Toughening of an epoxy thermoset with poly[styrene-alt-(maleic acid)]-block-polystyrene-block-poly(n-butyl acrylate) reactive core–shell particles

Abstract

A series of tailored reactive core–shell particles (RCSPs), structurally core-crosslinked reactive block copolymer micelles, were synthesized via reversible addition–fragmentation chain transfer (RAFT) miniemulsion polymerization mediated by an amphiphilic macro-RAFT agent. The core–shell structures of RCSPs were confirmed by SEC, DSC, and TEM observations. The particle sizes of RCSPs were all in unimodal distributions according to DLS measurement. The designed RCSPs were then blended with epoxy resin. RCSPs with 8 wt% more reactive segments were found to be well-dispersed in the cured epoxy matrix in the scale of pre-existing core–shell particles. By considering the similar RCSP particle size and RCSP volume fraction in the epoxy matrix, it was judged that the reactive segment of RCSPs played a key role in toughness improvement. It was noted that the epoxy blends are significantly toughened by RCSPs in terms of the measurement of the critical stress field intensity factor (K_IC).

---

1 Introduction

Epoxy polymers are widely used for the matrices of fibre-reinforced composite materials and adhesives. Epoxy toughening is a topic with wide commercial applications, arousing a lot of attention and research.^1–4 Core–shell particles (CSP) exhibited an excellent toughening effect on epoxy blends.^5–10 It is convenient to produce CSPs by grafting a compatible shell to a rubbery or rigid core via multistep soap-less emulsion polymerization.^11–14 Lin et al. focused on increasing the miscibility between CSP and the epoxy matrix, where glycidyl methacrylate (GMA) was copolymerized into the glassy shell of CSP.^15,16 The impact resistance of UV-cured epoxy systems were enhanced by crosslinked core/shell PBA/PMMA-PGMA particles.¹⁷ A theoretical model was proposed and used to predict the fracture toughness increment of epoxy modified by CSP with PMMA shell.⁵ However, PMMA shell always require specific hardeners to avoid particle aggregation after curing of epoxy matrix.¹⁸ And CSPs containing glycidyl methacrylate embedded in shell is not able to covalently link to epoxy precursor, reducing its miscibility.^16,19 To overcome these difficulties, we propose a reactive core–shell particle composed of a rubbery core, a glassy shell, and a surface of reactive segments.

Analogue to core–shell structure of latex particles, the self-assembled block copolymer micelles in epoxy matrix have been proved to have considerable toughening effect by many researchers.^20–27 The recent advent in RAFT emulsion polymerization provides a good opportunity in synthesis of tailored block copolymers.^28,29 Traditionally, the core–shell structures of amphiphilic block copolymer latex particles generated in RAFT emulsion polymerization are inevitably destroyed when they are blended with epoxy resin. By core cross-linking, the morphologies of CSP would like to be persistent when transferred to epoxy precursor.³⁰ Therefore, reactive core–shell particles could like to be stable and well-dispersed after curing of epoxy thermosets without considering the self-assembly and reaction induced phase separation of block copolymers.^18,20,31,32 And poly[styrene-alt-(maleic anhydride)] (SMA) shell in reactive block copolymer core–shell particles are able to create covalent links between both DGEBA and amine hardener.³³ The obtained nanoparticle with significantly increased interfacial adhesion can induce the particle cavitation and plastic void growth, leading to toughness improvement.^21,34–36

In this study, a series of poly[styrene-alt-(maleic acid)]-block-polystyrene-block-poly(n-butyl acrylate) reactive block copolymer core–shell nanoparticle was synthesized via RAFT emulsion polymerization. The latex particles with core–shell structures were then core-crosslinked by EGDMA, as illustrated in Scheme 1. The structures of RCSPs were measured by GPC, DLS, DSC, and TEM observations. The reactive block copolymer core–shell nanoparticles were then redispersed in epoxy matrix at 10 wt% loading. The morphology of epoxy blends was verified by TEM and FESEM observations. Thermal properties of epoxy blends were investigated using dynamical mechanical thermal analysis. Fracture toughness of epoxy blends were evaluated by single edge notched three-point blending method. The epoxy blends modified by RCSPs with different reactive segments are compared in this paper. This research would like to provide a new application of core–shell particles prepared by RAFT polymerization for both industrial and commercial perspectives.

