Assembly of Fe₃O₄ nanoparticles on PEG-functionalized graphene oxide for efficient magnetic imaging and drug delivery

Abstract

Assembly of Fe₃O₄ nanoparticles (NPs) on PEG-functionalized graphene oxide for efficient magnetic imaging and drug delivery are reported. Nanoscale graphene oxide (GO) was first synthesized and functionalized by branched, biocompatible polyethylene glycol (PEG) to render high aqueous solubility and stability in physiological solutions. Next, the meso-2,3-dimercaptosuccinnic acid-modified Fe₃O₄ NPs were anchored onto GO sheets via formation of an amide bond in the presence of 1-ethyl-3-(3-dimethyaminopropyl)carbodiimide (EDC). It was found that the Fe₃O₄–PEG–GO nanocomposites possess good physiological stability and low cytotoxicity. Furthermore, controlled loading of aromatic drugs, doxorubicin (DOX), a typical anticancer drug, onto the Fe₃O₄–PEG–GO nanocomposites via π–π stacking and hydrophobic interactions is investigated. It is demonstrated that Fe₃O₄–PEG–GO/DOX loaded with the anticancer drug shows remarkably high cytotoxicity. The Fe₃O₄–PEG–GO/DOX nanocomposites show significantly enhanced cellular MRI, being capable of detecting cells at the iron concentration of 10 μg mL⁻¹ with cell density of 2 × 10⁵ cells per mL. The current work provides a general approach toward rational design and synthesis of a versatile theranostic nanoplatform based on Fe₃O₄–PEG–GO NPs with good biocompatibility and the capability of simultaneously performing imaging and therapy for clinical outcomes.

---

1. Introduction

The term “theranostic” is defined as a material that combines the modalities of therapy and diagnostic imaging. The field of theranostics remains relatively young but is developing at an incredibly fast rate. Researchers have created numerous nanoparticle-based systems that can be used for therapy and diagnostic.^1–5 Among various all-in-one theranostic nanoparticles (NPs), Fe₃O₄ NPs have emerged as one of most appealing candidates. Fe₃O₄ NPs are widely used in biomedical fields such as magnetic resonance imaging (MRI), biological separation, and hyperthermia therapy.^4,6–8

It is reported that formation of aggregates of Fe₃O₄ NPs leads to enhanced relaxation rate (r₂).^9–14 Ai and colleagues¹⁰ found that the T₂ relaxivities of Fe₃O₄-loaded polymeric micelles (r₂ = 169 Fe mM⁻¹ s⁻¹) were significantly larger than that of the dextran-coated Fe₃O₄ NPs (30–50 Fe mM⁻¹ s⁻¹). Liu et al.¹¹ reported that under a magnetic field of 3 T, multiple Fe₃O₄ NPs containing micelles have an r₂ up to 345 Fe mM⁻¹ s⁻¹, 4-fold of single Fe₃O₄ NP-containing micelles (84 Fe mM⁻¹ s⁻¹). And our team reported that the formation of aggregates of Fe₃O₄ NPs onto graphene oxide (GO) exhibit a considerable enhanced T₂ relaxivity compared to the isolated Fe₃O₄ NPs.¹⁴

There are some reports on formation of composites of GO and magnetic nanoparticles and their applications in MRI, magnetically targeted drug delivery, removal of contaminants from waste water.^14–21 Cong et al.¹⁵ reported for the first time the synthesis and MRI effect of composites of Fe₃O₄ NPs and chemically reduced GO. Here, we report that the Fe₃O₄–PEG–GO nanocomposites show significantly enhanced cellular MRI, being capable of detecting cells at the low iron concentration. GO as a nanocarrier can load anticancer drugs via noncovalent physisorption for biomedical applications, Chen and colleague investigated the loading and release of doxorubicin (DOX), an anticancer drug, on the GO. Their work demonstrates that GO can be used as efficient nanocarriers for the loading and delivery of water-insoluble aromatic drugs.²² Based on the above advantages of GO, assembly of Fe₃O₄ nanoparticles onto PEG-functionalized GO for efficient magnetic imaging and drug delivery is reported. Most of the reports are based on in situ formation of magnetic GO nanoparticles,^23,24 we have previously reported a method that magnetic nanoparticles were coupled onto GO by EDC condensation,²⁵ this paper carried out further work, that the anticancer drugs were loaded onto the nanocomposites for cancer therapies, formation of a new type of diagnosis-treatment integrated nanomaterials.

