 0.9SSEBS, and SEBS. Note that these images were taken at different magnifications. The transparent epoxy/10.8SSEBS blend shows heterogeneous morphology at nanoscale. Similar to the epoxy blends with 21.9SSEBS, a spherical morphology where bright spherical cores are surrounded by thin dark shells and well-dispersed in the gray continuous matrix (Figure 8a) is observed. Compared to the 90/10 epoxy/21.9SSEBS blend (see Figure 1c), the average size of the spherical phase domains in 90/10 epoxy/10.8SSEBS blend increases from ∼11 to ∼18 nm. By contrast, the dispersed dark areas are highly aggregated in the epoxy blend with 10 wt % of 5.8SSEBS (Figure 8b). TEM images of epoxy blends with 0.9SSEBS and SEBS shown in Figures 8c and 8d clearly demonstrate that a micrometer scale phase-separated morphology exists in these two blends. The macroscopically phase-separated morphology of these epoxy blends was further investigated by SEM. Figure 9 shows SEM images of cryogenic fractured surfaces of epoxy blends containing 10 wt % of 5.8SSEBS, 0.9SSEBS, and SEBS in which the heterogeneous morphology is seen in all cases. For further examination, the fracture surfaces were etched with THF to reveal more clearly the morphology. The images on the right-hand

at high temperature assignable to glass transition of the epoxyrich phase in the DMTA spectra of epoxy/SEBS and epoxy/ 0.9SSEBS blends. For both of these blends, a transition at 179 °C, almost the same as that of the neat epoxy, is observed. The other transition occurs at 148 and 146 °C, respectively, for epoxy/SEBS and epoxy/0.9SSEBS blends. The presence of these two major transitions may be due to the fact that there are compositional gradients in epoxy matrix.57 It is proposed that the epoxy matrix close to the interface of the two components is possibly mixed with some PS or SPS chains exhibiting a lower α relaxation temperature due to the plasticization effect, whereas that far from the surface remains unmixed with these chains and displays almost the same glass transition as neat epoxy. It is noted that the epoxy/5.8SSEBS blend does not show two glass transitions of the epoxy-rich phase, as seen in the DSC study. However, there is seemingly an ambiguous shoulder peak next to the major transition, which may account for the presence of two glass transitions in the DSC measurements. Morphology of Epoxy Blends Containing SSEBS of Different Degree of Sulfonation. The morphology of the cured epoxy/SSEBS blends was investigated by TEM. Figure 8 shows TEM images of epoxy blends containing 10 wt % of 8197

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Figure 8. TEM images of epoxy blends containing 10 wt % of (a) 10.8SSEBS, (b) 5.8SSEBS, (c) 0.9SSEBS, and (d) SEBS.

these epoxy blends, SSEBS of different sulfonation degree was also subjected to SAXS examination and the obtained SAXS profiles are shown in Figure 10a. For the parent SEBS, the firstorder peak is centered at a value of the scattering vector q corresponding to a long spacing of 24.9 nm. Also, a secondorder reflection is clearly visible at the double angular position of this first-order maximum. The pronounced first-order peak and the double angular position of secondary scattering indicate a well-ordered system which might microscopically phase separate into a simple cubic phase, a gyroid structure, or a lamellar structure as is known for SEBS.38 Similar SAXS scattering patterns are observed for SSEBS of different degree of sulfonation. The difference is that the first-order scattering peak is situated at a slightly higher q value for SSEBS with a higher sulfonation degree. The q value reaches 0.28 nm−1, corresponding to a distance in real space at 22.6 nm for 21.9SSEBS. Shown in Figure 10b are the SAXS profiles of the epoxy blends containing SSEBS of various sulfonation degrees. It can be seen from these profiles that quite different scattering patterns are exhibited depending on the sulfonation degree of the ionomers. When SSEBS with a sulfonation degree of 5.8 mol % or lower was used, a sharp first-order scattering is found situated at almost the same q value as the corresponding neat block ionomer. However, the secondary scattering peak becomes so weak that it is not clearly visible, which indicates that the structure has been changed and becomes less ordered in the epoxy/SSEBS blends. In previous work by Guo et al.,39 it was found that the epoxy immiscible block copolymer SB (polystyrene-b-polybutadiene) is microphase separated in macroscopically phase separated epoxy/SB blend. This is deduced from the result that epoxy/SB blends show similar SAXS scattering patterns as neat SB. In the present system, although a two-phase morphology was also obtained, the morphology of the block ionomers was changed with the

side in Figure 9a2,b2,c2 show the SEM images of the etched specimens of these blends, displaying characteristics of a macroscopically phase-separated structure. For the epoxy/SEBS blend, the SEBS domains with irregular shapes and varied sizes are dispersed in a continuous cured epoxy matrix (Figure 9a1). The size of the phase structure is observed to be of the order of 5−10 μm. Vacant holes (dark areas) with irregular shapes are clearly seen in the SEM image of the etched specimen (Figure 9a2). With increasing sulfonation degree of SSEBS, the size of the heterogeneous phase structure in the epoxy/SSEBS blend decreases. For epoxy/0.9SSEBS blend, a micrometer scale phase-separated morphology is also observed (see Figure 9b1). The SEM image taken from the etched fracture surface (Figure 9b 2) shows vacant holes (dark areas) with irregular shapes but smaller sizes (∼1−3 μm) in the epoxy matrix. It is noted that small particles of cured epoxy are also visible inside the holes in Figure 9a2,b2. When SSEBS with higher sulfonation degree (e.g., 5.8SSEBS) is used, the block ionomer domain size further decreases and a relatively flatter surface can be observed (Figure 9c). It is also noted that the difference in the surface morphology for unetched and etched surfaces becomes less distinguishable. This may be due to the different extents of phase separation depending on the miscibility between the components. With increasing degree of sulfonation of SSEBS, more sulfonic acid groups in the SPS block are available to interact with epoxy network, giving more SPS blocks penetrating into the epoxy network. It is hence more difficult to remove the SSEBS phase. For the same reason, a phase structure with a smaller size exists in epoxy blends containing SSEBS with a higher sulfonation degree. The morphologies of epoxy blends having SSEBS of different sulfonation degree were further studied by SAXS. To have a better understanding of the formation of the morphology in 8198

