View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/C4RA12261E (Paper) RSC Adv., 2015, 5, 8471-8478

Effect of SiO2 nanoparticles on the reaction-induced phase separation in dynamically asymmetric epoxy/PEI blends

Weizhen Li*, Zonglian Xia, Ao Li, Yang Ling, Baoyu Wang and Wenjun Gan*
Department of Macromolecular Materials and Engineering, College of Chemistry and Chemical Engineering, Shanghai University of Engineering Science, 201620, Shanghai, China. E-mail: liweizhen@sues.edu.cn; wjgan@sues.edu.cn

Received 13th October 2014 , Accepted 24th December 2014

First published on 24th December 2014

Abstract

Hydrophilic silica nanoparticles were introduced into the epoxy/polyetherimide (epoxy/PEI) binary system to study their effect on the reaction induced phase separation (RIPS) by differential scanning calorimetry (DSC), optical microscopy (OM), and time-resolved light scattering (TRLS). Depending on the specific interaction between the silica surface and the epoxy, the silica nanoparticles selectively distribute in the epoxy-rich domain, leading to a slow down of the diffusion of the epoxy molecules. And the coarsening mechanism was forced towards the diffusion-controlled regime, which enhanced the viscoelastic effect and produced a more dynamic asymmetric epoxy/PEI system. Based on this, the final morphology was stopped at an earlier stage of an inverted phase structure (isolated epoxy-rich droplet and PEI-rich matrix). The silica nanoparticles showed a critical impact on the balance of the diffusion and geometrical growth of epoxy molecules. Further, the activity energy of the curing reaction and the phase separation temperatures were decreased by existence of the silica nanoparticles.

1. Introduction

Epoxy resins have been widely used in many industrial areas in combination with polymeric modifiers to enhance their toughness.1 Typical polymeric modifiers are thermoplastics, including polyimide,2 polyethersulfone, polysulfone etc.3–5 A valuable two phase structure6 can be obtained by the reaction-induced phase separation (RIPS) of the initially homogeneous epoxy/thermoplastic blends. Since the mechanical properties of the materials are determined by their final morphologies, much work has been focused on phase separation and morphology control. Generally, a phase inversion structure or a co-continuous phase7 is a determining morphology to help the achievement of the toughness, due to an efficient combination of the stiffness of the thermosetting and the toughness of thermoplastics. Owing to the complex phase behavior during the RIPS of epoxy/thermoplastics blends, the final morphology is generally controlled by the phase separation kinetics. Thus, the study on the phase-separation kinetics and the control of the morphology evolution has been focused by many researchers.

Recently, literatures reported that the phase behavior of binary polymer mixtures can be significantly altered by nano- or microfillers, such as fibers, silica particles and nanoclay.8,9 The addition of nanofillers leads to significant effects on the system with UCST or LCST, such as increasing or decreasing the temperatures of phase separation, modifying the shape of the phase diagram or changing the kinetics of phase separation. Lipatov et al. found the phase boundary can be shifted either up or down depending on the particles concentration with addition of silica nanoparticles in the chlorinated poly(ethylene)/poly(ethylene-co-vinyl acetate) (EVA) blends.10,11 Ivan Kelnar et al. showed that the addition of the layered silicate in polyamide/polystyrene blends leaded to remarkable refinements of both particulate and co-continuous structures.12

Generally, the final phase separation morphology of an epoxy/thermoplastic binary system is determined by the competition between cure-reaction rate and phase separation rate.13 Nanofillers could tailor the phase behavior by altering the diffusion or the cure-reaction rate of the components. Guijun Yu et al. reported that an enhancement effect of methylenedianiline-modified graphene oxide (GO–MDA) on the complex viscosity and cure-reaction rate of the diglycidyl ether of a bisphenol A/polyetherimide (DGEBA/PEI) binary system, therefore arresting the final morphology of the composites at an earlier stage of co-continuous structure.14 Mao Peng et al. studied the effect of organic modified layered silicates (OLS) on the RIPS of epoxy and poly(ether imide) (PEI). The onset of phase separation and the gelation or vitrification time were greatly brought forward and the periodic distance of phase-separated structure was reduced when OLS was incorporated.15,16 Since that, nanofillers attract much attention to regulate the morphology of the complex mixture, and control the macroscopic properties of the polymer blends. It is essential to obtain a co-continuous or phase inversion morphology for toughening purposes in material designing and property control with small amount of modifier.