In our strategy, biocompatible PEG with amino groups at both ends was introduced to GO, which rendered it stable under physiological conditions, then carboxyl group-modified Fe₃O₄ NPs were anchored onto above PEGylated GO sheets, via formation of amide bond by EDC condensation reaction. At last, the anticancer drugs DOX were attached to the Fe₃O₄–PEG–GO nanocomposites for cancer treatment. The Fe₃O₄–PEG–GO/DOX nanocomposites enhanced magnetic resonance T₂ relaxation rate. Thereby, improving biological imaging sensitivity provides guidance in clinical diagnosis. Moreover, the nanocomposites system is not only a good contrast agent for MRI, but also can serve as an ideal nanoplatform for theranostics.

2. Experimental section

2.1 Materials

Native graphite flake was purchased from Alfa Aesar. meso-2,3-Dimercaptosuccinnic acid (DMSA), dimethylsulfoxide (DMSO), 1-ethyl-3-(3-dimethyaminopropyl)carbodiimide (EDC), and N-hydroxysuccinimide (NHS), and oleic acid were purchased from Sigma-Aldrich. 10k PEG (NH₂–PEG_10 000–NH₂) was purchased from Jenkem Technology. RPMI 1640 culture medium and fetal bovine serum were purchased from Invitrogen. WST-1 was purchased from Biyuntian Biotechnology Institute. All other reagents were analytical grade and used as received. Ultrapure water (18.2 MΩ cm⁻¹) was used in all experiments.

2.2 General characterization

The morphology and composition of NPs were characterized by Tecnai G² F20 S-Twin transmission electron microscopy (TEM) equipped with an energy dispersive spectrometry. The iron concentration was measured using atomic absorption spectrometry (SpectrAA-Duo 220 FS, Varian). WST assay was performed with a Biotek Elx 800 Microplate Reader. Cell lines were cultured in a water-jacketed CO₂ incubator (Thermo 3111). The MRI was carried out on a 11.7 T Bruker micro 2.5 micro-MRI system with a conventional spin-echo acquisition. Relaxivity (r₂) with unit mM⁻¹ s⁻¹ was calculated through the curve-fitting of the reciprocal of the relaxation time versus the iron concentration (mM Fe). Magnetic measurement was carried out on a Magnetic Property Measurement System (MPMS SQUID VSM, Quantum Design).

2.3 Preparation of DMSA-coated Fe₃O₄ NPs

Ten nm monodisperse Fe₃O₄ NPs were synthesized by thermal decomposition of an iron oleate precursor according to the known method.²⁶ Thus-prepared Fe₃O₄ NPs were then coated with DMSA following the approach by Chen et al.²⁷ In brief, 1.08 g of FeCl₃·6H₂O and 3.65 g of sodium oleate were dissolved in a mixture solvent containing 8 mL ethanol, 6 mL distilled water, and 14 mL hexane. The solution was heated to 70 °C and stirred for 4 hours. Then the upper red-brownish organic layer was separated, washed with ultrapure water in a separatory funnel and dried overnight. The obtained 2.8 g waxy iron oleate was dissolved in 20 mL octadecanol and 0.5 mL oleic acid at 70 °C. The mixture was heated to 320 °C for 30 min, then the precipitate containing OA-coated Fe₃O₄ nanoparticles was obtained. 20 mg of the Fe₃O₄ NPs was dissolved in 2 mL toluene, then 20 mg DMSA and 2 mL DMSO was added to the solution and stirred for 12 hours. This precipitant was then washed with ethyl acetate. Finally, the DMSA-coated Fe₃O₄ NPs were transferred into 2 mL ultrapure water.

2.4 Synthesis of Fe₃O₄–PEG–GO nanocomposites

Water soluble GO was prepared according to the procedure described in our previous work.²⁸ 5 mL GO (0.5 mg mL⁻¹) was adjusted to pH 8.0 with triethylamine, to which 40 mg of EDC, 40 mg of NHS, and 2 mL of PEG (NH₂–PEG_10 000–NH₂) (1 mg mL⁻¹) were added. The reaction solution was stirred at room temperature for 48 h, and the obtained raw product, GO–PEG NPs, was transferred to Millipore centricon (100 000 molecular mass cut-off), then washed three times by ultrapure water. For preparation of Fe₃O₄–PEG–GO nanocomposites, 1.5 mL of the GO–PEG NPs (0.5 mg mL⁻¹), 40 mg of EDC, 40 mg of NHS, and 2 mL of the DMSA-coated Fe₃O₄ NPs (1.5 mg mL⁻¹) were mixed and stirred overnight. The Fe₃O₄–PEG–GO nanocomposites were obtained after centrifugation and through washing with ultrapure water.