Scheme 1 Schematic illustration of the synthesis procedure of triblock copolymer and the formation of reactive core–shell particle.

2 Experimental

2.1 Materials

Styrene (St, monomer) was distilled under reduced pressure before use. n-Butyl acrylate (nBA, monomer) was washed with sodium hydroxide aqueous solution (5 wt%) to remove the inhibitor. Ethylene glycol dimethacrylate (EGDMA, monomer) was used to crosslink the core after the removal of inhibitor. Potassium persulfate (KPS, initiator), acetone and ammonia solution (25 wt% in water) were used without further purification. The procedure of preparing poly(styrene-alt-maleic anhydride) dithiocarbonate macro-RAFT agent was described in detail in our previous studies.³⁴ The epoxy resin used in this study was diglycidyl ether of bisphenol A (DGEBA) (E51, Shanghai Resin Production, epoxy value 0.51 mol/100 g). 4,4′-Diaminodiphenylmethane (DDM, 97%, Aldrich) was used as hardener. Amino hydrogen to epoxy stoichiometric ratio of one was selected for all the systems.

2.2 Preparation of RCSP

The synthesized RCSPs mediated by SMA-RAFT macro-RAFT agents with molecular weight of 2 kDa, 5 kDa, and 10 kDa are designated as R2CSP, R5CSP, and R10CSP, respectively. Scheme 1 presents the route to obtain the RCSP. 2 × 10⁻⁴ mol of the SMA-RAFT and 0.038 mol of styrene were dissolved in 0.07 mol of acetone. Then, 20 g of deionized water and 0.15 g of ammonia solution were added in the organic styrene/acetone solution under stirring. The obtained emulsion was ultra-sonicated for 15 min and then transferred to a flask. The resulting stable emulsion was deoxygenated by N₂ purge for 20 min. The initiator KPS (4 × 10⁻⁵ mol) dissolved in 5 g of water was added to start polymerization, then the temperature reached 70 °C. Later, 0.063 mol of n-butyl acrylate as the second monomer was added. After 90 min, 0.02 mol of EGDMA as the cross-link agent was added. The polymerization proceeded to complete monomer conversion. Samples were taken at regular time to analysis the conversion gravimetrically. The nonreactive core–shell particle, designated as R0CSP, was prepared via conventional polymerization process. 1-Phenylethyl phenyldithioacetate (PEPDTA) was used to mediate RAFT polymerization. Sodium dodecyl sulfate (SDS) was used as surfactant of the latex, and hexadecane was used as co-stabilizer of emulsion.

2.3 Preparation of RCSP modified epoxy blends

The RCSP and DGEBA mixture was heated to 90 °C under magnetic stirring until homogeneity was achieved. Then a stoichiometric amount of hardener DDM was added to the mixture, magnetic stirred, until the hardener was completely dissolved. Uniform and bubble-free plaques were obtained by pouring the mixture into a preheated mould and kept in a vacuum oven for 1 h. The blends were cured at 50 °C for 2 h, 100 °C for 3 h and 150 °C for 3 h in the oven and then slowly cooled down to room temperature. The 10 wt% R0CSP, R2CSP, R5CSP, and R10CSP modified epoxy blends are designated as R0EP, R2EP, R5EP, and R10EP, respectively.

