Controlled assembly of Fe₃O₄ magnetic nanoparticles on graphene oxide

Abstract

We describe a facile approach to controllable assembly of monodisperse Fe₃O₄ nanoparticles (NPs) on chemically reduced graphene oxide (rGO). First, reduction and functionalization of GO by polyetheylenimine (PEI) were achieved simultaneously by simply heating the PEI and GO mixture at 60 °C for 12 h. The process is environmentally friendly and convenient compared with previously reported methods. Meso-2,3-dimercaptosuccinnic acid (DMSA)-modified Fe₃O₄ NPs were then conjugated to the PEI moiety which is located on the periphery of the GO sheets via formation of amide bonds between COOH groups of DMSA molecules bound on the surface of the Fe₃O₄ NPs and amine groups of PEI. The magnetic GO composites were characterized by means of TEM, AFM, UV-vis, FTIR, Raman, TGA, and VSM measurements. Finally, preliminary results of using the Fe₃O₄-rGO composites for efficient removal of tetracycline, an antibiotic that is often found as a contaminant in the environment, are reported.

---

Integration of nanoparticles (NPs) and graphene into nanocomposites has recently become a hot topic of research due to their new and/or enhanced functionalities that cannot be achieved by either component alone, and therefore holds great promise for a wide variety of applications in catalysis, optoeletronic materials, surface enhanced Raman Scattering, biomedical fields, and so on.^1–10 Among them are nanocomposites of magnetic NPs (MNPs) and graphene oxide (GO) (Fe₃O₄-GO), for potential applications in enhanced optical limiting, MRI, drug delivery, and removal of contaminants from waste water.^8–11 Formation of Fe₃O₄-GO nanocomposites is usually achieved by in situreduction of iron salt precursors^12–14 or assembly of the MNPs on the GO surface,¹¹ similarly to its carbon nanotube counterpart.^15,16 Cong et al. reported for the first time depositing of magnetic nanoparticles on a chemically reduced GO and application of the magnetic functionalized rGO in MRI.⁹ Most recently He and colleagues attached iron oxide NPs to GO viaconjugation chemistry and explored their application for removal of methylene blue from waste water.¹¹ There are, however, some problems with the approaches reported so far, such as poor control over the size, size distribution, and location of magnetic nanoparticles on the GO sheets. In addition, thus-prepared Fe₃O₄-GO composites may agglomerate and gradually precipitate from their aqueous suspension.⁹ All these drawbacks make the use of magnetic GO in biomedical and other fields difficult.

Herein we describe a facile strategy for assembly of monodisperse Fe₃O₄ NPs onto a polyetheylenimine (PEI)-grafted reduced GO sheet, the schematic of which is illustrated in Fig. 1. In this straightforward route, the Fe₃O₄ NPs were firstly prepared viathermal decomposition of Fe precursors and then modified by hydrophilic 2,3-dimercaptosuccinnic acid (DMSA), followed by assembling the DMSA-modified Fe₃O₄ NPs onto PEI-grafted GO sheets. There are two significant advantages of our approach over other previously reported ones: (1) well controlled size, size distribution, and morphology of the Fe₃O₄ NPs in the Fe₃O₄-rGO composites. This is readily achieved by first preparing the monodispersed Fe₃O₄ NPs by well established synthetic procedures and then assembly onto the GO surface. (2) controlled decoration of Fe₃O₄ NPs on the edge of GO sheets, which allows us to efficiently load aromatic drugs or to adsorb aromatic pollutants from waste water, on the basal plane of the GO sheet via π–π stacking, the latter will be discussed later in this paper. To our knowledge, this may be the first report on a facile and general synthesis route for NPs-GO composites by which NPs with precisely-controlled size and morphology can be deposited onto specific sites of the GO sheets.

Fig. 1 Schematic diagram of synthesis of Fe₃O₄-rGO composites.

In our approach, 6 nm Fe₃O₄ NPs and 18 nm Fe₃O₄ NPs were prepared by thermal decomposition of Fe(acac)₃ using a well-established procedure¹⁷ and surface modified by DMSA¹⁸ to allow further covalent bonding to GO. TEM images in Fig. S1† indicate that the as-prepared Fe₃O₄ NPs are uniform particles with sizes of 6 nm and 18 nm, respectively.