2.5 Controlled loading and release of DOX

Loading of a single drug on the Fe₃O₄–PEG–GO nanocomposites was carried out by adding DOX to the water soluble Fe₃O₄–PEG–GO nanocomposites with stirring for 24 h. The result solution was then dialyzed against ultrapure water to remove free drug. The loading ratio of DOX was estimated from the absorbance at 490 nm in the UV/Vis spectra after subtracting the absorption contribution from the Fe₃O₄–PEG–GO nanocomposites.

2.6 Cell line and cell culture

The HeLa cell line (human cervical carcinoma cell) was kindly provided by Professor Haiyan Liu, Soochow University. Cells were cultured in RPMI 1640 medium with 10% fetal bovine serum and incubated in a humidified atmosphere at 37 °C with 5% CO₂.

2.7 WST assay

WST assay was performed to evaluate the cytotoxicity of Fe₃O₄–PEG–GO/DOX nanocomposites. The cells were seeded in 96-well plates at a density of 1 × 10⁴ cells in 100 μL culture medium and maintained for 24 h. Then, cells were incubated for 24 h with the Fe₃O₄–PEG–GO/DOX nanocomposites at different DOX concentration of 0.25 μg mL⁻¹, 0.5 μg mL⁻¹, 1 μg mL⁻¹, and 2 μg mL⁻¹ respectively, then washed with PBS buffer for three times and added fresh RPMI 1640 medium supplemented with 10% fetal bovine serum. The relative cellular viability was examined by the WST assay. The data were presented as mean ± SD.

2.8 Cell labeling

The minimum detectable iron concentration was obtained by incubation of separate batches of cells (2 × 10⁵ cells per mL) at different iron concentration of 2.5, 5, 10, 20, and 40 μg mL⁻¹ for 24 h, respectively. Imaging parameters of T₂-weighted images (T₂WI) by multi slice multi echo (MSME) experiments on 11.7 T Bruker micro 2.5 micro-MRI system with repetition time (TR) = 2500 ms, echo time (TE) = 90 ms, imaging matrix = 128 × 128, slice thickness = 1 mm, and file of vision (FOV) = 2.0 cm × 2.0 cm.

3. Results and discussion

3.1 Synthesis and characterization of the Fe₃O₄–PEG–GO nanocomposites

The thickness of prepared GO sheets was at 1–2 nm, and the lateral width was at 50–250 nm (Fig. S1†). The morphology and size of the oleic acid-coated Fe₃O₄ NPs and Fe₃O₄–PEG–GO nanocomposites were characterized by TEM (Fig. 1). The TEM data indicate that the as-prepared oleic acid coated Fe₃O₄ NPs are uniform cubic NPs with size of 10 nm. It is clearly seen from Fig. 1 (right) that the Fe₃O₄ NPs were assembled and formed aggregates onto the GO surface.

Fig. 1 TEM images of the oleic acid coated Fe₃O₄ NPs (left) and Fe₃O₄–PEG–GO nanocomposites (right).

Fig. 2 depicts FTIR spectra of GO (red line) and GO–PEG (black line), two characteristic peaks at 2857 cm⁻¹ and 2924 cm⁻¹ corresponds to symmetric and asymmetric stretching modes of CH₂ of GO, which shifts to 2874 cm⁻¹ and 2918 cm⁻¹ for the GO–PEG confirming successful covalent linking of PEG to GO. The peaks at 1460 cm⁻¹ assignable to C–N stretching mode of the GO–PEG, also suggesting PEG successful covalently conjugated to the GO via formation of amide bond between carboxylic acid groups of GO and amine groups of PEG.

Fig. 2 FT-IR spectra of GO and GO–PEG.

