Thermo-oxidative stabilization for natural rubber nanocomposites by polydopamine interfacial tailored clay

Abstract

Polydopamine (PDA) is applied to natural rubber (NR)/clay nanocomposites through a facile dip-coating approach, which is inspired by the mussel adhesion proteins. This novel modification of clay perfectly captures the interfacial tailoring reinforcement principles for rubber/filler compounds. The hydrophilic PDA coating assists clay platelets to disperse uniformly in aqueous NR latex with respect to TEM observations. Furthermore, PDA modified montmorillonite (PDA-MMT) interacts with NR chains by robust hydrogen-bonding adhesion, which serves as an efficient physical cross-linker demonstrated with transverse magnetization decay analyses. The expectable enhancements are directly visualized in strengthened tensile properties with 46% increasement of tensile strength by adding 6 phr PDA-MMT and only 18% increasement without PDA modification. Besides, the E_α→0 values of NR based composites follow such sequence: NR/MMT < NR < NR/PDA-MMT. Synthetic substitute of natural melanin as PDA is, PDA-MMT layers remarkably stabilize the NR matrix via radical-scavenging pathway during heat-treatment.

---

1. Introduction

In the past decades, natural rubber (NR, composed of cis-polyisoprene and trivial protein, ash) has been a popular candidate for engineering applications, due to the well-known high green-strength arising from its molecular regularity and crystallinity. However, reinforcing of NR is still widely implemented to meet the application requirements. Carbon black (CB) as well as silica represent traditional particulate reinforcing fillers in the rubber industry. To optimize the mechanical properties of NR, a considerable amount of such fillers is mingled with the gum rubber.^1–4 Laminate fillers have emerged in response to the increasing demands on impact resistance, fatigue resistance, dimensional stability and gas barrier properties of rubber composites, typically the layered silicate clay (e.g. montmorillonite, MMT).^5,6 However, it should be incorporated modified clay but dispensing with pristine clay to enhance the interfacial interactions.^7–10

Polydopamine (PDA) surface modification, a facile surface modification approach inspired by the adhesion mechanism of mussels,^11,12 can be regarded as an inherent surface modification of clay related in polymer/clay nanocomposites.^13–15 Therefore, we conducted the interfacial reinforcement design for NR/clay nanocomposite by PDA modification, which emphasized the important roles of clay dispersion, clay–matrix interactions and the overall networks, respectively.¹⁶ Firstly, bio-mimetic substance as PDA it is, it demonstrates an intrinsic hydrophilic ability due to the abundant hydroxyl groups.^17–19 Thus, PDA coating can remarkably improve the aqueous coagulation compounding process of clay and NR latex. Secondly, effective physical interactions are indispensable for rubbery reinforcement.^20,21 PDA exhibits a forceful adhesion onto various substrates, benefit from its robust hydrogen bonding of hydroxyl groups as well. Yang et al. have evidenced such bonding between PDA modified MMT (PDA-MMT) and polyether epoxy matrix by means of the wavenumber-downshift in FT-IR as well as the gel-like rheological behaviors of PDA-MMT.²² Besides, the typical fillers (unmodified CB and silica) usually interact with NR chains via physical interactions. Thus, the interfacial interplay of PDA-MMT coincides with that of traditional reinforcing fillers.

Moreover, the cross-linking network plays a vital role in rubber matrix, which is correlated with excellent mechanical properties of matrix. The network components (such as crosslinking domains and dangling chains, with different mobility) can reflect themselves in the relaxation curves describing molecular dynamics. For instance: López-Manchado et al. evidenced the uniform networks of NR/Organ-MMT by dielectric loss spectra, in which NR/Organ-MMT demonstrated less segmental transitions corresponding to less network defects.²³ On analogy, we observed the magnetic relaxation of protons from NR main-chains to evaluate the regularity of networks and to calculate the crosslinking density.

Many theoretical models have been put forward to clarify the mechanisms of rubber reinforcement.^24–26 It is noteworthy that the classical stress-transfer model for plastic polymers is not applicable to rubber for their absolutely different thermodynamic states. Rubbery polymers possess mobile chains in ambient environment due to their much lower glass transition temperatures. To explain the obtained drastic reinforcement for NR/PDA-MMT systems, the “induced stretching and parallel-arraying chains” model²⁷ associated with “improving stress-induced-crystallization” mechanism²⁸ was adopted in our research.