2.4 Morphology of RCSP

Molecular weight (M_n) and molecular weight distribution (PDI) of RCSP (without crosslinking agent EGDMA) were measured by GPC (Waters 1525 Binary HPLC Pump, Waters 717 Autosampler, Waters 2414 Refractive Index Detector, Waters 2487 Dual λ Absorbance Detector for UV 311 signals). The samples were prepared following the same procedure as the core cross-linked ones with EGMDA replaced by nBA, then dried in an oven at 100 °C for 3 h and then dissolved in tetrahydrofuran (THF). The block copolymers, which were not crosslinked, were designated as R2CSP′, R5CSP′, and R10CSP′. The eluent was THF with a flow rate of 1.0 mL min⁻¹ and the testing temperature was 35 °C. The molecular weights and PDIs were derived from a calibration curve based on narrow polystyrene standards with molecular weight from 1200 to 3 940 000 g mol⁻¹.

The transmission electron microscopy (JEOL JEMACRO-1230) was used to observe the core–shell structure of RCSP at an operating voltage of 80 kV. The latex were diluted and mounted on a copper grid, then dried at room temperature for 24 h.

Differential scanning calorimetry (DSC) was carried out with a TA Q200 instrument. The glass transition temperature (T_g) was reported at the inflection point of the heat capacity jump with a heating rate of 10 °C min⁻¹. The particle sizes and distributions were measured by a Malvern ZETASIZER 3000 HAS at 25 °C. The samples were dried in vacuum at 40 °C for 3 h to remove the residual monomers before analysis.

2.5 Fourier transform infrared spectroscopy (FT-IR)

FT-IR measurements were performed on a Nicolet 5700 FT-IR spectrometer with an OMNIC workstation. Blend containing 20 wt% of R5CSP and 80 wt% of DGEBA was measured after stirring in 80 °C for several hours. Samples were casted on KBr pellet and scanned against this blank KBr pellet background (wavenumber 4000 to 400 cm⁻¹; resolution 4.0 cm⁻¹; accumulate 256 scans).

2.6 Dynamic mechanical analysis (DMA)

The temperature dependences of the viscoelastic properties (storage modulus: G′ and loss tangent: tan δ) of the cured epoxy or cured blends were examined by TA Q800 (TA Instruments). Dimensions of specimens were 35.5 × 12 × 1.5 mm³. The analysis were evaluated in torsion mode, a fixed frequency 1 Hz, amplitude 15 μm, and run from −100 °C to 200 °C with temperature increases of 5 °C per minute. The peak in tan δ was considered to be the glass-transition temperature (T_g) of blends.

2.7 Morphologies of RCSP modified epoxy blends

FESEM images were obtained using ULTRA 55 FESEM based on the renowned GEMENI®FESEM column with beam booster (Nano Technology Systems Division, Carl Zeiss NTS GmbH, Germany) with a tungsten gun and applying 5 kV as the accelerating voltage. The fracture samples for morphologies observation were prepared after cooling in liquid N₂ then coated with gold by vapour deposition before observation. TEM observation were done using JEOL JEMACRO-1230 at an operating voltage of 80 kV.

2.8 Fracture tests

Fracture toughness measurements were performed following the linear elastic fracture mechanics (LEFM) approach by single edge notched three-point blending (SEN-3PB) method according to ASTM D 5045-99. The dimensions of parallelepiped specimens are 60 × 10 × 5 mm³. Zwick/Roell Z020 universal material tester was used to perform the tests at a test speed of 10 mm min⁻¹ (23 °C). The cracks were carefully generated first cutting a notch of approximately 4.5 mm using diamond saw blade then tapping a liquid N₂ chilled fresh razor blade. Critical stress intensity factor was calculated by averaging data obtained by at least 3 specimens.