Water soluble GO was prepared from graphite according to the modified Hummers method reported in our previous work.¹⁹ 40 mL aqueous suspension of GO (0.1mg mL⁻¹) was then mixed with 40 mL PEI (1% aqueous solution) and heated to 60 °C for 12 h.²⁰ The color of the mixture changed gradually from brown to black during the heating process, indicating reduction of GO to reduced GO (rGO) during heating with PEI. It was reported that GO can be reduced to rGO in the presence of amine-containing molecules.²⁰ After purification by centrifugation and dialysis, the product was extensively characterized by AFM, UV-vis, FTIR, Raman, and electrochemical measurements. AFM measurement show that the thickness of GO changed from 1.79 nm to 0.815 nm (Fig. 2) upon heating of the mixture of GO and PEI. It is reasonable that reduction of GO to rGO by PEI, that is, removal of oxygen-containing groups such as epoxide from both sides of the GO sheet, leads to a decrease in GO thickness. The characteristic peak at 231 nm for GO, was shifted to 266 nm in the UV-vis spectra (Fig. S2†), suggesting reduction of GO and restoration of the large electronic conjugation.^21,22 Significant shift of the D band from 1336 cm⁻¹ to 1322 cm⁻¹ and the G band from 1603 cm⁻¹ to 1591 cm⁻¹ in the Raman spectra (Fig. S3†), and an increase of intensity ratio of the D band to G band (I_D/I_G) from 1.26 to 1.53 are typical features of reduction of GO to rGO.^23,24 From IR spectra (Fig. S4†), substantial increase in two peaks at 2852 cm⁻¹ and 2923 cm⁻¹ assignable to symmetric and asymmetric stretching modes of CH₂ of the PEI chains, and dramatic decrease in the peak at 1730 cm⁻¹ due to C=O of COOH-GO suggests reduction of the GO and chemical bonding of PEI to the rGO sheets.^25,26 The zeta potential of the original GO is −39.7 mV, which was shifted to 55.8 mV when the GO was heated with PEI, indicative of formation of PEI-rGO complex and introduction of positive charges into initially negatively charged GO sheets. The PEI-rGO complex obtained dispersed well in water and was stable when stored in ambient conditions for 3 months. Agglomeration and precipitation usually occurs when GO is chemically reduced to rGO.²⁷ Based on the discussion above, the role of the PEI in this work is threefold: reducing GO to rGO, stabilizing the chemically reduced GO in aqueous solution, and acting as a linker molecule for anchoring the MNPs on the GO sheet.

Fig. 2 AFM images of GO and PEI-grafted GO sheets.

Next, formation of Fe₃O₄-rGO nanocomposites was carried out by chemical bonding of PEI-rGO and DMSA-modified Fe₃O₄ NPs in the presence of 1-ethyl-3-(3-dimethyaminopropyl) carbodiimide (EDC) and N-hydroxysuccinnimide (NHS) (Fig.1). The composition, morphologies, and magnetic properties of the Fe₃O₄-rGO composites were extensively characterized by TEM, EDX, FTIR, TGA, and VSM measurements. Fig. 3 shows TEM images of the 6nm Fe₃O₄-rGO and 18nm Fe₃O₄-rGO composites. The monodisperse Fe₃O₄ NPs (Fig. S1†) were preferentially decorated to the periphery of the graphene surface. This can be explained as follows. The COOH groups are thought to populate along the edge of the GO sheets in the original GO,²⁸ which serves as a template for PEI grafting, and then PEI guided assembly of Fe₃O₄ NPs occurs on the outer boundary of the graphene sheets. It is also noted that some Fe₃O₄ NPs are located on the basal plane due to the irregular distribution of the carboxylic acid groups on the GO sheets caused by the vigorous oxidation of graphite. Typical EDX pattern (inset in the left TEM image of Fig. 3) indicates that the 6nm Fe₃O₄-rGO composites contain Fe, O, and C elements, while the Cu element comes from the copper grids. From TGA data shown in Fig. S5,† slight weight loss (5.2%) at temperature below 120 °C was observed, due to loss of water or other volatile residuals. Further heating of the 6nm-Fe₃O₄-rGO from 120 °C to 550 °C results in 25.3% weight loss, mainly due to removal of oxygen-containing functional groups such as COOH, OH, C=O, and epoxide of the GO sheets. The amount of Fe₃O₄ component in the 6nm-Fe₃O₄-rGO composites was thus estimated to be 61% from the TGA data.^10,23 This ratio is rather high compared with other magnetic composites, and may be beneficial for their applications in various fields.¹⁰Fig. 4 gives magnetic hysteresis loops for 6nm Fe₃O₄-rGO composites measured at 298 K. The profiles of the magnetization curves of the 6 nm Fe₃O₄ and the corresponding 6nm-Fe₃O₄-rGO composites are characteristic of superparamagnetic nanoparticles. The saturation magnetization of the 6 nm Fe₃O₄ NPs in our work is significantly lower when compared with their bulk counterpart, which was also observed previously by others and can be explained by size-dependent magnetization of the magnetic NPs and the presence of organic ligands on the surface of the Fe₃O₄ NPs.^17,29,30 The 6 nm Fe₃O₄ NPs loaded on the rGO support showed a decrease in saturation magnetization (from 39 emu g⁻¹ to 28 emu g⁻¹), due to a decrease in the amount of magnetic component in the composites,^9,10 but it is still very high compared with that of the Fe₃O₄-rGO obtained viaassembly of surface modified Fe₃O₄ NPs on GO in previous work.¹¹