In this paper, PDA modification satisfactorily complied with the proposed interfacial tailoring reinforcement principles. The dispersion of clay layers was observed by TEM. The interfacial interactions as well as network structure was estimated by magnetic relaxation parameters (micro-scale), and then ascertained by DSC curves (macro-scale). Finally, since the anchored NR chains slide away upon the PDA-MMT surface to consume the external loadings, and the adding clay could facilitate the crystallization of NR, NR matrix was endowed with excellent mechanical properties. The simultaneously enhanced thermal stability of NR/PDA-MMT was characterized through calculating degradation activate energy E_a (Flynn–Wall–Ozawa method). Furthermore, the extrapolation of E_a was carried out to explore the initial thermodynamic state of NR/PDA-MMT.

2. Experimental

2.1 Materials

Sodium montmorillonite (MMT) was supplied by Siping Liufangzi Aska Bentonite Co., Ltd, Jilin Province, China. Natural rubber latex (solid content ≈ 60 wt%) was obtained from Hainan Rubber Products Corporation, China. Dopamine hydrochloride and Tris(hydroxymethyl)-aminomethane (Tris) were purchased from Alfar Aser as received. Other reagents were all commercially available.

2.2 Preparation of PDA-MMT

PDA-MMT was synthesized using the same method reported in our previous publication.²⁹ The supernatant suspension of MMT with a concentration of 2 wt% was obtained by stirring in DI water for 8 hours at 1600 rpm and rest for at least 24 hours. Then, the clay suspension was exposed to ultrasonic treatment (800 W) for 15 minutes. Dopamine (1.5 g L⁻¹) and Tris (1.2 g L⁻¹) were added into the as-received clay suspension, followed by continuous agitating for 4 hours at ambient temperature. The resulting PDA-MMT was used as a sludge for the later compounding with NR latex.

2.3 Preparation of NR/PDA-MMT nanocomposites

The NR/clay nanocomposites were prepared by latex compounding. A pre-determined amount of PDA-MMT suspension was blended with the NR latex by stirring for 30 minutes. Then, sulphuric acid solution (3 wt%) was incorporated to flocculate the NR/PDA-MMT rubber particles. The solid products were rinsed until their surface pH appeared to be neutral and then dried for 24 hours at 50 °C.

The flocculate rubber was masticated in a 6 inch double-roll open mill, followed with mingling all the agents according to the formulation listed in Table 1 in turn. After resting for 24 hours to assure a good diffusion of these ingredients in the matrix, the gross rubber was vulcanized at 143 °C for its corresponding T₉₀ (the optimum curing time). The control samples (NR and NR/MMT) were prepared under the same conditions. In addition, the sample is simply abbreviated as NR, followed by one-digit number indicating the clay mass fraction per hundred of rubber matrix (phr). Especially, the NR matrix (100 phr) containing 2 phr PDA-MMT (or MMT) is abbreviated as NR/PDA-MMT-2 (or NR/MMT-2).

Table 1 Composition of PDA-MMT filled NR systems

Ingredient	Loading^a (phr)
a Weight parts per 100 weight parts of rubber.b N-Isopropyl-N′-methylphenyl-p-phenylene diamine.c Dibenzothiazole disulfide.d Diphenyl guanidine.e Tetramethyl thiuram disulphide.
NR	100
PDA-MMT	0/2/4/6/8
Zinc oxide	5
Stearic acid	2
Anti-oxidant 4010NA^b	1
Accelerator DM^c	0.5
Accelerator D^d	0.5
Accelerator TT^e	0.2
Sulfur	2

---

2.4 Characterization

The NR/clay samples were sliced by microtome (LEICA EM FC7) into ultrathin sections prior to TEM observation (H-800, HITACHI, Japan). The glass transition temperature of NR based composites was achieved by using DSC equipped with STAReSystem (Mettler-Toledo, Switzerland). The samples were heated from −100 °C to room temperature in nitrogen atmosphere with a heating rate of 10 °C min⁻¹. The mechanical properties were carried out by a universe electronic testing machine (SANS CMT 4104, China) with a tensile speed of 500 mm min⁻¹ at room temperature. For each data reported, five sample measurements were averaged.