3 Results and discussion

3.1 Synthesis of RCSP

As described in experimental part, RCSP was prepared by RAFT emulsion polymerization followed by core crosslinking, the procedure is shown in Scheme 1. The maleic anhydride in macro-RAFT agent was fully hydrolysed during the process, resulting maleic acid in SMA segment when mixed with epoxy precursor. During the polymerization, the ionized carboxyl groups stayed on the surface of latex and chain growth took place inside the particles. Therefore, reactive brush was formed by amphiphilic macro-RAFT agent, encapsulating the hydrophobic monomers. The early added monomer formed the polystyrene (PS) shell to enhance its redispersibility, and the later added monomer formed the poly(n-butyl acrylate) (PnBA) core. The core morphology was then stabilized by the crosslinking agent, EGDMA, resulting in a core–shell nanostructure. The carboxyl acid groups in the reactive brush were able to covalently link to epoxy precursor or anime hardener.

The conversion of monomers was measured gravimetrically during the whole polymerization, as shown in Fig. S1.† During styrene shell polymerization, the conversion increased rapidly and conversion reached the maximum of 85% after 120 min for R2CSP, 80% after 200 min for R5CSP, and 75% after 300 min for R10CSP. To reduce undesired terminations, the second monomer, n-butyl acrylate, was added into emulsion after 120 min for R2CSP, 200 min for R5CSP, and 300 min for R10CSP. The PnBA inevitably constituted the cores. After 4 hours of polymerization, the conversions of n-butyl acrylate for 3 systems were all higher than 80%. Unfortunately, the molecular weights of as-prepared core crosslinked core–shell particles were not obtained because they could not be completely dissolved in solvents. As a reference, a model polymerization for preparation of non-crosslinked reactive core–shell particle without adding EDGMA cross-linker were conducted to measure the block copolymer formation.

3.2 Morphology of RCSP

To elucidate the block copolymer formation of RCSP, the non-crosslinked RCSPs were prepared following the similar procedure and monitored by GPC. Table S1† presents the M_n and PDI of macro-RAFT agent, SMA-PSt diblock copolymer, and SMA-PSt-PnBA triblock copolymer. The GPC curves are also illustrated in Fig. S2.† After chain-extension, the GPC peaks of each intermediate product shifted toward the region of higher molecular weight and all peaks were in unimodal distribution during the emulsion polymerization progress. The small shoulder peak observed for triblock copolymer in R5CSP′ and R10CSP′, indicating that few dead chains might be generated. But the majority of macromolecular chains did not have dead chains. Similarly, the crosslinked ones would like to be composed of crosslinked triblock copolymers.

The dimension of the RCSP latex was measured by DLS, shown in Fig. 1. After core formation, the average particle size was 250 nm. The size distribution is narrow. Interestingly, the finally obtained RCSPs were similar in size. In RAFT emulsion polymerization, the size of particle stabilized by macro-RAFT agent with higher hydrophilic content was expected to be smaller.³⁷ The stabilizing block molecular weight dispersity was larger in core–shell particles of higher reactivity.³⁴ This leaded to the increase in particle size of RCSP with macro-RAFT agent as surfactant in higher M_n and higher PDI.³⁸ And the neutralization effect of ammonia and emulsion pH on the stabilization performance of macro-RAFT surfactant influenced the particle size.³⁹ All these factors lead to the similar average particle size of three kinds of core–shell particles.

Fig. 1 Particle size distribution of RCSP before and after the formation of PnBA rubbery core. The left peaks were measured when polystyrene shell formed. The right peaks were measured when all monomers had complete the reaction.

Fig. 2 shows the DSC curves of the RCSPs, where the T_g of PnBA rubbery core and PS shell are considered to be −38 °C and 102 °C, respectively. Three arrows on the left indicate the glass transition of rubbery poly(n-butyl acrylate) core, arrows on the right indicate the glass transition of glassy polystyrene. The thermal transition temperature of SMA segment are probably overlapped with that of polystyrene. Due to core crosslinking, the T_g value of PnBA was higher than that of liner block copolymers previously reported.⁴⁰ The clear thermal transition temperatures suggest the phase separation of RCSPs.