With vibration sample magnetometer (VSM), their magnetization behavior was characterized. A typical plot of magnetization versus applied magnetic field (magnetic hysteresis loop) was shown in Fig. S2.† The saturation magnetization of the DMSA–Fe₃O₄ NPs is determined to be 67.27 emu g⁻¹, much higher than Fe₃O₄–PEG–GO nanocomposites, being 43.23 emu g⁻¹, due to decrease of magnetic component in the nanocomposites. To investigate the effect of aggregation of Fe₃O₄ NPs on the MRI property, we compared the MRI intensities of the DMSA–Fe₃O₄ NPs and Fe₃O₄–PEG–GO nanocomposites. T₂ relaxivity of the Fe₃O₄–PEG–GO nanocomposites is determined to be 98.1 Fe μg mL⁻¹ s⁻¹, much higher than that of the DMSA–Fe₃O₄ NPs (r₂ = 79 Fe μg mL⁻¹ s⁻¹) (Fig. S3†), indicative of enhanced T₂ relaxivity caused by formation of Fe₃O₄ aggregates.

3.2 Loading and releasing of anticancer drugs

DOX are widely used in the clinic for cancer treatment and showed much better anticancer effect. The amount of the drug loaded was estimated by absorbance at 490 nm for DOX. By the way of physical adsorption, the DOX was adsorbed to Fe₃O₄–PEG–GO nanocomposites. DOX has a distinct absorption peak at 547 nm, as shown from the UV/Vis absorption spectra in Fig. 3a, but the maximum absorption peak was obviously red shift. The red shift of DOX absorption peak may be because of the strong molecular interaction between GO and DOX. In this thesis, the adsorption behavior of DOX on GO was further studied by fluorescence spectra. From the fluorescence spectra of Fe₃O₄–PEG–GO/DOX nanocomposites, it can be seen that DOX was excited at 488 nm, the fluorescence quenching was almost completely (Fig. 3b). This is mainly because of the presence of electrons and energy transfer between GO and DOX. When the Fe₃O₄–PEG–GO/DOX was eluted with ethanol, the emission peak at 554 nm and 588 was assignable to the eluted DOX, ethanol can release DOX from Fe₃O₄–PEG–GO/DOX nanocomposites, thus enabling fluorescence recovery. We measured the concentration of DOX at 490 nm by UV/Vis absorption spectrophotometer. The drug loading rate of Fe₃O₄–PEG–GO/DOX nanocomposites was about 95%. A group led by Chen investigated the loading and release of DOX on GO,²² they found that the loading ratio (the weight ratio of loaded drug to carriers) of GO could reach 200%. In contrast, our drugs loading rate is lower, it may be because of the Fe₃O₄ loaded onto the GO, occupying certain space, affecting the adsorption of DOX on the Fe₃O₄–PEG–GO/DOX nanocomposites.

Fig. 3 (a) UV/Vis spectra of the DOX and Fe₃O₄–PEG–GO/DOX in aqueous solution. (b) Fluorescence spectra of aqueous solution of Fe₃O₄–PEG–GO/DOX before (black line) and after (red line) adding of ethyl alcohol (final concentration of ethyl alcohol is 50%). Excitation at 480 nm.

3.3 Cytotoxicity of the Fe₃O₄–PEG–GO composites

We examined the physiological stability of the nanocomposites, and we found that the nanocomposites were still well dispersed in RPMI 1640 medium with 10% fetal bovine serum, a physiological solution, after stored for 24 h. The aggregation and precipitation of the nanocomposites were not found under the microscope (Fig. S4†). The molar ratio of Fe and DOX control in 20 : 1, in order to make the Fe₃O₄–PEG–GO/DOX nanocomposites not only have antitumor activity, but also effective cellular imaging function. The antitumor activity of the Fe₃O₄–PEG–GO/DOX and Fe₃O₄–PEG–GO nanocomposites was determined by WST assay. It was found that the Fe₃O₄–PEG–GO nanocomposites showed very low cytotoxicity even at iron concentration up to 40 μg mL⁻¹, the Fe₃O₄–PEG–GO/DOX nanocomposites showed obvious cytotoxicity at iron concentration of 20 μg mL⁻¹ (the concentration of DOX is 1 μg mL⁻¹), when the iron concentration reached to 40 μg mL⁻¹ (the concentration of DOX is 1 μg mL⁻¹), the relative survival rate of HeLa cells was 76%, there was a significant difference compared to the low concentration (Fig. 4). All these results suggest that the Fe₃O₄–PEG–GO/DOX nanocomposites employed in our work are practically cytotoxic, which is ideal for their application in biomedical applications such as cellular MRI contrast agent, antitumor drugs.