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Figure 9. SEM images of cryogenic fractured surfaces of epoxy blends containing 10 wt % of (a) SEBS, (b) 0.9SSEBS, and (c) 5.8SSEBS. Note that the left-hand side images are taken from unetched surfaces whereas the right-hand side images are from surfaces etched by THF for 24 h.

which interactions of SSEBS with epoxy are sufficiently favorable that microscopic phase-separated structures exist in the epoxy blend. Otherwise, macroscopic phase-separated structures are formed. Mechanical and Fracture Toughness Properties of Epoxy Blends Containing SSEBS of Different Degree of Sulfonation. The effects of sulfonation degree of SSEBS on the mechanical properties of the epoxy blends were investigated. Table 2 summarizes the tensile properties (i.e., Young’s modulus, tensile strength, and elongation at break) and fracture toughness (KIC and GIC). Compared with neat epoxy, all epoxy/SSEBS blends exhibit lower modulus and tensile strength but better fracture toughness. Figure 11 shows the variations of KIC and GIC as a function of sulfonation degree of the used block ionomer. The sulfonation degree at 0 mol % refers to the parent SEBS. It can be clearly seen that the epoxy/SSEBS blends show various degree of improvement on fracture toughness depending on the sulfonation degree. The toughness of the epoxy blends increases with increasing sulfonation degree of SSEBS up to 5.8 mol % and decreases thereafter. For example, Table 2 shows that GIC for

incorporation of epoxy. Possible interactions between sulfonic acid groups with epoxy network may account for this change. For epoxy blends containing SSEBS with a sulfonation degree of 10.8 mol % or higher, multiple scattering peaks can be observed, indicating nanostructures are present in these blends. Compared with the aforementioned 90/10 epoxy/ 21.9SSEBS blend, broader and weaker scattering peaks can be found in the SAXS profile of the epoxy/10.8SSEBS blend. Further, the long spacing increases from 26.3 to 33.9 nm. SAXS data of epoxy/10.8SSEBS blend were analyzed using the core− shell model mentioned earlier. Results (not shown here) indicate that this model also fits the SAXS data very well. Structural parameters, Rc, Ts and Rhs, derived from SAXS analysis are 5.21, 3.40, and 13.17 nm, respectively, compared to 2.43, 3.16, and 12.55 nm for the 90/10 epoxy/21.9SSEBS blend. Therefore, smaller microdomains are formed with increasing sulfonation degree of SSEBS, probably caused by the increased content of the sulfonic acid groups in the SPS block, leading to enhanced miscibility with the epoxy matrix. From the above discussions, it can be concluded that there is a critical sulfonation degree (10.8 mol % in this study) beyond 8199

dx.doi.org/10.1021/ma401478t | Macromolecules 2013, 46, 8190−8202

Macromolecules

Article

Figure 10. SAXS profiles of (a) SSEBS of different sulfonation degree and (b) corresponding epoxy blends. The scattering profiles were shifted vertically for clarity. The dashed lines in (b) indicate macroscopically phase-separated epoxy system with microphase-separated SSEBS while the solid lines demonstrate microscopically phase-separated epoxy system with spherical microdomains.

Table 2. Mechanical Properties of Epoxy Blends Containing SSEBS of Various Sulfonation Degree epoxy blends with 10 wt % neat epoxy 21.9SSEBS 10.8SSEBS 5.8SSEBS 0.9SSEBS SEBS a

Young’s modulus (GPa) 2.48 2.39 2.32 2.33 2.22 2.25

± ± ± ± ± ±

tensile strength (MPa)

0.05 0.01 0.09 0.07 0. 09 0.02

62.78 47.36 44.74 43.96 40.50 39.98

± ± ± ± ± ±

1.6 2.1 1.5 2.3 1.9 1.7

elongation at break (%) 3.69 6.85 6.12 5.57 5.01 4.46

± ± ± ± ± ±

0.24 0.20 0.18 0.16 0.22 0.25

KIC (MPa·m1/2) 0.77 1.39 1.58 1.65 1.31 1.01

± ± ± ± ± ±

0.02 0.01 0.01 0.02 0.01 0.03

GIC (kJ/m2)

GIC/GIC,ma

± ± ± ± ± ±

1.00 3.52 4.52 4.90 3.24 1.95

0.21 0.74 0.95 1.03 0.68 0.41

0.03 0.01 0.02 0.02 0.03 0.01

GIC,m is plane strain energy release rate for neat epoxy.

The dependence of mechanical properties of the epoxy/ SSEBS blends on the sulfonation degree reflects their correlations with the phase-separated morphology and structure. First, in the opaque system with macroscopic phaseseparated morphology, increasing the sulfonation degree of SSEBS from 0 to 0.9 mol % gives epoxy blends with micrometer-sized phase domains x (1.0 < x