Fig. 2 DSC curve of R2CSP, R5CSP, and R10CSP. The glass transition regions are indicated by arrows.

The core–shell structure of R5CSP was clearly revealed by TEM observation, shown in Fig. 3. The inner core is estimated as EGDMA cross-linked PnBA, and the shell is estimated as PS. The diameter of R5CSP observed by TEM is approximately 150 nm, while smaller and larger particles could also been observed. This is in accordance with DLS measurements, because RCSP in emulsion was swelled by water.

Fig. 3 TEM micrographs of R5CSP prepared by RAFT emulsion. Scale bar is 500 nm. The inset provides an interpretation of the core–shell morphology, dark poly(n-butyl acrylate) cores surrounded by a light polystyrene shell.

3.3 Reaction between RCSP and DGEBA

The reaction between RCSP and DGEBA was confirmed by FT-IR measurements. FT-IR spectra of RCSP, DGEBA, and DGEBA (80 wt%)/R5CSP (20 wt%) binary blend were obtained as example and vertically stacked in Fig. 4. Compared to DGEBA epoxy precursor, the binary blend's absorption near 1735 cm⁻¹ (carboxylic acid C=O stretching) and 860 cm⁻¹ (epoxy C–O–C ring deformation) decreased, while absorption near 1770 cm⁻¹ (ester C=O stretching) increased. The unreacted carboxylic groups, a small absorption near 1730 cm⁻¹ in blend, had opportunities to covalently link to amine hardeners during curing, giving additional interfacial adhesion between RCSP and epoxy matrix.

Fig. 4 FT-IR spectra of R5CSP, DGEBA, and DEGBA/R5CSP binary blends.

3.4 Morphology of epoxy thermoset containing RCSP

The reactive brush covalently connected to the end of a linear PS-PnBA-PS triblock copolymer had been proved to be efficiently decreased the inclusion size when blending with DGEBA/DDM thermosets. Through the similar strategy, the RCSP with surface functionalized reactive brushes could be miscible with epoxy matrix. FESEM photographs of the cured blends in Fig. 5 illustrate the well dispersed inclusions of RCSP. As predicted, R0EP had local coagulations and left epoxy matrix with less particles. The poor miscibility of PSt shell caused the R0CSP to have a clear boundary with the epoxy matrix. Small amount of particle aggregation was found in R2CSP modified blends. This suggested that the R2CSP were not able to separate from each other during the mixing process. But for R5EP and R10EP, the particles were well dispersed in matrix. Almost no particle agglomeration was observed in these two blends. The homogeneous distribution of R5CSP and R10CSP in the matrix could be attributed to the strong chemical interaction between RCSP and epoxy. The dimensions of RCSP dispersed in matrix are smaller than 200 nm, will be further confirmed by TEM observations. The RCSP particle size in epoxy was significantly smaller than latex because they were possibly embedded in epoxy matrix and blurring particle-matrix boundaries. This observation also indicates that the core–shell structures of most RCSPs were maintained after demulsification and blending process. Since the RCSPs prepared in this study were of similar in size, one can conclude that it is interfacial adhesion between RCSPs and polymer network play the important role in controlling morphology. It is analogy to RCSPs in water during emulsion polymerization stabilized by SMA macro-RAFT agent. This is in accordance with the results by other researchers.^16,19 And morphology determined the mechanical properties of RCSP modified epoxy blends.

Fig. 5 FESEM photographs of epoxy thermosets (a) R0EP, (b) R2EP, (c) R5EP, and (d) R10EP.

TEM photographs of the cured blends in Fig. S3† clearly reveal the difference in the size of the internal phase structures. The flocculated and locally structured phase was observed in R0CSP modified epoxy blends. Several coalescences were found in R2CSP modified epoxy blends. The nano particles of R5CSP and R10CSP were found to be uniformly dispersed in epoxy matrix. TEM observations further confirmed the morphologies in Fig. 5.