Fig. 4 Relative viability of HeLa cells incubated with the Fe₃O₄–PEG–GO/DOX (black) and Fe₃O₄–PEG–GO (red) at the different DOX concentration for 24 h, respectively. Error bars were based on quartet samples.

3.4 In vitro MR imaging

In order to observe the MRI labeling behavior of the Fe₃O₄–PEG–GO/DOX nanocomposites, we used the HeLa cells incubated with various concentrations of nanocomposites, and then the labeled HeLa were moved to the magnetic tube, scanned by MRI. In our experiment, it was found that the cells incubated with nanocomposites enhance the cell-to-background contrast and makes them visible in MR images (Fig. 5). The nanocomposites showed relaxation rate enhancement, depending on the dose density and was capable of detecting cells at the iron concentration of 10 μg mL⁻¹ with cell density of 2 × 10⁵ cells per mL. The Fe₃O₄–PEG–GO/DOX nanocomposites show significantly enhanced cellular imaging, capability of simultaneously performing imaging and therapy for clinical outcomes.

Fig. 5 The T₂ weighted MR images: HeLa cells (2 × 10⁵ cells per mL) incubated with the Fe₃O₄–PEG–GO/DOX composites at different concentrations for 24 h. The imaging parameters with 11.7 T magnet system: T_R = 2500 ms, T_E = 90 ms, imaging matrix = 128 × 128, slice thickness = 1 mm, FOV = 2.0 cm × 2.0 cm.

4. Conclusions

We have developed a strategy for preparation of Fe₃O₄–PEG–GO nanocomposites as efficient MRI contrast agent. The nanocomposites were synthesized by coupling reaction of the DMSA-coated Fe₃O₄ NPs with PEG-modified GO via EDC chemistry. Thus-prepared magnetic composites exhibit good physiological stability and did not affect the cellular viability and proliferation. More importantly, compared to the corresponding Fe₃O₄ NPs, the Fe₃O₄–PEG–GO nanocomposites exhibit significantly improved T₂ weighted MRI contrast, which is explained by that the Fe₃O₄ NPs formed aggregates on the GO sheets, resulting in a considerable enhanced T₂ relaxivity. This work demonstrated potential application of Fe₃O₄–PEG–GO nanocomposites as biocompatible and efficient cellular MRI contrast agent. Using physical adsorption, the anticancer drugs doxorubicin was loaded onto Fe₃O₄–PEG–GO nanocomposites for oncotherapy, the drug loading rate is higher (95%). WST cytotoxicity experiments demonstrated the efficient anticancer effect using the Fe₃O₄–PEG–GO/DOX nanocomposites. Further work on in vivo cancer treatment by the Fe₃O₄–PEG–GO nanocomposites is desired for their clinical application.

Acknowledgements

Financial support of this work by Natural Science Foundation of China (No. 21405074, No. 21375057), Shandong Province Natural Science Foundation (No. ZR2013BL007), and National undergraduate training programs for innovation and entrepreneurship (No. 201410452006) are gratefully acknowledged.

