A simple fabrication of electrospun nanofiber sensing materials based on fluorophore-doped polymer

Abstract

A new method for fabricating fluorescent nanofibers as sensing materials was developed via the electrospinning technique using fluorophore-doped polymers. Fluorescent nanofibers with diameters in the range 200–600 nm were produced with fluorophores dispersed uniformly in the matrix. Sensing nanofibers with secondary porous structures were further achieved by introducing a porogen or deacetylation treament as a new approach, exhibiting unique morphologies with intra-fiber pores and obviously improved quenching sensitivity. Methyl violet was used as a model to investigate the sensing performance of a nanofiber membrane composed of cellulose acetate doped with 9-chloromethylanthracene. This work provides a simple and applicable strategy for the fabrication of sensing materials for optical sensor devices.

---

Introduction

One-dimension (1D) nanostructures have attracted much attention due to their fascinating properties and applications.^1,2 Among those strategies used to generate 1D nanostructures, the electrospinning technique provides a simple approach for versatile nanofiber production.^3–9 Electrospun nanofibers should also be 3D for their inherent porosity. Electrospun nanofibers possess remarkable features, such as desirable fiber length, well-controlled diameter and large surface area-to-volume ratio. As a result, the electrospinning technique finds applications in many fields, including reinforced composites,^10,11 tissue engineering,^12–16filtration industry,¹⁷ supports for enzymes and catalysts,^18–20 electronic devices^21–23 and sensing materials.^24–30 Electrospun nanofibers with specific secondary structures involve core–sheath, hollow and porous structures with increased surface areas,⁷ and are receiving much attention during recent years.

Electrospun nanofibers for fluorescence sensor applications have emerged in recent years as a fascinating progress. The reports on fluorescence signal-based electrospun nanofibers are mainly based on covalence and assembly approaches.^24–27 By covalently binding pyrene methanol to poly(acrylic acid), Wang et al. developed a highly responsive optical sensor for Fe³⁺, Hg²⁺ and 2,4-dinitrotoluene.²⁴ Tao et al. reported that electrospun nanofibers of porphyrin-doped nanocomposites exhibited a sensitive response toward 2,4,6-trinitrotoluene vapor.²⁵ Wang et al. developed a strategy to immobilize H-PURET onto the electrospun membranes with a sensitive response to methyl viologen and cytochrome c by a layer-by-layer assembly approach.²⁶ Heng et al. reported an electrospun composite film with a lotus leaf like structure for Fe³⁺ and Hg²⁺ detection with a high sensitivity and reproducibility.²⁷ However, up to now, preparation of electrospun sensing materials always involves complicated polymer syntheses which make the fabrication time-consuming, and sometimes the leakage of sensing molecules affects the sensing performance. New approaches for the fabrication of nanofiber sensing materials are highly desirable to overcome the drawbacks. On the other hand, some researches have focused on fibers with secondary porous structures. Nanofibers with secondary porous structures can be produced from one-component polymers by choosing appropriate processing parameters such as humidity,^31–33 and from bicomponent systems by calcination or porogen treatment,^34–37 but these reports are very limited. More available systems and new fabrication approaches need exploring. Moreover, it is a great challenge to develop fluorescence sensing materials based on electrospun nanofibers with secondary porous structures.

Herein we successfully developed a new simple doping fabrication approach of electrospun nanofiber materials for fluorescence sensor devices by using a common polymer and fluorophore. The fluorophore, 9-chloromethylanthracene (9-CMA) was doped into the host matrix network of cellulose acetate (CA) rather than covalently attached to or assembled to the polymer before the electrospinning process. The sensing performance of the nanofiber membrane was evaluated by its quenching behavior towards methyl violet, a model quencher. The sensing behavior was substantially improved by introducing the secondary porous structures into nanofibers, which were produced by porogen or deacetylation treatment. The stability, reusability and reproducibility of the nanofiber materials were investigated as well. This is a report of a novel fluorescent sensor based on electrospun nanofibers with observed secondary structures and thus may contribute to the progress of fluorescence sensing materials.