In this study, we selected a ternary mixture of DGEBA, PEI, and hydrophobic silica particles as a model system to evaluate the influence of silica nanoparticles on the process of phase separation and the final morphology of the thermosetting/thermoplastic blends with optical microscopy (OM), time-resolved light scattering (TRLS), and scanning electron microscopy (SEM). Besides, the curing behavior of the epoxy/PEI system was studied by differential scanning calorimetry (DSC).

2. Experimental

2.1 Materials

Diglycidyl ether of bisphenol A (DGEBA) with epoxide equivalent weight of 171–175 was purchased from Dow Chemical Company, China. The curing agent was methyl tetrahydrophthalic anhydride (Me-THPA), and the accelerator N,N-dimethyl benzyl amine (DMBA) was also added in the blend. The polyetherimide, PEI Ultem 1040, was supplied by SABIC's Innovative Plastics. Hydrophilic colloidal silica nanoparticles with a diameter of 10–15 nm, dispersing in MEK with approx. 40 wt% silica, was provided by Nissan Chemical (USA) and a surface covered with methyl and hydroxyl groups, referred to as MEK–ST.17

2.2 Preparation of DGEBA/PEI/MEK–ST silica blends

DGEBA/MEK–ST silica blends were blended and followed by the ultrasonic. PEI was added to the DGEBA/silica according to the formulation in Table 1. Afterwards, the obtained blend was dissolved in chloroform and then stirred at 60 °C to obtain a homogeneous mixture. Next, the solvent was removed, the curing agent Me-THPA was added and stirred for 10 minutes at 110 °C, followed by pouring them into moulds and cured at 150 °C isothermally for 5 h. The remaining pre-curing mixture was immediately cooled down and stored in fridge at −18 °C to avoid further curing for kinetics studies.
Table 1 Formulation of epoxy/PEI/MEK–ST silica
Sample DGEBA (pbw) PEI (pbw) MEK–SiO2 (pbw) Me-THPA (pbw)
M0025 100 25 0 80
M0125 100 25 1.0 80
M0225 100 25 2.0 80

2.3 Characterizations

2.3.1 Scanning electron microscopy (SEM). Each isothermally cured sample was cooled and fractured in liquid nitrogen at atmospheric pressure. The clean samples were sputter-coated with platinum prior to SEM (Hitachi S-3400, Japan) examination.
2.3.2 Optical microscopy (OM). A few milligrams of the epoxy/PEI or epoxy/PEI/silica blends sandwiched between two glass slides were cured in a hot stage and meanwhile observed by optical microscope SBM-80I (Shanghai WeiTu Optics & Electron Technology Co., Ltd, China). The images were taken by a digital camera controller. Cloud points Tcp were estimated by the OM observation of a thin layer of the sample and were selected as an average value from 3–4 measurements.
2.3.3 Differential scanning calorimetry (DSC). DSC measurements were performed via Linseis PT-10 (Germany). Dry nitrogen was used as purge gas and samples of 4–5 mg were analyzed. Dynamic DSC was done at 2.5, 5, 10, 15, and 20 °C min−1 from 30 to 300 °C.
2.3.4 Time-resolved light scattering (TRLS). The phase separation process during the isothermal curing reaction was observed in situ on the self-made time-resolved light scattering (TRLS) instrument with controllable hot chamber. The samples for TRLS observation were sandwiched between two glass cover slides and pressed into thin film.

3. Results and discussion

3.1 Influence of silica nanoparticles on the cure kinetics

Hydroxyl groups are known to have a catalytic effect on the epoxy ring-opening reactions, acting as hydrogen-bond donor molecules.18 Therefore, the influence of silica nanoparticle on the curing reaction in the binary PEI/DGEBA blend could be interested and was studied by DSC. At a certain heating rate, the peak temperature (Tp, where the curing-reaction rate is maximal) of the exothermic curve was attained as shown in Fig. 1 (10 °C min−1) and summarized in Table 2. Tp was shifted to lower temperature with the addition of silica nanoparticle, implicating that silica nanoparticle may accelerate the curing-reaction of the composites, due to the presence of Si–OH of MEK–ST silica.17
 
image file: c4ra12261e-t1.tif(1)
image file: c4ra12261e-f1.tif
Fig. 1 Dynamic DSC thermograms of epoxy/PEI (100[thin space (1/6-em)]:[thin space (1/6-em)]25), epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]1) and epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]2) blends at a heating rate of 10 K min−1.
Table 2 Tp and Ea values of the blends obtained from the DSC measurement at different heating rates
Tp (K) 2.5 K min−1 5 K min−1 10 K min−1 15 K min−1 20 K min−1 Ea (kJ mol−1)
M0025 426.6 437.1 450.2 460.1 465.8 78.8
M0125 417.6 430.8 444.8 456.3 463.7 65.2
M0225 415.8 428.3 443.3 454.8 461.7 64.2