The crosslinking density of NR samples was measured using transverse ¹H-NMR decay along with IIC XLDS-15 NMR crosslinking density spectrometer (IIC Innovative Imaging Corp. KG, Germany). The transverse magnetization decay (TMD) was provoked by inter- and intra-dipolar interactions of protons in network. The rubbery network comprises with crosslinking chains and some dangling chain ends, which exhibit different molecular mobility and thus differ from TMD signals. The motion of crosslinking chains restricted by various cross-linkers present a Gaussian feature in TMD signal. Here, the ratio q is proposed to define the residual dipolar interaction after averaging by the abovementioned constraints. Conversely, since the dangling chain ends are usually free and their motion could be averaged out completely, they demonstrate an exponential feature (q = 0). The resulting TMD of samples is composed of Gaussian and exponential fractions using Marquardt–Levenberg algorithm equation below:

where A₁ and A₂ represent the relative fraction of crosslinking chains (Gaussian) and dangling chain ends (exponential), respectively, T₂ corresponds to the transverse relaxation time, q × M₂ is the residual dipolar interaction that is crucial for the following determination of cross-link parameters of rubber. Namely, the Gaussian fraction of TMD indeed reveals the crosslink property of rubber. M_c (average molecular weight between two cross-linkers) can be figured out upon the ratio q, concomitant with the calculation of ν_c (the number of active network chains per unit volume of the rubbery matrix), since νc is inversely related to M_c.^30–32

Rod-shape rubber sample (ca. 8 mm × 5 mm × 2 mm) in glass tube was inserted in magnetic field for 3 min to reach equilibrium. The measuring temperature was 60 °C for NR samples. In order to calculate the exact crosslinking density, relaxation parameters for NR samples were figured out using the IIC Analysis Software package.

The NR/clay samples were heated from 40 °C to 600 °C (10 °C min⁻¹) by TGA (METTLER-TOLEDO, Switzerland), with a purging rate of 50 ml min⁻¹ in air and nitrogen atmosphere, respectively. To compute the degradation apparent activation energy E_a by Flynn–Wall–Ozawa method, the samples were heated in air at different heating rate of 5 °C min⁻¹, 10 °C min⁻¹, 20 °C min⁻¹, 30 °C min⁻¹, respectively. By plotting −ln β versus 1000/T for a given conversion using TGA data (air atmosphere), the specific E_a can be determined from the slope of the fitting line in light of the equation below:

d(ln β)/d(1/t) = −E_a/R (2)

where β is the heating rate (K min⁻¹), T is the temperature for a given conversion (K), E_a is activation energy (kJ mol⁻¹), R is the universal gas constant (8.314 kJ mol⁻¹ K⁻¹).

3. Results and discussion

3.1 Interfacial reinforcement design of NR/clay

PDA-MMT imparts limited mechanical enhancement to SBR matrix,²⁹ which attributes to the poor stress-induced crystallization ability of SBR chains. Exceptions to that trend have been discovered in NR matrix. The interfacial tailoring reinforcement for NR/clay nanocomposites includes three pivotal principles: the dispersion state of clay platelets, the interfacial interplay between clay and NR as well as the optimal crosslinking density of rubber. PDA modification satisfactorily achieves all the above objectives and acquires simultaneous enhanced properties, which is given in detail as follows.

3.1.1 Dispersion state of clay platelets. Homogeneous dispersion of clay is an essential prerequisite to rubber/clay reinforcement. Since clay has an intrinsic tendency of self-aggregation in polymeric matrix,³³ ultrasonic of pristine MMT before PDA coating is used to improve the dispersion of layers in this study.

Considering the weak XRD signal of rubber/clay composites in our previous study about SBR/clay,²⁹ TEM observation was established to gain direct insight of the dispersion state and morphology. At lower magnification (Fig. 1(a) and (c)), the laminate structures of clay can be barely identified. With improving the magnification, it can be observed that the clay platelets evenly dispersed in matrix ranging from 100–300 nm. The platelet size coincides with the ultrasonic treated MMT, most of which is in the form of a mono-layer distribution with a width ranging from 200 nm to 1000 nm. Especially, PDA-MMT exhibits a more uniform and thorough exfoliation state in NR (Fig. 1(d)), whereas MMT shows an intercalated and aggregated exfoliation state (Fig. 1(c)).

Fig. 1 TEM graphs of (a) (b) NR/MMT, and (c) (d) NR/PDA-MMT, in which the remaining strains should be same ingredients incorporated in mix process.