Experimental

Chemicals

Cellulose acetate (CA) (M_w = 150 kDa) was purchased from Sinopharm Chemical Reagent Co. Ltd. (Beijing, China). 9-Chloromethylanthracene (9-CMA) was purchased from Acros (New Jersey, USA). Methyl violet (MV) and Triton X-100 (TX-100) were obtained from Beijing Chemical Reagent Company (Beijing, China). N,N-Dimethylacetamide (DMAc) was obtained from Beijing Yili Fine Chemical Company (Beijing, China). Polyethylene glycol-4000 (PEG-4000) was obtained from Tianjin Tiantai Fine Chemical Company (Tianjin, China). All reagents were of analytical grade and used as received without further purification. Deionized water was used throughout. Prior to use, the glass slides (1.4 cm × 3 cm) were treated with trimethylchlorosilane to make the surface hydrophobic.

Preparation of electrospun nanofiber sensing membranes and cast film

The preparation of sensing membranes was as follows. An electrospun solution was prepared by dissolving 12.5 wt% or 8.7 wt% CA in 1:2 (v/v) DMAc/acetone, and then doped with 9-CMA (the weight ratio of 9-CMA and CA was 1/10). The as-prepared solution was then loaded to the syringe. The electrospinning setup was shown in Fig. 1. A high-voltage power supply generated DC voltage up to 20 kV. The working distance between the syringe tip and collector was 15 cm. The electrospun solution was fed at a constant rate of 1.2 mL h⁻¹ by a syringe pump. Nanofibers were collected on glass slides mounted on aluminium foil with a collection time of 8 min, and then were dried in an oven at 60–70 °C for 12 h to remove the trace solvent.

Fig. 1 Schematic illustration of the electrospinning setup.

For comparison, the cast film was also prepared. The cast solution was prepared with the same method as that for the sensing nanofibers mentioned above. 0.5 mL of the solution was dropped onto glass slide and left to dry at 60–70 °C for 12 h to form a continuous thin film.

Preparation of nanofibers with secondary porous structures

The nanofibers with secondary porous structures were prepared by either introducing a porogen or deacetylation treatment. An electrospun solution was prepared by dissolving 12.5 wt% CA in 1:2 (v/v) DMAc/acetone, doped with 9-CMA (the weight ratio of 9-CMA and CA was 1/10) and then porogen, PEG-4000 or TX-100, was added. The weight ratio of PEG-4000 and CA was 1/2 and that of TX-100 and CA was 1/1. The as-prepared solution was electrospun into nanofibers under the same condition and treatment as those for electrospun nanofibers. The nanofiber membrane was then rinsed in water for 24 h to remove the porogen. On the other hand, deacetylation was conducted by treating the nanofiber membrane in 0.05 M NaOH aqueous solution for 12 h and thoroughly rinsing.

Characterization

The fluorescence measurements were performed using an F4500 fluorescence spectrophotometer (Hitachi, Tokyo, Japan). The emission spectra were obtained in the wavelength region of 380 to 600 nm with an excitation wavelength of 366 nm. The morphology of the nanofibers was observed using an S570 scanning electron microscope (SEM) (Hitachi, Tokyo, Japan) and JEM-100CX transmission electron microscope (TEM) (JEOL, Tokyo, Japan). The mass spectra were obtained using a ZAB-HS mass spectrometer (Micromass, England). The fluorescence images were taken with a DMLS fluorescence microscope (Leica, Germany). The excitation source was a high-voltage mercury lamp, and light with a wavelength of around 300–400 nm was emitted with the help of an optical filter. The exposure time was the same for all the images.

Results and discussion

Fabrication and characterization of sensing nanofibers

Cellulose acetate (CA) is commonly used in electrospinning and herein is accordingly employed as host matrix. Used as a fluorophore, 9-CMA was introduced to the electrospinning process. To obtain the optimal electrospinning conditions, several operation parameters were investigated including voltage, working distance and flow rate. Results showed that no significant variation in the morphology of nanofibers was observed within the tested range (15–25 kV voltage, 10–20 cm working distance, and 0.6–1.5 mL h⁻¹ flow rate). As a result, the moderate processing parameters (20 kV voltage, 15 cm working distance and 1.2 mL h⁻¹ flow rate) were used in the following procedures.

The effect of matrix CA concentration on the morphology of nanofibers was also studied. As shown in the SEM images (Fig. 2), the nanofibers prepared from the 12.5 wt% CA solution were uniform and smooth, with diameters in the range 200–600 nm (Fig. 2a), while those from the 8.7 wt % CA solution exhibited some distinctively unfavorable beads (Fig. 2b), indicating that the intrinsic properties of solutions influenced the morphology more significantly than the operation condition. Therefore, the concentration of CA played a major role in the nanofiber formation and the chosen concentration of CA was 12.5 wt% in the following experiments.