Fig. 3 TEM image of the 6 nm Fe₃O₄-rGO and 18 nm Fe₃O₄-rGO composites. The inset in the left image shows EDX pattern of the 6nm-Fe₃O₄-rGO sample.

Fig. 4 Magnetization curves of 6 nm Fe₃O₄ NPs and 6 nm Fe₃O₄-rGO composites. The inset shows photographs of the aqueous solution of the 6 nm Fe₃O₄-rGO composites before (left) and after attracted by a magnetic bar (right).

Finally, we carried out a preliminary study on the use of the 6nm-Fe₃O₄-rGO nanocomposites in magnetically-guided removal of aromatic antibiotic contaminants from waste water (Fig. 5). We chose tetracycline (TC) as a model contaminant because TC and related derivatives are consumed in huge amounts (∼10,000 ton) annually worldwide, much of which is discharged into soil and water, becoming a major environmental hazard.³¹ We demonstrated previously that GO exhibits a super-strong ability to adsorb aromatic anticancer drugs.¹⁹ For the 6nm-Fe₃O₄-rGO composites, the basal plane of the GO is almost unoccupied by the magnetic NPs (TEM image in Fig. 3), and therefore facilitates selective adsorption of aromatic TC molecules via π–π stacking and other molecular interactions. The magnetic GO loaded with TC can be removed from the waste water readily by applying a magnetic bar. In our experiment, the aqueous suspension of 6nm-Fe₃O₄-rGO is simply mixed with aqueous solutions of TC of different concentrations and stirred for 24h. After adsorption of TC, the magnetic composites were then separated with the aid of a magnetic bar. Adsorbed amount of TC was found to increase with increasing equilibrium concentration of TC. The maximum amount of TC adsorbed to GO was determined to be 95 mg g⁻¹ (Fig. 5) under current experimental conditions. It was found that the adsorption of TC by 6nm Fe₃O₄-rGO composites follows Freundlich isotherm (Fig. S6†). The adsorption of TC by GO is so strong that sonication for 30 min cannot remove the TC from the GO sheet (data not shown). Further experiments are under way in this lab to investigate the efficient use of the magnetic GO in selective removal of TC and other environmental contaminants with aromatic structures from waste water.

Fig. 5 Adsorption of tetracycline by 6 nm Fe₃O₄-rGO as a function of equilibrium concentration at 25 °C.

In summary, we demonstrated that simple heating of the mixture of PEI and GO in aqueous suspension leads to chemical reduction of GO, and binding of PEI to the GO sheet mainly at its outer boundary, which then guided prefential assembly of the magnetic NPs on the periphery of the GO sheet. This structural feature of the GO-based magnetic composites facilitates efficient loading of anticancer drugs or adsorbing aromatically structured chemical pollutants onto the basal region of the GO sheets. This approach may be generalized to form a wide variety of NP-rGO composites, by simply changing the type of nanoparticles and the ratio of NPs to GO, adaptable to different applications.

Experimental section

Materials

Native graphite flake was purchased from Alfa Aesar. Concentrated sulfuric acid (H₂SO₄), potassium persulfate (K₂S₂O₈), phosphorus pentoxide (P₂O₅), potassium permanganate (KMnO₄), branched polyethylenimine 25 K (PEI), absolute ethanol, hexane, dichloromethane, iron(III) acetylacetonate (Fe(acac)₃), oleic acid, oleylamine, meso-2,3-dimercaptosuccinic acid (DMSA), dimethylsulfoxide (DMSO), 1-ethyl-3-(3-dimethyaminopropyl) carbodiimide (EDC), and N-hydroxysuccinnimide (NHS) were purchased from Sigma-Aldrich. Ultrapure water (18MΩ cm⁻¹) was used in all experiments.