Notes and references

1. K. Kaaki, K. Hervé-Aubert, M. Chiper, A. Shkilnyy, M. Soucé, R. Benoit, A. Paillard, P. Dubois, M.-L. Saboungi and I. Chourpa, Langmuir, 2012, 28, 1496–1505.
2. S. S. Kelkar and T. M. Reineke, Bioconjugate Chem., 2011, 22, 1879–1903.
3. T. Lammers, S. Aime, W. E. Hennink, G. Storm and F. Kiessling, Acc. Chem. Res., 2011, 44, 1029–1038.
4. C. Sanson, O. Diou, J. Thévenot, E. Ibarboure, A. Soum, A. Brûlet, S. Miraux, E. Thiaudière, S. Tan, A. Brisson, V. Dupuis, O. Sandre and S. b. Lecommandoux, ACS Nano, 2011, 5, 1122–1140.
5. J. Shi, A. R. Votruba, O. C. Farokhzad and R. Langer, Nano Lett., 2010, 10, 3223–3230.
6. T. Chen, M. I. Shukoor, R. Wang, Z. Zhao, Q. Yuan, S. Bamrungsap, X. Xiong and W. Tan, ACS Nano, 2011, 5, 7866–7873.
7. Y. M. Huh, Y. W. Jun, H. T. Song, S. Kim, J. S. Choi, J. H. Lee, S. Yoon, K. S. Kim, J. S. Shin, J. S. Suh and J. Cheon, J. Am. Chem. Soc., 2005, 127, 12387–12391.
8. J. Liu, S. Z. Qiao, Q. H. Hu and G. Q. Lu, Small, 2011, 7, 425–443.
9. H. Jaganathan, D. L. Hugar and A. Ivanisevic, ACS Appl. Mater. Interfaces, 2011, 3, 1282–1288.
10. H. Ai, C. Flask, B. Weinberg, X. T. Shuai, M. D. Pagel, D. Farrell, J. Duerk and J. Gao, Adv. Mater., 2005, 17, 1949–1952.
11. G. Liu, Z. Y. Wang, J. Lu, C. C. Xia, F. B. Gao, Q. Y. Gong, B. Song, X. N. Zhao, X. T. Shuai, X. Y. Chen, H. Ai and Z. W. Gu, Biomaterials, 2011, 32, 528–537.
12. J. Lu, S. L. Ma, J. Y. Sun, C. C. Xia, C. Liu, Z. Y. Wang, X. N. Zhao, F. B. Gao, Q. Y. Gong, B. Song, X. T. Shuai, H. Ai and Z. W. Gu, Biomaterials, 2009, 30, 2919–2928.
13. Z. Y. Wang, G. Liu, J. Sun, B. Y. Wu, Q. Y. Gong, B. Song, H. Ai and Z. W. Gu, J. Nanosci. Nanotechnol., 2009, 9, 378–385.
14. X. Y. Yang, Y. S. Wang, X. Huang, Y. F. Ma, Y. Huang, R. C. Yang, H. Q. Duan and Y. S. Chen, J. Mater. Chem., 2011, 21, 3448–3454.
15. H. P. Cong, J. J. He, Y. Lu and S. H. Yu, Small, 2010, 6, 169–173.
16. F. A. He, J. T. Fan, D. Ma, L. M. Zhang, C. Leung and H. L. Chan, Carbon, 2010, 48, 3139–3144.
17. J. F. Shen, Y. Z. Hu, M. Shi, N. Li, H. W. Ma and M. X. Ye, J. Phys. Chem. C, 2010, 114, 1498–1503.
18. V. K. Singh, M. K. Patra, M. Manoth, G. S. Gowd, S. R. Vadera and N. Kumar, New Res. Carbon Mater., 2009, 24, 147–152.
19. T. Szabó, A. Bakandritsos, V. Tzitzios, E. Devlin, D. Petridis and I. Dékány, J. Phys. Chem. B, 2008, 112, 14461–14469.
20. Y. Zhang, B. Chen, L. M. Zhang, J. Huang, F. H. Chen, Z. P. Yang, J. L. Yao and Z. J. Zhang, Nanoscale, 2011, 3, 1446–1450.
21. H. He and C. Gao, ACS Appl. Mater. Interfaces, 2010, 2, 3201–3210.
22. X. Y. Yang, X. Y. Zhang, Z. F. Liu, Y. F. Ma, Y. Huang and Y. S. Chen, J. Phys. Chem. C, 2008, 112, 17554–17558.
23. J. M. Shen, G. Huang, X. Zhou, J. Zou, Y. Yang, Y. F. Chen and S. K. Men, RSC Adv., 2014, 4, 50464–50477.
24. J. M. Shen, F. Y. Gao, L. P. Guan, W. Su, Y. J. Yang, Q. R. Li and Z. C. Jin, RSC Adv., 2014, 4, 18473–18484.
25. W. H. Chen, P. W. Yi, Y. Zhang, L. M. Zhang, Z. W. Deng and Z. J. Zhang, ACS Appl. Mater. Interfaces, 2011, 3, 4085–4091.
26. N. R. Jana, Y. Chen and X. Peng, Chem. Mater., 2004, 16, 3931–3935.
27. Z. P. Chen, Y. Zhang, S. Zhang, J. G. Xia, J. W. Liu, K. Xu and N. Gu, Colloids Surf., A, 2008, 316, 210–216.
28. L. M. Zhang, J. G. Xia, Q. H. Zhao, L. W. Liu and Z. J. Zhang, Small, 2009, 6, 537–544.

---

Footnotes

† Electronic supplementary information (ESI) available: AFM image for GO, photographs for Fe₃O₄–PEG–GO, etc. See DOI: 10.1039/c5ra09901c

‡ Both authors contributed equally to this work.

---