Fig. 2 SEM images of electrospun sensing materials produced from (a) 12.5 wt% and (b) 8.7 wt% CA solutions.

Compared with pure 9-CMA, 9-CMA doped electrospun nanofibers showed peaks at m/z of 226 (molecular ion peak of 9-CMA) and 191 ([C₁₅H₁₁]⁺ the fragment of 9-CMA) with distinctly reduced intensities, and a peak at m/z of 208 (molecular ion peak of 9-hydroxymethyl anthracene (9-HMA)) with a high intensity (data not shown). This indicates that the chlorine of 9-CMA is very active in nature and therefore 9-CMA in the nanofibers was converted to 9-HMA in a substantial percentage under the high voltage of electrospinning and the attack of water and alkali in the solvent and humidity in air.

Fabrication and characterization of sensing nanofibers with secondary porous structures

The secondary porous structures of the nanofibers were produced by two approaches, i.e. introducing a porogen or deacetylation treatment. For the porogen addition method, by introducing PEG-4000 or TX-100 to the electrospun solution followed by electrospinning, CA/PEG-4000 or CA/TX-100 nanofiber membranes were prepared. The porogens were removed by water due to their hydrophilicity and the sufficient removal was verified by IR (data not shown). For the other, deacetylation was carried out by treating the membrane with alkaline solution. The secondary porous structures generated by the above mentioned methods were observed by TEM as shown in Fig. 3. Considering the porous structures, the nanofiber membranes with intra-fiber pores should be 3D as well.

Fig. 3 TEM images of prepared electrospun sensing materials. Electrospun sensing nanofibers with PEG-4000 (a) before and (b) after washing, nanofibers with TX-100 (c) before and (d) after washing, and nanofibers (e) before and (f) after deacetylation treatment. (Scale bar: 200 nm).

The CA/PEG-4000 and CA/TX-100 systems were reported for the first time. As shown in Fig. 3a and Fig. 3c, the composite nanofibers before removal of the porogen were solid in morphology. After sufficient removal of the porogen, the fibers became porous with the 1D structure maintained. The intra-fiber pores were uniformly dispersed through the nanofibers, introducing the secondary porous structures which would be desirable for sensing materials. As shown in Fig. 3b and Fig. 3d, these nanofibers exhibited unique morphologies with the dispersity of the pores different from those reported in the literature.^35–37 Phase separation is commonly observed in component blends due to the change of mixing entropy. By selective removal of dispersed phase domains of electrospun nanofibers of component blends, intra-fiber pores can be generated accordingly.³⁸ There are limited reports on electrospun nanofibers with secondary porous structures produced by porogen treatment. Producing porosity by calcination limited the application of nanocomposites containing organic groups because the fluorophores might be decomposed during calcination. The removal of the porogen by water in our work can overcome this drawback as the experimental condition for generating porosity is mild and this approach might even inspire the development of porosity production.

To the best of our knowledge, deacetylation treatment has been seldom employed in porosity fabrication. Before deacetylation, the nanofibers were solid in morphology (Fig. 3e). By directly treating the sensing nanofiber membrane in a 0.05 M NaOH aqueous solution without other reagent addition, bubble-like porous structures can be observed by TEM (Fig. 3f). The bubble-like pores were distributed along through the nanofibers, and is the first example of porosity production by deacetylation treatment as a simple and applicable approach. The nanofibers after alkaline treatment showed a peak at m/z of 208 (molecular ion peak of 9-HMA) with high intensity, and the peak at m/z of 226 (molecular ion peak of 9-CMA) was almost undetectable (data not shown). This indicates that the majority of the residual 9-CMA in the nanofibers after electrospinning was converted to 9-HMA due to the nucleophilic attack of hydroxyl from the strong alkaline solution. The parent structure of anthracene was not altered, and accordingly no change in fluorescence property was observed. In one-component systems for fabrication of nanofibers with secondary porous structures, deacetylation may provide a more facile way compared with conventional humidity selection.

Sensing performance of electrospun nanofibers

The absorption and emission spectra of the sensing nanofiber membrane were investigated as shown in Fig. 4. It can be clearly seen that 9-CMA exists in nanofibers mainly as monomers.³⁹

Fig. 4 Absorption (dashed line) and emission (solid line) (λ_Ex = 366 nm) spectra of a 9-CMA doped nanofiber membrane.