More information can be obtained from the calculation of the activation energy. According to Kissinger equation,19 a plot of ln(β/Tp2) versus 1000/Tp allows the determination of activation energy, which is shown in Fig. 2. The activation energy calculated for epoxy/PEI (100[thin space (1/6-em)]:[thin space (1/6-em)]25), epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]1) and epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]2) blends were 78.8 kJ mol−1, 65.2 kJ mol−1, 64.2 kJ mol−1, respectively. The decrease of the activation energy with the incorporation of silica nanoparticles could be explained by the status that the hydroxyl groups on the surface of the particle participate in the curing reaction and act as an epoxy ring-opening accelerator (quickly open the ring of the anhydrides or the epoxy ring), therefore bringing down the energy barrier of curing reaction. This result is consistent with the work of Alice Mija et al. that silica particles act as a catalytic in the epoxy ring-opening reaction, although the curing agent used in that study was m-phenylenediamine.18 Scheme 1 shows the proposed mechanism for the reaction between epoxy and anhydrides in the presence of silica nanoparticles.


image file: c4ra12261e-f2.tif
Fig. 2 Determination of the Ea values for epoxy/PEI (100[thin space (1/6-em)]:[thin space (1/6-em)]25), epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]1) and epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]2) blends by using Kissinger's equation.

image file: c4ra12261e-s1.tif
Scheme 1 Proposed mechanisms for the reaction between epoxy and anhydrides in the presence of silica nanoparticles.

Besides, as shown in Table 2, similar Tp and Ea were observed in the epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]1) and epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]2) systems. The extent of silica content does not show a further coincident decrease of the peak temperature and activation energy of curing reaction. This may indicate that the aggregation of the silica particles occurred at higher loading content and only part of the Si–OH group work as an accelerator of the curing reaction.

3.2 Effect of silica nanoparticles on the phase diagrams of DGEBA/PEI blends

The cloud-point temperatures (Tcp) of epoxy/PEI with and without 2 wt% MEK–ST silica nanoparticles were evaluated by OM and presented in Fig. 3. As can be seen, both the non-filled and filled blends show typical UCST behavior. That is to say, the one-phase region is located above the curves. The increase of the PEI content leads to an increase then decrease in the temperature required to generate a one-phase state. Introduction of a certain amount of MEK–ST silica nanoparticles induces an obvious decrease of the cloud-point temperatures, which means the temperatures of phase separation of the filled blends are lower than the non-filled blend.
image file: c4ra12261e-f3.tif
Fig. 3 The cloud-point temperatures (Tcp) of epoxy/PEI blends with and without 2 wt% MEK–ST silica.

In other words, the thermal stability of the epoxy/PEI systems is increased with the addition of MEK–ST silica. The changes of phase separation temperatures may be explained as followed. It is known that, there are border layers on the surface of the fillers, which differs with the polymer bulk. In this study, the hydroxyl groups on the silica surface induce the formation of the border layer. As Lipatov Y. S. et al.10,11 declared that the border layer will differ from the bulk in relation to the molecular weight distribution, due to preferential adsorption of the polymer fraction of higher molecular weight. In this case, DSC results show that the activation energy was decreased with the addition of MEK–ST silica because of the hydroxyl groups on the surface of the particle.17 As explained, the hydroxyl groups of silica surface self-participate the epoxy ring-opening reaction and accelerate the curing rate, therefore leading to a selective adsorption of the epoxy fraction of higher molecular weight and diminishing the molecular mobility of epoxy. As a result, the phase separation of the bulk occurs at a lower temperature in the MEK–ST silica-filled blends, meaning that the thermodynamics interaction of the bulk was changed. To further study the phase separation of the binary blends, the evolution of phase separation was studied by OM.