Fig. 6 presents the storage modulus and tan δ curves for epoxy thermosets R2EP, R5EP, and R10EP. The values obtained from DMA are listed in Table 1. The unmodified epoxy had a thermal transition temperature of 142.35 °C, and room temperature modulus of 2.89 GPa. The first T_g around −50 °C corresponds to β transition of epoxy network with small fractions of PnBA rubbery core. The second T_g is estimated as α transition of epoxy network. All epoxy blends modified by RCSP had loss in modulus and T_g. The loss in thermal transition temperature is in contrast with the trend reported in ref. 5 and 16. In current work, the RCSPs were prepared in certain size distributions. The matrix plasticity increased due to the existence of larger particles, which might be in micro meter scale. However, compared to surface functionalized RCSPs, significant decrease in T_g was observed in R0EP. This in accordance with FESEM results, where some local coagulation of R0CSP to larger inclusions was observed in epoxy blends. R0CSP was squeezed out of epoxy precursor during the cosolvent evaporation process, forming micro-sized inclusions which decreased T_g of the blend. It was found that T_g of blends increased as the reactivity of RCSP increased. Addition of SMA reactive brush in shell of RCSP increased T_g from 126 °C for R0EP to 136.0 °C for R10EP.

Fig. 6 DMA plots of neat epoxy, and epoxy modified with 10 wt% of RCSP. (a) Storage modulus (G′) and (b) tan δ curves as a function of temperature are presented.

Table 1 Mechanical and thermal properties, T_g, E, K_IC, and G_IC of RCSP modified epoxy resins investigated in this study

Sample	Tg^a	E^b	GIC^c (J m^−2)	KIC^d
a Tgs were obtained from DMA analysis.b Elastic modulus were obtained from DMA analysis.c Strain energy release rate (GIC) was calculated from KIC and E using equation: GIC = KIC^2(1 − ν^2)/E, and Possion's ratio ν = 0.34.d Stress intensity factor (KIC) was obtained from SENB tests on at least 5 samples.
Unmodified	142.35	2.89	212.41	0.83
R0EP	108.80	2.49	433.02	1.10
R2EP	135.43	2.62	948.54	1.67
R5EP	135.73	2.72	1109.15	1.84
R10EP	136.00	2.82	1276.64	2.01

---

3.5 Fracture toughness of epoxy thermosets modified by RCSP

The K_IC values of cured epoxy blends modified by core–shell particles are compared in Fig. 7. The fracture toughness values of RCSP modified epoxy are all higher than unmodified neat epoxy (approximately 0.8 MPa m^1/2), and increase gradually with length of reactive segment covering the glassy shell. The unreactive CSP increased the toughness of epoxy resin by 30%. However, cured blends containing R2CSP had 100% increase in critical stress intensity factor. Furthermore, significant enhancement in K_IC was observed for R5EP and R10EP, 121% and 142% compared to neat epoxy, respectively. The higher standard deviation of R0EP was higher than that of other samples, which was probably due to the fact that the crack tip radius was similar to the size of R0CSP phase. Depending on exactly where the crack tip is located with respect to the macroscopic structure, the resistance against crack propagation would vary.⁴¹ By increasing the content of SMA in the RCSP, the fracture toughness of modified epoxy resins was simultaneously increased. Our experimental results seemed to contradict with Sue's observation that chemical bonding of CSP to epoxy matrix did not significantly contribute to the toughening performance.¹⁶

Fig. 7 Fracture toughness values of neat epoxy and epoxy thermosets modified with 10 wt% of RCSP.