The sensing performance of the electrospun nanofiber membrane without secondary structures was investigated and compared with its two analogues, i.e.membranes of nanofibers with beads (Fig. 2b) and cast films. Three glass slides coated with the different sensing membranes were placed in an optical cell, respectively. Upon introducing a series of MV solutions with concentrations from 10⁻⁶ to 10⁻⁴ M, fluorescence quenching of the sensing membrane was obviously observed. With the probe (9-CMA) as an electron donor, and the quencher (MV) as electron acceptor, the quenching process is based on a photo-induced electron transfer (PET) mechanism. The fluorescence intensity decreased with the increase of MV concentration (Fig. 5). F₀ and F are the fluorescence intensities in the absence and in the presence of the quencher, respectively. The nanofiber membrane exhibited a much higher sensitivity than any of other two membranes, which could be attributed to the differences in their morphologies. Nanofibers possess a 1D morphology, proper diameter and a relatively uniform distribution (Fig. 2a), which causes a more satisfactory porous structure and a higher surface area-to-volume ratio. These attractive advantages make the quencher easily diffuse into the surface and interior of fibrous membranes, and thus result in more efficient fluorescence quenching. On the contrary, the beads in the fibrous structure reduce the surface area⁴⁰ and consequently lead to the lower sensitivity. Besides, electrospun nanofibrous membranes provide 1 to 2 orders of magnitude higher surface area than that of cast films.^24,26,41 In consequence, compared with the prominent quenching of the sensing nanofiber membrane, the fluorescence of the cast film was less quenched. Nanofiber-based sensing materials are more sensitive than thin film based sensors. To conclude, fluorescent nanofiber membranes produced by the doping approach is simple in practice and provides higher sensitivity than nanofiber membranes with beads and cast film.

Fig. 5 Plots of F₀/F of fluorophore-doped sensing membranes against MV concentration. (■) Sensing nanofibers, (●) nanofibers with beads and (▲) cast film. (λ_Ex/λ_Em = 366/417 nm).

The sensing performace of nanofibers with secondary porous structures was further studied. The three membranes with secondary porous structures showed almost the same quenching sensitivities, which were approximately 3 fold compared with that of the membrane without secondary structures (Fig. 6). It has been clearly stated in the literature that the surface area of nanofibers could be greatly increased when the structure turned from solid to porous.⁷ By introducing a porogen or by a post-deacetylation treatment, the porosity was greatly increased while the 1D nanofibrous structure remained. Addition of a porogen (PEG-4000 or TX-100) to nanofibers and selective removal of it made the solid fibers porous, and by alkaline treatment, the nanofibers became bubble-like maintaining the 1D nanofibrous structure. The improved fluorescence quenching sensitivities undoubtedly benefited from these secondary porous structures, which provided increased surface areas, greatly facilitated quenchers to diffuse into nanofibers, accelerated the contact of quenchers with fluorophores, and thus effectively improved the sensing performance.

Fig. 6 Plots of F₀/F of (●) deacetylated, (■) PEG-4000 treated, (▲) TX-100 treated and (▼) untreated electrospun nanofibers against MV concentration. (λ_Ex/λ_Em = 366/417 nm).

The sensitivity improvement was further visualized by fluorescence microscopy. After treating the deacetylated membrane with a 10⁻⁴ M MV solution for 15 min, a remarkable quenching effect could be observed by fluorescence images (Fig. 7a and Fig. 7b). While the non-treated sensing membrane without secondary porous structures, after the same MV quenching process, showed a much less quenched fluorescence emission than that of the deacetylated membrane (Fig. 7c and Fig. 7d). Besides, the fluorescence images before the quenching process indicated the evident fluorescence emission and the uniform dispersion of fluorophores in CA (Fig. 7a and Fig. 7c), which is beneficial to the sensing performance.

Fig. 7 Fluorescence images of deacetylated sensing electrospun nanofibers (a) before and (b) after immersion in a 10⁻⁴ M MV solution for 15 min, and non-treated sensing nanofibers (c) before and (d) after immersion in a 10⁻⁴ M MV solution for 15 min.