3.3 Evolution of phase morphologies observed by OM

The effect of nanoparticles on the phase evolution of the dynamic asymmetry blends (epoxy/PEI) was investigated by OM firstly. The epoxy/PEI (100/25) blend at near-critical composition was studied and a typical evolution process of the RIPS was shown in Fig. 4. At the beginning of the phase separation (Fig. 4a), a microcontinuous phase structure was observed, which present as the near-critical composition via spinodal decomposition (SD) mechanism. As the cure proceeding, the molecular weight of epoxy precursors increased, followed by diffusing out from the PEI-rich phase (viscous phase) and coarsening rapidly. At time around 200 s, the continuous epoxy-rich phase aggregated and appeared as irregular shapes in the PEI-rich matrix. Fig. 4e shows a phase inversion structure (PEI-rich phase as matrix) and the final phase structure was fixed after 250 s. It has been demonstrated as the dynamic asymmetry between thermoplastics and thermoset precursors, due to their large differences in molecular mobility and molecular weight. During cure processing, the viscoelasticity of the slow dynamic phase (PEI-rich phase) increased when the epoxy precursor diffused out from the PEI-rich phase and eventually behaved as an elastic matrix. Meanwhile, the fast dynamic phase (epoxy-rich phase) started to coarsen and the elastic force balance dominated the morphology instead of the interface tension, leading to a phase inversion structure fixed at the earlier time. Fig. 4e shows the final phase structure containing irregular worm-like epoxy-rich domain (broken cocontinuous structure) dispersed in PEI-rich matrix. It should be noted that secondary phase separation occurs in both epoxy-rich and PEI-rich domain, which shows many small droplets appearing in the both domains at the late stage of phase separation in Fig. 4d, which coincides well with the state of viscoelastic phase separation proposed by Tanaka.20
image file: c4ra12261e-f4.tif
Fig. 4 Evolution of phase morphologies of epoxy/PEI (100[thin space (1/6-em)]:[thin space (1/6-em)]25) blends at 110 °C.

As mentioned, the studied epoxy/PEI blend is a dynamic asymmetric system. During the cure process, large differences occur in mobility between epoxy molecular and PEI. At the late stage of phase separation, the molar mass of epoxy increases, therefore the diffusion of epoxy molecules is not able to follow the growth of the size of the epoxy-rich domain. As a result, the secondary phase separation occurred. It is shown that the small droplets of epoxy-rich phase appear in the PEI-rich phase, or in the opposite way, seen in Fig. 4e. And as the same reason, the slow dynamic phase (PEI-rich phase) was fixed as a phase inversion structure despite of its lower proportion.

The evolution of phase structure was influenced by the addition of MEK–ST silica nanoparticles. It showed a similar microcontinuous phase structure (Fig. 5a) at the beginning of phase separation compared with the neat epoxy/PEI blend. However, during the curing reaction of epoxy molecules, the microcontinuous epoxy-rich phase coarsened and aggregated into anisotropic-shape domains quickly, which was shown as isolating dispersed epoxy-rich droplets in the PEI-rich matrix at phase separation time of 100 s (Fig. 5c). As followed, the size of the epoxy-rich domains increased till the chemical gelation of the epoxy. Compared to the neat epoxy/PEI blend, the extent of coarsening was suppressed by the MEK–ST silica nanoparticles (the microcontinuous phase was broken earlier and more isolated epoxy-rich droplets were formed).


image file: c4ra12261e-f5.tif
Fig. 5 Evolution of phase morphologies of epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]2) at 110 °C.

For the dynamic asymmetric system (epoxy/PEI), Tanaka proposed that the effect of viscoelastic would play an important role in the phase separation and the diffusion of the fast dynamic phase (epoxy-rich phase) was prevented by the elastic force of the slow dynamic phase (PEI-rich phase). At the moment, dynamic asymmetric is the main driving force of the secondary phase separation.21 In this study, the hydroxyl groups of the silica nanoparticles participated the reaction between the anhydrides and epoxy molecules and restricted the molecular mobility of epoxy molecules, therefore slowing down the diffusion of the epoxy, which will be further demonstrated in TRLS experiments. Since that, the diffusion of epoxy molecules (fast dynamic phase) was even unable to follow the growth of the size of the epoxy-rich domain. Afterwards, the microcontinuous phase had to break into isolated droplets earlier, which grew slowly till the chemical gelation. As a result, the final phase separation morphology stopped at an earlier stage of inverted phase structure. Moreover, the secondary phase separations occurred as the same as in the neat PEI/epoxy blend.