Although some clay agglomeration exists in matrix, the global exfoliation amplitude of NR systems can eclipse that of SBR. The following impressive enhancement of properties is predictable due to such fine dispersion state of clay.

3.1.2 Interfacial interplay & crosslinking density of NR/clay. Effective interfacial interplay is a vital part of rubber reinforcement. The NR chains anchored onto clay layers align and then slide away to average the external stress. Thus, in order to guarantee that effective slippage, a compromising of chemical bonding and physical adhesion must be taken into consideration in the case of interfacial interaction, in which the thorough chemical-bonding hinders the chains' slippage and the weak physical-bonding results in the chains' easy slip-away.

Here strong hydrogen bonding from the abundant catechol groups of PDA and the π–π conjugation of the pendant aromatic structures give rise to impressive interfacial interactions in NR/clay systems.

The transverse magnetization decay (TMD) spectra of NR/clay composites are plotted in Fig. 2. In general, with the increment of clay fractions, all the magnetic signals decay rapidly, suggesting clay platelets exert more hindrance upon NR molecules. However, the superposition degree in unmodified systems exceeds that of PDA-containing systems, which indicates the more significant impact of PDA-MMT on NR composite network.

Fig. 2 Transverse magnetization decay spectra of (a) NR/MMT, and (b) NR/PDA-MMT, in which the inset graphs enlarge the TMD signal over short relaxation times.

Given the ambiguity of TMD spectra, the transverse relaxation time T₂ is selected to quantitate the advantage of PDA modification. T₂ is designated as the relaxation time of fast-motion domains (the dangling chain ends) of chains corresponding to the exponential fraction of eqn (1), which reveals the information concerning molecular mobility as well as the filler–matrix interaction. The variation of T₂ listed in Table 2 further consolidates the evolution trend of TMD spectra by representing the solid-like (fast-decaying) component should not affect the outcome of the analysis for the liquid-like domains. In detail, T₂ of NR/PDA-MMT is substantially lower than that of NR/MMT under identical loading levels, because PDA-MMT exerts too much constraint upon NR chains so that the magnetic signal decays swiftly.

Table 2 The transverse relaxation time of NR/clay composites

Clay content^a (phr)	T2 (ms)
	NR/MMT	NR/PDA-MMT
a The T2 of virgin NR is 4.54 ± 0.21 ms.
2	3.95 ± 0.18	3.72 ± 0.16
4	3.95 ± 0.17	3.54 ± 0.16
6	3.81 ± 0.13	3.52 ± 0.16
8	3.63 ± 0.12	2.92 ± 0.08

---

Additionally, a 1.9 °C improvement of T_g (glass transition temperature, the onset temperature for macro-scale) in NR/PDA-MMT relative to virgin NR can be observed in Fig. 3. Thus the strong interaction between PDA-MMT and NR is corroborated from not only the molecular-scale (T₂) but also macro-scope (T_g).

Fig. 3 DSC curves of NR/clay systems.

Crosslinking density ν_c is an important indicator of rubbery materials, which is closely related to the mechanical properties. The density ν_c and the fraction of crosslinking chains (A₁, also the coefficient of Gaussian fraction in eqn (1)) were employed to analyze crosslinking structure (seen in Fig. 4).

Fig. 4 The curves of cross-link density as well as cross-link fraction A₁ versus clay content, respectively (cross-link density solid lines, cross-link fraction dot lines).

As the loading of clay augments, the crosslinking density value ν_c of NR/PDA-MMT composites increases fast. Such mounting trend implies the addition of clay (especially PDA-MMT) improves the vulcanization process of NR. On the other hand, the fraction value A₁ (according to cross linking chains) climbs steadily because of the present of clay in the case of NR/PDA-MMT, whereas A₁ quickly reaches a plateau in unmodified systems. The increasing of A₁ (cross-link domains) represents the diminishing of A₂ (dangling chains), demonstrating PDA coating can lessen the network defect. PDA-MMT platelets pose as efficient cross-linkers for overall network of NR, and also help to compensate the interfacial incompatibility arising from introducing inorganic fillers.

Sufficient cross-linkers in NR based matrix prolong the effective chain length, which amounts to curtail the inter-distance of neighbor clay layers in favor of bridging aligning NR chains. However, excessive cross-linkers may handicap the anchoring or even the slippage of chains, which in turn undermines the reinforcement of rubber.