The mechanical and thermal properties of neat epoxy and RCSP modified blends in this study are presented in Table 1. The elastic modulus, E, of R0CSP modified epoxy was significantly lower than reactive surface functionalized CSP modified thermosets. The G_IC value of the unmodified epoxy resin was 212 J m⁻². Compare to neat epoxy, 3.5 folds increase in G_IC was observed for R2EP. The highest G_IC was obtained for R10EP thermoset, 1276 J m⁻², which was 6 times the fracture energy of neat epoxy. Considering similar size and core–shell structure of RCSPs in this study, the difference in reactivity of the RCSPs should be responsible for the different toughness. So the increased cavitation of inner core of sphere inclusions and the increased shear banding of epoxy matrix would like to be the major mechanisms in toughening, where the cavitated rubbery core relieved the plane strain constraint from nearby matrix, therefore induced matrix plastic deformation.

4 Conclusions

The poly[styrene-alt-(maleic acid)]-block-polystyrene-block-poly(n-butyl acrylate) reactive block copolymer core–shell particles were synthesized via RAFT polymerization mediated by macro-RAFT agent. The resulted RCSPs, identified with core–shell structure, gave unimodal size distributions. A series of RCSPs were carefully designed to have similar size and core–shell structure but different reactivity. The surface functionalized RCSPs were able to covalently link to epoxy matrix, making the RCSPs well-dispersed in cured epoxy matrix. The fracture toughness of modified epoxy resin increased with respect to neat epoxy, the measured K_IC had 142% increase in R10EP compared to unmodified epoxy thermoset. The 10 wt% R10CSP modified epoxy blends had 6 times the fracture energy of neat epoxy and 3 times the G_IC of structurally similar non-reactive core–shell particles. This provided us a new strategy to prepare reactive core shell particle as toughening agent for epoxy resins.

Acknowledgements

The authors would like to thank the National Natural Science Foundation of China (NSFC) for Award No. 21276224, 21476195, 21576236 and Zhejiang Provincial National Science Foundation of China Y14B060038 for supporting this research.