3.4 SEM and TEM morphologies

3.4.1 The distribution of silica particles in the epoxy/PEI blends. In the majority of the blends, the distribution of nanoparticles is uneven between the different polymer phases. TEM was used to study the distribution of 2 wt% silica particles in epoxy/PEI (100/20) blends, in which the epoxy-rich phase appears as the matrix and the PEI-rich phase appears as droplet structure.

Fig. 6 shows the PEI-rich droplet (black domain) dispersed in the epoxy-rich matrix (grey domain). And it seems that the MEK–ST silica nanoparticles prefer to be dispersed in the epoxy matrix, although showing the silica aggregation. This preferential location of silica nanoparticles is the result of that the hydroxyl group catalyzed curing reaction on the silica surface. The strong hydrogen bonding interaction between the hydroxyl groups on MEK–ST silica nanoparticles and the carbonyl groups of the Me-THPA-cured epoxy could also be the reason.


image file: c4ra12261e-f6.tif
Fig. 6 TEM images of epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]20[thin space (1/6-em)]:[thin space (1/6-em)]2) blends cured at 150 °C for 5 hours.
3.4.2 SEM morphologies of fracture surface. SEM micrographs of fracture surface of the cured samples for DGEBA/PEI/MEK–ST silica (100/25/2) blend were shown in Fig. 7. Obviously, the secondary phase separation occurs in both PEI-rich phases (Fig. 7(a)i) and epoxy-rich phases (Fig. 7(a)ii). This is to say, lots of epoxy droplets are dispersed in the continuous PEI-rich phase (Fig. 7(a)i) and many small PEI particles are distributed in primary epoxy-rich domain (Fig. 7(a)ii). This result supports the previous OM results.
image file: c4ra12261e-f7.tif
Fig. 7 (a) SEM images of the fracture surface of epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]2) blends cured at 150 °C for 5 hours ((i): PEI-rich phase, (ii): epoxy-rich phase), (b) magnification of PEI-rich phase ((a)-(i)), (c) magnification of epoxy-rich phase ((a)-(ii)).

3.5 Phase separation dynamics by TRLS

To better understand the structure development of the reaction-induced phase separation, TRLS were performed on the neat epoxy/PEI and epoxy/PEI/SiO2 blends during isothermal curing reaction. The scattering profiles are shown in Fig. 8 and 9. Fig. 8 indicates the scattering profiles of neat epoxy/PEI (100/25) with curing temperature at 110 °C. The peak scattering vector, corresponding to the wave number of concentration fluctuation, is defined by image file: c4ra12261e-t2.tif, where λ and θ are the wavelength of incident beam and the peak scattering angle, respectively.22,23 The reciprocal of the scattering vector qm is assigned to the periodic distance Λm of the dispersed particles. The smaller the characteristic vector qm is, the larger the periodic distance Λm of the pinned structure will be image file: c4ra12261e-t3.tif.
image file: c4ra12261e-f8.tif
Fig. 8 Phase structure evolution with time of epoxy/PEI (100[thin space (1/6-em)]:[thin space (1/6-em)]25) by TRLS cured at 110 °C.

image file: c4ra12261e-f9.tif
Fig. 9 The scattering profiles of epoxy/PEI with 1 wt% MEK–ST silica (a and b) and 2 wt% (c and d) curing at 110 °C.

As shown in Fig. 8, at the beginning of the curing process, qm shifted to a lower value and the light intensity increases, indicating that the reaction induced phase separation was initiated and the periodic structure size Λm grew. When the light intensity got to the maximum, a second peak at a higher scattering vector value appeared at about 200 s, which implied the occurrence of secondary phase separation. Afterwards, the scattering intensity of qm stopped increasing after 150 s and then began to decrease slightly before it finally became invariant, which resulted from the decrease in the difference of the reflection indices of the two phases.24 The profiles after 300 s kept similar since the final phase structure was fixed till the occurrence of chemical gelation. These results are in good agreement with the OM observations.

Fig. 9 showed the scattering profiles of epoxy/PEI with 1 wt% MEK–ST silica and 2 wt% curing at 110 °C. For the silica-filled epoxy/PEI (1 wt% MEK–ST silica), at the early stage of phase separation, it was same as the neat blend that qm shifted to lower value with a quick increase of light intensity. However, when the light intensity got to the maximum, a second peak at high scattering vector value appeared at the same time (46 s), which is much faster than the neat epoxy/PEI blend. Corresponding to the OM results (Fig. 5c and d), the secondary phase separation occurred after 50 s, showing that the number of the epoxy droplets gradually increased in the PEI-rich domain. In Fig. 9b, the qm got to a higher value with a quick increase of light intensity, meaning that the appearance of small particles made the average phase domain size decrease. When the content of MEK–ST silica was increased to 2 wt%, the morphology evolution was similar to that of epoxy/PEI/SiO2 (100/15/1) blend. But the onset time of the secondary phase separation was earlier (25 s) and the average periodic distance Λm was much smaller.