Instrumentation

The Fe₃O₄ NPs and their GO composites were characterized by atomic force microscopy (AFM) and Tecnai G2 F20 S-Twin transmission electron microscopy (TEM) equipped with Energy Dispersive Spectrometry (EDX). Samples for TEM were prepared by placing a drop of water dispersion onto a carbon-coated copper grid and drying at room temperature. Zeta potential was obtained using a ZEN3600 (NanoZS). UV-vis spectra were collected using a Lambda 25 UV-vis-near infrared spectrophotometer. Raman spectra were measured using a confocal microprobe Raman system (HR 800), which is equipped with a holographic notch filter and a CCD detector. The excitation wavelength was 632.8 nm from a He–Ne laser. Thermogravimetric analysis (TGA) was conducted using a TG/DTA 6200 system under dry air with a heating rate of 10 °C min⁻¹. Magnetic measurement was carried out at room temperature with a vibrating sample magnetometer (VSM).

Preparation of GO. Water soluble GO was prepared by oxidizing graphite according to the modified Hummer's method.^32,33

Preparation of Fe₃O₄ NPs. Synthesis of 6 nm Fe₃O₄ nanoparticles. Briefly, Fe(acac)₃ (3 mmol) was injected into a mixture containing 20 mL of octylamine and 10 mL of phenyl ether at 110 °C for 1 h and then, heated to reflux (300 °C) for another 1 h. This solution was cooled to room temperature. The solution was treated with excess ethanol and separated by centrifugation. The black product was dissolved in cyclohexane in the presence of oleic acid (∼0.05 mL) and oleylamine (∼0.05 mL). Centrifugation (4000 rpm, 10 min) was applied to remove the undissolved residue from the solution.¹⁷ Synthesis of 18 nm Fe₃O₄ nanoparticle is similar.

Modification of the Fe₃O₄ NPs by meso-2,3-dimercaptosuccinic acid (DMSA). In this experiment, 20 mg of the as-prepared Fe₃O₄ NPs was dissolved in 2 mL toluene followed by addition of a solution containing 20 mg DMSA and 2 mL DMSO. The resulting solution was stirred at 25 °C for 12 h. This precipitant was then carefully washed with ethyl acetate three times and separated by magnetic separation. Finally, the DMSA-Fe₃O₄ NPs were collected using a permanent magnet bar and transferred into 2 mL water.¹⁸

Synthesis of Fe₃O₄-rGO composites. The mixture of 10 mL aqueous solution of Fe₃O₄ NPs (∼3 × 10⁻⁶ mol mL⁻¹), 30 mg of EDC, 30 mg of NHS, and 15 mL of PEI-rGO (∼0.2 mg mL⁻¹) was stirred at room temperature for 48 h, and the raw product was centrifuged, then washed 3 times using milli Q water.

Removal of tetracycline by Fe₃O₄-rGO composites. Adsorption of an antibiotic, tetracycline (TC), on 6nm-Fe₃O₄-rGO composites was carried out in a 5 mL vial at 25 °C, shaking using a rotary shaker. The vials were wrapped in aluminium foil to prevent possible photodegradation of TC. The initial TC solution was at a concentration ranging from 12.5 mg L⁻¹ to 75 mg L⁻¹, which was determined by UV-vis absorbance at 357 nm, using a calibration curve built with tetracycline solutions of different concentrations. The aqueous solution of the 6nm-Fe₃O₄-rGO composites (1mL, 1.6 mg mL⁻¹) was added to 3mL of TC solution with different concentrations in vials. After shaking for 24 h using a rotary shaker, the clear supernatant TC solution was separated using a magnet.

Then the mass of adsorbed tetracycline per mass unit of the 6nm-Fe₃O₄-rGO is expressed as function of tetracycline concentration in supernatant as

q_e = V(C₀ − C_e)/m

Where m is the mass of 6nm-Fe₃O₄-rGO in dispersion, C₀ the initial concentration of TC, and C_e the concentration of TC in the supernatant at equilibrium state.

The adsorption of TC to the 6nm-Fe₃O₄-rGO can be expressed by Freundlich model equation:

lnq_e = lnK_f + 1/n(lnC_e)

where q_e is the amount of TC adsorbed to the GO at equilibrium state, and C_e is the equilibrium concentration. The Freundlich constants, K_f and n, are determined to be 2.071 (L mg⁻¹) and 0.993, respectively.