Notes and references

1. L. Pan, S. Lu, X. Xiao, Z. He, C. Zeng, J. Gao and J. Yu, RSC Adv., 2015, 5, 3177.
2. R. Konnola, J. Joji, J. Parameswaranpillai and K. Joseph, RSC Adv., 2015, 5, 61775.
3. Manju, P. K. Roy and A. Ramanan, RSC Adv., 2014, 4, 52338.
4. K. Sethuraman, P. Prabunathan and M. Alagar, RSC Adv., 2014, 4, 45433.
5. D. Quan and A. Ivankovic, Polymer, 2015, 66, 16.
6. G. Giannakopoulos, K. Masania and A. C. Taylor, J. Mater. Sci., 2011, 46, 327.
7. R. Bagheri, B. T. Marouf and R. A. Pearson, Polym. Rev., 2009, 49, 201.
8. J. Choi, A. F. Yee and R. M. Laine, Macromolecules, 2004, 37, 3267.
9. N. Fu, G. Li, Q. Zhang, N. Wang and X. Qu, RSC Adv., 2014, 4, 1067.
10. J. Guo, J. Long, D. Ding, Q. Wang, Y. Shan, A. Umar, X. Zhang, B. L. Weeks, S. Wei and Z. Guo, RSC Adv., 2016, 6, 21187.
11. Z. Liu, Y. Deng, Y. Han, M. Chen, S. Sun, C. Cao, C. Zhou and H. Zhang, Ind. Eng. Chem. Res., 2012, 51, 9235.
12. D. Shi, E. Liu, T. Tan, H. Shi, T. Jiang, Y. Yang, S. Luan, J. Yin, Y. Mai and R. K. Y. Li, RSC Adv., 2013, 3, 21563.
13. R. A. Ramli, W. A. Laftah and S. Hashim, RSC Adv., 2013, 3, 15543.
14. Y. Guo, S. Sun and H. Zhang, RSC Adv., 2014, 4, 58880.
15. K. Lin and Y. Shieh, J. Appl. Polym. Sci., 1998, 69, 2069.
16. K. Lin and Y. Shieh, J. Appl. Polym. Sci., 1998, 70, 2313.
17. M. Naguib, S. Grassini and M. Sangermano, Macromol. Mater. Eng., 2013, 298, 106.
18. S. Ritzenthaler, F. Court, L. David, E. Girard-Reydet, L. Leibler and J. P. Pascault, Macromolecules, 2002, 35, 6245.
19. J. Y. Qian, R. A. Pearson, V. L. Dimonie and M. S. El-Aasser, J. Appl. Polym. Sci., 1995, 58, 439.
20. V. Rebizant, A. Venet, F. Tournilhac, E. Girard-Reydet, C. Navarro, J. Pascault and L. Leibler, Macromolecules, 2004, 37, 8017.
21. N. Song, J. Yang, P. Ding, S. Tang, Y. Liu and L. Shi, Ind. Eng. Chem. Res., 2014, 53, 19951.
22. J. P. Zhou, H. R. Wu, W. L. Fu and S. J. Jia, Appl. Mech. Mater., 2014, 496–500, 71.
23. M. Ochi, N. Ichikawa, M. Harada, M. Hara and H. Uchida, J. Appl. Polym. Sci., 2012, 124, 4572.
24. Y. S. Thio, J. Wu and F. S. Bates, J. Polym. Sci., Part B: Polym. Phys., 2009, 47, 1125.
25. J. Wu, Y. S. Thio and F. S. Bates, J. Polym. Sci., Part B: Polym. Phys., 2005, 43, 1950.
26. J. M. Dean, R. B. Grubbs, W. Saad, R. F. Cook and F. S. Bates, J. Polym. Sci., Part B: Polym. Phys., 2003, 41, 2444.
27. J. P. Pillai, J. Pionteck, R. Hässler, S. M. George, U. Cvelbar and S. Thomas, Ind. Eng. Chem. Res., 2011, 50, 4432.
28. L. Yang, J. Xu, P. Sun, Y. Shen and Y. Luo, Ind. Eng. Chem. Res., 2014, 53, 11259.
29. L. Yang, J. Xu, J. Han, Y. Shen and Y. Luo, Ind. Eng. Chem. Res., 2015, 54, 5536.
30. C. F. van Nostrum, Soft Matter, 2011, 7, 3246.
31. S. Ritzenthaler, F. Court, E. Girard-Reydet, L. Leibler and J. P. Pascault, Macromolecules, 2002, 36, 118.
32. V. Rebizant, V. Abetz, F. Tournilhac, F. Court and L. Leibler, Macromolecules, 2003, 36, 9889.
33. Z. Luo, Q. Lu, F. Ma and Y. Jiang, J. Appl. Polym. Sci., 2014 DOI:10.1002/app.40898.
34. X. Zhan, R. He, Q. Zhang and F. Chen, RSC Adv., 2014, 4, 51201.
35. Y. Yu, Q. Zhang, Z. Wang, X. Zhan, R. He, W. Zhang and F. Chen, J. Macromol. Sci., Part A: Pure Appl. Chem., 2012, 49, 60.
36. P. Das and J. P. Claverie, J. Polym. Sci., Part A: Polym. Chem., 2012, 50, 2802.
37. L. Zhang and A. Eisenberg, J. Am. Chem. Soc., 1996, 118, 3168.
38. K. Yan, X. Gao and Y. Luo, J. Polym. Sci., Part A: Polym. Chem., 2015, 53, 1848.
39. X. Wang, Y. Luo, B. Li and S. Zhu, Macromolecules, 2009, 42, 6414.
40. Y. Luo, X. Wang, Y. Zhu, B. Li and S. Zhu, Macromolecules, 2010, 43, 7472.
41. J. D. Liu, Z. J. Thompson, H. Sue, F. S. Bates, M. A. Hillmyer, M. Dettloff, G. Jacob, N. Verghese and H. Pham, Macromolecules, 2010, 43, 7238.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra05048d

---