It is generally known that the increase of the domain size of the separated phase is mainly caused by the hydrodynamic forces and diffusion. And the phase growth can be analyzed with a power-law scaling equation. The position of the scattering ring, qmax(t), as a function of time, can be linked by the following equations:

 
qmax(t) ∝ t−α (2)
where α is a scaling coefficient. For binary polymer blends, quenched at the critical concentration, α varies from 0.33, with coarsening controlled by diffusion (viscous force effect), to 1.0, with coarsening controlled by hydrodynamic forces.25,26

With regards to the TRLS results, the plots of the peak scattering vector, qmax, versus time t was shown in Fig. 10. The qmax follows the power law and the slope of log[thin space (1/6-em)]qmax versus log[thin space (1/6-em)]t (power-law scaling coefficient α) are 0.77, 0.62, 0.47 for epoxy/PEI (100/25) without and with 1%, 2% MEK–SiO2, respectively. The scaling coefficient α varies from 0.47 to 0.77, which means that all systems are between diffusion-controlled and hydrodynamics-controlled. The α decreases significantly when the blend is filled with a certain amount of MEK–SiO2 nanoparticles, implying that the nanoparticles force coarsening mechanism towards the diffusion-controlled regime. That is to say, the migration of MEK–SiO2 nanoparticles to the epoxy-rich phase can lead to a concomitant reduction of the mobility of epoxy and then indeed slow down the diffusion of the fast dynamic phase (epoxy-rich phase).


image file: c4ra12261e-f10.tif
Fig. 10 The changes of peak scattering vector qm, versus time t: epoxy/PEI (100[thin space (1/6-em)]:[thin space (1/6-em)]25), epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]1) and epoxy/PEI/SiO2 (100[thin space (1/6-em)]:[thin space (1/6-em)]25[thin space (1/6-em)]:[thin space (1/6-em)]2) blends at 110 °C.

Based on the above results, the significant effect of the MEK–SiO2 can be found in the epoxy/thermoplastic blend, which leads to a more complicated phase separation. The TRLS results show the secondary phase separations occur earlier in this silica-filled system compared to the neat blend. As discussed before, the main driving force of the secondary phase separation is the dynamic asymmetric of thermosetting/thermoplastic system, which is forced to more asymmetric because of the reduction of the diffusion of the fast dynamic phase. For that reason, the more diffusion-controlled regime makes the secondary phase separations occur earlier and the final morphology will be fixed at earlier time, showing an inverted structure (the component with high weight proportion as dispersed phase and the component with low weight proportion as matrix).

4. Conclusions

The effect of MEK–ST silica nanoparticles on the reaction-induced phase separation in a dynamically asymmetric thermosetting/thermoplastic (epoxy/PEI) blend was investigated. It was shown that silica nanoparticles prefer distributing in the epoxy-rich phase due to the existence of hydroxyl groups on the surface of silica. Based on DSC results, the decrease of the activation energy with the addition of silica can be explained by the existence of hydroxyl groups on the surface of hydrophilic silica, which accelerated ring-opening reaction of the epoxy and anhydrides. Further, the introduction of MEK–ST silica nanoparticles induces an obvious decrease of the temperatures of phase separation, which means the thermal stability of the epoxy/PEI systems is increased with the MEK–ST silica. This fact could be related to the formation of the surface layers of the silica nanoparticles, which attract the higher molecular weight fraction of epoxy and diminish the molecular mobility of epoxy. The phase evolution was investigated via optical microscopy and time-resolved light scattering. MEK–ST silica nanoparticles were found to suppress the diffusion of epoxy-rich phase (the fast dynamic phase). Therefore, the final morphology stops at an earlier stage of inverted phase structure and the secondary phase separation occurs earlier. Silica nanoparticles show as an important factor to obtain a phase inversion morphology for toughening purposes in epoxy/PEI blend with small amount of PEI modifier.