3.2 Simultaneously enhanced properties of NR/clay nanocomposite

3.2.1 Mechanical properties. Predicated on the above reinforcement design, the stress–strain curves of NR/clay nanocomposites are shown in Fig. 5. With the increasing of clay loading amount, the tensile properties (the elongation and the tensile strength) rise up accordingly. Besides, the increasing amplitude becomes more obvious in the pattern of NR/PDA-MMT.

Fig. 5 The stress–strain curves of NR/clay nanocomposites.

As the “induced stretching and parallel-arraying chains” theory of Wang et al.,²⁷ the incorporated fillers are constituted by the neighbor fillers and the anchored chains between them which can form stretching chains during tension. The even-dispersion of PDA-MMT and optimal crosslinking density of NR decrease the inter-distance of neighbor clay layers, which facilitates the formation of anchored stretching NR-chains. These stretching chains begin to reorient and slide away to average the stress during tensile process, in case of the successive chain fractures as well as material destructions. Accordingly, adequate interfacial interactions of NR/PDA-MMT guarantee the quantity as well as the mobility of such chains.

Furthermore, especially for NR/clay based composites, López-Manchado et al. complemented the details. The uniformly distributed clay layers improve the stress induced crystallization (SIC) of NR during the later stage of tensile process, which greatly strengthen the mechanical properties. Thus the incorporating PDA-MMT expedites NR chains to reorient and then crystallizes under stress, followed by the significantly improved tensile properties. Unfortunately, excess amount of PDA-MMT give rise to detrimental cross-linkers. That's why the most pronounced tensile increment is realized by adding only 6 phr PDA-MMT.

3.2.2 Thermal stabilities. Simultaneously enhanced thermal stability of NR/PDA-MMT systems is also expected, attributable to its multi-functional biomimetic interface serving as not only physical barrier but also chemical scavenger.^34–37

As shown in Fig. 6, the samples demonstrate rather plain decomposition process in either air or nitrogen atmosphere, implying a relatively simple mechanism for NR based nanocomposites. It can be inferred that NR/clay undergoes a mono-decomposition pathway (chain scission) relative to a multi-decomposition pathway of SBR/clay (chain scission followed by cross-linking). The degradation of NR/clay proceeds from 270 to 500 °C in air while it exhibits a 50 °C delay in nitrogen approximately, which results from the shift of rate-limiting step in different atmospheres. The limiting step of degradation should be assigned to the random scission of polymer chains (higher energy) in nitrogen, whereas it converts to the decomposition of reactive hydroperoxide radicals (lower energy) in air. Additionally, the inclusion of PDA-MMT helps to curtail the degradation amount of NR matrix, especially in oxidative atmosphere.

Fig. 6 TGA curves of NR/clay systems in (a) nitrogen atmosphere, and (b) air atmosphere.

Thermal kinetic analyses (Flynn–Wall–Ozawa method) were utilized to systematically elucidate the thermal stability of NR/clay, as our previous study did for SBR.^29,38,39 The usage of PDA-MMT brings impressive thermal enhancement to NR matrix over that of unmodified MMT. A wide energy difference (22–47 kJ mol⁻¹) can be observed between NR/PDA-MMT-8 and NR/MMT-8, as tabulated in Table 3.

Table 3 The degradation activation energy of NR/clay by Flynn–Wall–Ozawa method

Conversion (%)	Ea (kJ mol^−1)
	NR	NR/MMT	NR/PDA-MMT
10	125.1	134.1	156.6
15	149.4	155.0	182.9
20	163.0	174.8	210.1
25	182.5	196.1	239.9
30	198.9	214.0	261.4
α → 0	92.0	90.7	114.5

---

Based on the iso-conversion model, Flynn–Ozawa–Wall method excludes the definite selection of complicated mechanism functions. However, due to the different heating rates, the sample system is away from balance and the obtained E_a values thus deviate from true values. Given that E_a value exhibits linear proportion to the exact conversion α (the weight loss percentage in TGA), extrapolation assay was carried out to simulate the critical E_a in initial state (E_α→0) (seen in Fig. 7 and Table 3). The E_α→0 values of NR based composites follow such sequence: NR/MMT < NR < NR/PDA-MMT, amplifying that unmodified NR/clay system shows inferior thermo-stability owing to the short of effective interfacial interaction.