The stability, reusability and reproducibility of the nanofiber materials were also investigated. The fluorescence intensity of the nanofiber membrane at 417 nm was investigated by time-scan mode for 4 h continuously (time interval is 1 s), indicating the good photostability of 9-CMA in nanofibers (Fig. 8). The fluorescence intensity of the sensing membrane remained unchanged during several months for experiments, indicating the firm adhesion of fluorophores in the nanofibers and the satisfactory stability. The fluorescence reusability of CA nanofibers was investigated by using a NaOH ethanol solution as eluent to regenerate the MV-quenched nanofibers. The CA nanofiber membrane was immersed in a 10⁻⁴ M MV solution and then washed with 0.05 M NaOH ethanol solution sequentially, and the quenching-regeneration cycle was repeated. The signal changes were recorded as shown in Fig. 9. In the tested 5 cycles, no obvious fluorescence loss was observed, indicating the excellent stability and reusability of the prepared fibrous membrane with no leakage of fluorophores in aqueous solutions. The relative standard deviation of quenching sensitivity among three batches of nanofiber membranes produced on different days was 0.3%, demonstrating that this nanofiber membrane had very good reproducibility as a sensing material (Table 1).

Fig. 8 Plot of fluorescence intensity of 9-CMA doped nanofiber membrane against the time. (λ_Ex/λ_Em = 366/417 nm).

Fig. 9 Repeated switching of fluorescence emission of the nanofiber membrane against the number of MV solution/eluent cycles.

Table 1 Reproducibility of the nanofiber sensing membranes^a

Batch	1	2	3	Average
a The mean of three replicates was used for each F0/F with relative standard deviation (RSD), where F0 and F represent fluorescence intensities in the absence and in the presence of the 10^−4 M MV, respectively.
F 0/F	1.84	1.84	1.85	1.84
RSD	2.4%	2.0%	1.9%	2.1%

---

Conclusions

This work provided a convenient and cost-effective fluorophore-doping approach for fast fabrication of electrospun fluorescent nanofiber materials for sensing applications. Novel secondary porous structures were fabricated successfully by a porogen or deacetylation treatment as a new approach to considerably improve the sensitivity. In addition, the prepared nanofibrous membranes showed excellent stability, reusability and reproducibility with no fluorophore leakage, and are suitable for sensing water-soluble species. The strategy for fabrication of secondary porous structures with attracting morphologies will enrich this research field and intrigue the development of sensors based on electrospun nanofibers with hierarchical structures. There is a wide selection of fluorphores combined with assorted membrane materials to facilitate the fabrication of fluorescent sensors with the desired sensing performance.

Acknowledgements

The authors thank Dr Y. Z. Xu's group of Peking University for obtaining the fluorescence images. This work was financially supported by the National Natural Science Foundation of China (20675003, 90713013 and 20275003) and Instrumental Analysis Fund of Peking University.