Acknowledgements

The authors wish to thank the Shanghai Municipal Education Commission (Overseas Visiting Scholar Project 20120407); Shanghai Young Teachers' Training-funded Projects (ZZGJD13018); Shanghai University of Engineering Science Developing founding (grant 2011XZ04), start-up project funding (grant 0501-13-018) and Interdisciplinary Subject Construction (grant 2012SCX005).

Notes and references

  1. P. Poornima Vijayan, D. Puglia, H. J. Maria, J. M. Kenny and S. Thomas, RSC Adv., 2013, 3, 24634 RSC.
  2. X. Xiao, S. Lu, B. Qi, C. Zeng, Z. Yuan and J. Yu, RSC Adv., 2014, 4, 14928 RSC.
  3. F. H. Chen, T. C. Sun, S. Hong, K. Meng and C. C. Han, Macromolecules, 2008, 41, 7469 CrossRef CAS.
  4. R. Rajasekaran, M. Alagar and C. Karikal Chozhan, eXPRESS Polym. Lett., 2008, 2, 339 CrossRef CAS.
  5. Y. Zhang, F. H. Chen, W. C. Shi, Y. R. Liang and C. C. Han, Polymer, 2010, 51, 6030 CrossRef CAS PubMed.
  6. C. Li, N. J. Kang, S. D. Labrandero, J. T. Wan, C. González and D. Y. Wang, Ind. Eng. Chem. Res., 2014, 53, 1040 CrossRef CAS.
  7. M. Giannotti, I. Mondragon, M. J. Galante and P. A. Oyanguren, Polym. Int., 2005, 54, 897 CrossRef CAS.
  8. L. Wang, X. X. Shui, X. Zheng, J. C. You and Y. J. Li, Compos. Sci. Technol., 2014, 93, 46 CrossRef CAS PubMed.
  9. P. P. Vijayan, D. Puglia, J. M. Kenny and S. Thomas, Soft Matter, 2013, 9, 2899 RSC.
  10. Y. S. Lipatov, J. Macromol. Sci., Part B: Phys., 2006, 45, 871 CrossRef CAS.
  11. Y. S. Lipatov, A. E. Nesterov, T. D. Ignatova and D. A. Nesterov, Polymer, 2002, 43, 875 CrossRef CAS.
  12. I. Kelnar, J. Rotrekl, J. Kotek and L. Kaprálková, Polym. Int., 2008, 57, 1281 CrossRef CAS.
  13. W. H. Jo and M. B. Ko, Macromolecules, 1994, 27, 7815 CrossRef CAS.
  14. G. J. Yu and P. Y. Wu, Polym. Chem., 2014, 5, 96 RSC.
  15. P. G. Mao, H. B. Li, L. J. Wu, Y. Chen, Q. Zheng and W. F. Gu, Polymer, 2005, 46, 7612 CrossRef PubMed.
  16. A. E. Nesterov, Y. S. Lipatov, V. V. Horichko and O. T. Gritsenko, Polymer, 1992, 33, 619 CrossRef CAS.
  17. K. Yoshitake and T. Yokoyama, US Pat. 60025455A, Nissan Chemical Inc., 2000.
  18. C. Alzina, N. Sbirrazzuoli and A. Mija, J. Phys. Chem. C, 2011, 115, 22789 CAS.
  19. H. E. Kissinger, Anal. Chem., 1957, 29, 1702 CrossRef CAS.
  20. T. Araki and H. Tanaka, Macromolecules, 2001, 34, 1953 CrossRef CAS.
  21. P. Jyotishkumar, J. Pionteck, P. Moldenaers and S. Thomas, J. Appl. Polym. Sci., 2013, 127, 3159 CrossRef CAS.
  22. W. J. Gan, Y. F. Yu, M. H. Wang, Q. Tao and S. J. Li, Macromol. Rapid Commun., 2003, 24, 952 CrossRef CAS.
  23. M. H. Wang, Y. F. Yu, X. Wu and S. J. Li, Polymer, 2004, 45, 1253 CrossRef CAS PubMed.
  24. N. Biolley, T. Pascal and B. Sillon, Polymer, 1994, 35, 558 CrossRef CAS.
  25. H. Wang and R. J. Composto, Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top., 2000, 61, 1659 CrossRef CAS.
  26. H. Chen and A. Chakrabarti, J. Chem. Phys., 1998, 108, 6006 CrossRef CAS PubMed.
This journal is © The Royal Society of Chemistry 2015
Click here to see how this site uses Cookies. View our privacy policy here.