Fig. 7 The activation energy of NR/clay systems at various conversions (a) virgin NR, (b) NR/MMT-8, (c) NR/PDA-MMT-8, (d) extrapolated E_α→0.

Furthermore, Fig. 8 displays the variation of tensile strength after ageing test (the dumbbell-tensile samples were subject to 100 °C heating treatment for 48 h in ageing-oven under air atmosphere). The loss-percentage of tensile strength descends as the loading of clay rises, in which PDA modified clay shows superior property maintenance.

Fig. 8 The variation curves of tensile properties after ageing test (100 °C × 48 h) (a) NR/MMT, (b) NR/PDA-MMT.

NR exacted from Hevea brasiliensis contains some transitional metal ions (Fe, Cu, Mn and so on) which could intensify the degradation of rubber matrix. Since polydopamine coating is convinced to chelate such metal ions (e.g. Fe³⁺), hence it provides NR with enhanced thermo-stability specially.^40,41

4. Conclusions

The PDA surface modification is applicable to NR/clay systems. The dispersion state of PDA-MMT in NR was conceived to be exfoliation (ca. 100–300 nm) demonstrated with TEM images. Modified MMT platelets interact with NR via robust hydrogen-bonding, which reflects in the significantly improved according constraint amplitude to NR chains, the declined T₂ and the increased T_g in NR/PDA-MMT. Besides, the cross linking density ν_c of NR was substantially elevated from 11.8 mol cm⁻³ to 17.6 mol cm⁻³ by incorporating 8 phr PDA-MMT. PDA modification perfectly captures the interfacial tailoring reinforcement principles, hence expected enhancements are found in mechanical properties as well as thermal stabilities. The tensile strength of NR was improved up to 28.7 MPa with adding 6 phr PDA-MMT, relative to 23.2 MPa with 6 phr MMT. The degradation energy of NR/PDA-MMT exceeded that of NR/MMT with a wide energy difference (22–47 kJ mol⁻¹) while the inferior extrapolated energy (E_α→0) of NR/MMT (90.7 kJ mol⁻¹), which stressed the importance of PDA modification. Moreover, PDA modification enhanced the ageing resistance of NR/clay composites.

Acknowledgements

The authors sincerely appreciate the financial supports from the Natural Science Foundation of China (Grant No. 51320105012, 51373010), and the BUCT fund for Disciplines Construction and Development (Project No. XK1511).