References

1. Y. N. Xia, P. D. Yang, Y. G. Sun, Y. Y. Wu, B. Mayers, B. Gates, Y. D. Yin, F. Kim and H. Q. Yan, Adv. Mater., 2003, 15, 353.
2. J. T. Hu, T. W. Odom and C. M. Lieber, Acc. Chem. Res., 1999, 32, 435.
3. X. B. Lu, J. H. Zhou, Y. H. Zhao, Y. Qiu and J. H. Li, Chem. Mater., 2008, 20, 3420.
4. A. L. Yarin, E. Zussman, J. H. Wendorff and A. Greiner, J. Mater. Chem., 2007, 17, 2585.
5. A. Greiner and J. H. Wendorff, Angew. Chem., Int. Ed., 2007, 46, 5670.
6. J. T. McCann, D. Li and Y. N. Xia, J. Mater. Chem., 2005, 15, 735.
7. D. Li and Y. N. Xia, Adv. Mater., 2004, 16, 1151.
8. C. Drew, X. Liu, D. Ziegler, X. Y. Wang, F. F. Bruno, J. Whitten, L. A. Samuelson and J. Kumar, Nano Lett., 2003, 3, 143.
9. X. F. Lu, Y. Y. Zhao, C. Wang and Y. Wei, Macromol. Rapid Commun., 2005, 26, 1325.
10. J.-S. Kim and D. H. Reneker, Polym. Compos., 1999, 20, 124.
11. M. M. Bergshoef and G. J. Vancso, Adv. Mater., 1999, 11, 1362.
12. O. D. Schneider, S. Loher, T. J. Brunner, P. Schmidlin and W. J. Stark, J. Mater. Chem., 2008, 18, 2679.
13. H. Yoshimoto, Y. M. Shin, H. Terai and J. P. Vacanti, Biomaterials, 2003, 24, 2077.
14. X. M. Mo, C. Y. Xu, M. Kotaki and S. Ramakrishna, Biomaterials, 2004, 25, 1883.
15. H.-J. Jin, J. S. Chen, V. Karageorgiou, G. H. Altman and D. L. Kaplan, Biomaterials, 2004, 25, 1039.
16. R. Chen and J. A. Hunt, J. Mater. Chem., 2007, 17, 3974.
17. http://www.donaldson.com/ .
18. T. E. Herricks, S.-H. Kim, J. Kim, D. Li, J. H. Kwak, J. W. Grate, S. H. Kim and Y. N. Xia, J. Mater. Chem., 2005, 15, 3241.
19. E. Formo, E. Lee, D. Campbell and Y. N. Xia, Nano Lett., 2008, 8, 668.
20. A. C. Patel, S. X. Li, C. Wang, W. J. Zhang and Y. Wei, Chem. Mater., 2007, 19, 1231.
21. Y. Wang and J. J. Santiago-Avilés, J. Appl. Phys., 2003, 94, 1721.
22. Y. X. Zhou, M. Freitag, J. Hone, C. Staii, A. T. Johnson, N. J. Pinto and A. G. MacDiarmid, Appl. Phys. Lett., 2003, 83, 3800.
23. Y. Zhao, J. Zhai, T. X. Wei, L. Jiang and D. B. Zhu, J. Mater. Chem., 2007, 17, 5084.
24. X. Y. Wang, C. Drew, S.-H. Lee, K. J. Senecal, J. Kumar and L. A. Samuelson, Nano Lett., 2002, 2, 1273.
25. S. Y. Tao, G. T. Li and J. X. Yin, J. Mater. Chem., 2007, 17, 2730.
26. X. Y. Wang, Y.-G. Kim, C. Drew, B.-C. Ku, J. Kumar and L. A. Samuelson, Nano Lett., 2004, 4, 331.
27. L. P. Heng, X. Y. Wang, Y. Q. Dong, J. Zhai, B. Z. Tang, T. X. Wei and L. Jiang, Chem. Asian J., 2008, 3, 1041.
28. A. C. Patel, S. X. Li, J.-M. Yuan and Y. Wei, Nano Lett., 2006, 6, 1042.
29. S. K. Chae, H. Park, J. Yoon, C. H. Lee, D. J. Ahn and J.-M. Kim, Adv. Mater., 2007, 19, 521.
30. Z. Y. Li, H. N. Zhang, W. Zheng, W. Wang, H. M. Huang, C. Wang, A. G. MacDiarmid and Y. Wei, J. Am. Chem. Soc., 2008, 130, 5036.
31. S. Megelski, J. S. Stephens, D. B. Chase and J. F. Rabolt, Macromolecules, 2002, 35, 8456.
32. C. L. Casper, J. S. Stephens, N. G. Tassi, D. B. Chase and J. F. Rabolt, Macromolecules, 2004, 37, 573.
33. M. Bognitzki, W. Czado, T. Frese, A. Schaper, M. Hellwig, M. Steinhart, A. Greiner and J. H. Wendorff, Adv. Mater., 2001, 13, 70.
34. D. Li and Y. N. Xia, Nano Lett., 2004, 4, 933.
35. S. Moon, J. Choi and R. J. Farris, Fiber. Polym., 2008, 9, 276.
36. M. Bognitzki, T. Frese, M. Steinhart, A. Greiner, J. H. Wendorff, A. Schaper and M. Hellwig, Polym. Eng. Sci., 2001, 41, 982.
37. C. Z. Chen, L. G. Wang and Y. Huang, Polymer, 2007, 48, 5202.
38. L. F. Zhang and Y. L. Hsieh, Nanotechnology, 2006, 17, 4416.
39. L. L. Li, H. Sun, Y. C. Bai, C. J. Fang and C. H. Yan, Chem.–Eur. J., 2009, 15, 4716.
40. T. Subbiah, G. S. Bhat, R. W. Tock, S. Parameswaran and S. S. Ramkumar, J. Appl. Polym. Sci., 2005, 96, 557.
41. P. Gibson, H. Schreuder-Gibson and D. Rivin, Colloids Surf., A, 2001, 187–188, 469.

---