References

1. Y. B. Liu, L. Li and Q. Wang, Plast., Rubber Compos., 2010, 39, 370–376.
2. G. Heinrich, M. Klüppel and T. A. Vilgis, Curr. Opin. Solid State Mater. Sci., 2002, 6, 195–203.
3. E. Guth, J. Appl. Phys., 1945, 16, 20–25.
4. N. Rattanasom, T. Saowapark and C. Deeprasertkul, Polym. Test., 2007, 26, 369–377.
5. A. Mousa, G. Heinrich and U. Wagenknecht, Polym.-Plast. Technol. Eng., 2011, 50, 1388–1392.
6. S. Varghese and J. Karger-Kocsis, J. Appl. Polym. Sci., 2004, 91, 813–819.
7. K. S. U. Devi, D. Ponnamma, V. Causin, H. J. Maria and S. Thomas, Appl. Clay Sci., 2015, 114, 568–576.
8. Y. P. Wu, Y. Q. Wang, H. F. Zhang, Y. Z. Wang, D. S. Yu, L. Q. Zhang and J. Yang, Compos. Sci. Technol., 2005, 65, 1195–1202.
9. Y. T. Vu, J. E. Mark, L. H. Pham and M. Engelhardt, J. Appl. Polym. Sci., 2001, 82, 1391–1403.
10. Y. Wang, L. Zhang, C. Tang and D. Yu, J. Appl. Polym. Sci., 2000, 78, 1879–1883.
11. H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426–430.
12. H. Lee, B. P. Lee and P. B. Messersmith, Nature, 2007, 448, 338–341.
13. L. Yang, S. L. Phua, J. K. Teo, C. L. Toh, S. K. Lau, J. Ma and X. Lu, ACS Appl. Mater. Interfaces, 2011, 3, 3026–3032.
14. S. L. Phua, L. Yang, C. L. Toh, S. Huang, Z. Tsakadze, S. K. Lau, Y. W. Mai and X. Lu, ACS Appl. Mater. Interfaces, 2012, 4, 4571–4578.
15. S. L. Phua, L. Yang, C. L. Toh, D. Guoqiang, S. K. Lau, A. Dasari and X. Lu, ACS Appl. Mater. Interfaces, 2013, 5, 1302–1309.
16. S. Huang, S. L. Phua, W. Liu, G. Ding and X. Lu, RSC Adv., 2014, 4, 1425–1431.
17. L. Zhu, Y. Lu, Y. Wang, L. Zhang and W. Wang, Appl. Surf. Sci., 2012, 258, 5387–5393.
18. Y. Yan, Z. Zheng, C. Deng, Y. Li, X. Zhang and P. Yang, Anal. Chem., 2013, 85, 8483–8487.
19. Y. T. Liu, K. C. Kung, C. Y. Yang, T. M. Lee and T. S. Lui, J. Mater. Chem. B, 2014, 2, 7927–7935.
20. N. Ning, Q. Ma, S. Liu, M. Tian, L. Zhang and T. Nishi, ACS Appl. Mater. Interfaces, 2015, 7, 10755–10762.
21. R. Sa, Y. Yan, Z. Wei, L. Zhang, W. Wang and M. Tian, ACS Appl. Mater. Interfaces, 2014, 6, 21730–21738.
22. L. Yang, S. L. Phua, J. K. H. Teo, C. L. Toh, S. K. Lau, J. Ma and X. Lu, ACS Appl. Mater. Interfaces, 2011, 3, 3026–3032.
23. M. A. López-Manchado, B. Herrero and M. Arroyo, Polym. Int., 2003, 52, 1070–1077.
24. J. Fröhlich, W. Niedermeier and H.-D. Luginsland, Composites, Part A, 2005, 36, 449–460.
25. P. Bindu and S. Thomas, J. Phys. Chem. B, 2013, 117, 12632–12648.
26. X. Zhang and L. S. Loo, Macromolecules, 2009, 42, 5196–5207.
27. Z. Wang, J. Liu, S. Wu, W. Wang and L. Zhang, Phys. Chem. Chem. Phys., 2010, 12, 3014–3030.
28. J. Rault, J. Marchal, P. Judeinstein and P. A. Albouy, Macromolecules, 2006, 39, 8356–8368.
29. L. Wang, L. Hu, S. Gao, D. Zhao, L. Zhang and W. Wang, RSC Adv., 2015, 5, 9314–9324.
30. B. P. Mayer, S. C. Chinn, R. S. Maxwell and J. A. Reimer, Chem. Eng. Sci., 2009, 64, 4684–4692.
31. R. S. Maxwell, S. C. Chinn, D. Solyom and R. Cohenour, Macromolecules, 2005, 38, 7026–7032.
32. K. Saalwächter and A. Heuer, Macromolecules, 2006, 39, 3291–3303.
33. S. Wang, Y. Zhang, W. Ren, Y. Zhang and H. Lin, Polym. Test., 2005, 24, 766–774.
34. S. L. Phua, L. Yang, C. L. Toh, G. Ding, S. K. Lau, A. Dasari and X. Lu, ACS Appl. Mater. Interfaces, 2013, 5, 1302–1309.
35. K. Shanmuganathan, J. H. Cho, P. Iyer, S. Baranowitz and C. J. Ellison, Macromolecules, 2011, 44, 9499–9507.
36. P. Meredith, B. J. Powell, J. Riesz, S. P. Nighswander-Rempel, M. R. Pederson and E. G. Moore, Soft Matter, 2005, 2, 37–44.
37. K. Y. Ju, Y. Lee, S. Lee, S. B. Park and J. K. Lee, Biomacromolecules, 2011, 12, 625–632.
38. H. Zhang, Y. Wang, Y. Wu, L. Zhang and J. Yang, J. Appl. Polym. Sci., 2005, 97, 844–849.
39. Q. X. Jia, Y. P. Wu, M. Lu, S. J. He, Y. Q. Wang and L. Q. Zhang, Instrum. Sci. Technol., 2008, 15, 193–205.
40. M. J. Sever, J. T. Weisser, J. Monahan, S. Srinivasan and J. J. Wilker, Angew. Chem., 2004, 116, 454–455.
41. L. Yang, J. Kong, D. Zhou, J. M. Ang, S. L. Phua, W. A. Yee, H. Liu, Y. Huang and X. Lu, Chem. –Eur. J., 2014, 20, 7776–7783.

---