Acknowledgements

Financial support of this work by Natural Science Foundation of China (No.20873090, 21073224), National Basic Research Program of China (2010CB933504), and CAS (KJCX2.YW.M12) is gratefully acknowledged. The authors thank Platform for Characterization and Test at SINANO for TEM measurement.

References

1. K. Jasuja, J. Linn, S. Melton and V. Berry, J. Phys. Chem. Lett., 2010, 1, 1853.
2. Y. C. Si and E. T. Samulski, Chem. Mater., 2008, 20, 6792.
3. C. Xu, X. Wang and J. W. Zhu, J. Phys. Chem. C, 2008, 112, 19841.
4. Z. H. Tang, S. L. Shen, J. Zhuang and X. Wang, Angew. Chem. Int. Ed., 2010, 49, 1.
5. B. S. Kong, J. X. Geng and H. T. Jung, Chem. Commun., 2009, 16, 2174.
6. C. X. Guo, H. B. Yang, Z. M. Sheng, Z. S. Lu, Q. L. Song and C. M. Li, Angew. Chem., Int. Ed., 2010, 49, 3014.
7. C. H. Lu, H. H. Yang, C. L. Zhu and G. N. Chen, Angew. Chem., Int. Ed., 2009, 48, 4785.
8. X. Y. Zhang, X. Y. Yang, Y. F. Ma, Y. Huang and Y. S. Chen, J. Nanosci. Nanotechnol., 2010, 10, 2984.
9. H. P. Cong, J. J. He, Y. Lu and S. H. Yu, Small, 2010, 6, 169.
10. X. Y. Yang, X. Y. Zhang, Y. F. Ma, Y. Huang, Y. S. Wang and Y. S. Chen, J. Mater. Chem., 2009, 19, 2710.
11. F. He, J. T. Fan, D. Ma, L. M. Zhang, C. Leung and H. L. Chan, Carbon, 2010, 48, 3139.
12. J. F. Shen, Y. Z. Hu, M. Shi, N. Li, H. W. Ma and M. X. Ye, J. Phys. Chem. C, 2010, 114, 1498.
13. T. Szabó, A. Bakandritsos, V. Tzitzios, E. Devlin, D. Petridis and I. Dékány, J. Phys. Chem. B, 2008, 112, 14461.
14. V. K. Singh, M. K. Patra, M. Manoth, G. S. Gowd, S. R. Vadera and N. Kumar, New Carbon Mater., 2009, 24, 147.
15. L. Q. Jiang and L. Gao, Chem. Mater., 2003, 15, 2848.
16. V. Georgakilas, V. Tzitzios, D. Gournis and D. Petridis, Chem. Mater., 2005, 17, 1613.
17. S. H. Sun, H. Zeng, D. B. Robinson, S. Raoux, P. M. Rice, S. X. Wang and G. X. Li, J. Am. Chem. Soc., 2004, 126, 273.
18. Z. P. Chen, Y. Zhang, K. Xu, R. Z. Xu, J. W. Liu and N. Gu, J. Nanosci. Nanotechnol., 2008, 8, 6260.
19. L. M. Zhang, J. G. Xia, Q. H. Zhao, L. W. Liu and Z. J. Zhang, Small, 2010, 6, 537.
20. D. S. Yu and L. M. Dai, J. Phys. Chem. Lett., 2010, 1, 467.
21. Q. Yang, X. J. Pan, F. Huang and K. C. Li, J. Phys. Chem. C, 2010, 114, 3811.
22. C. Z. Zhu, S. J. Guo, Y. X. Fang and S. J. Dong, ACS Nano, 2010, 4, 2429.
23. W. F. Chen, L. F. Yan and P. R. Bangal, Carbon, 2010, 48, 1146.
24. G. K. Ramesha and S. Sampath, J. Phys. Chem. C, 2009, 113, 7985.
25. H. L. Guo, X. F. Wang, Q. Y. Qian, F. B. Wang and X. H. Xia, ACS Nano, 2009, 3, 2653.
26. S. Stankovich, R. D. Piner, X. Q. Chen, N. Q. Wu, S. T. Nguyen and R. S. Ruoff, J. Mater. Chem., 2006, 16, 155.
27. J. X. Geng and H. T. Jung, J. Phys. Chem. C, 2010, 114, 8227.
28. S. J. Park and R. S. Ruoff, Nat. Nanotechnol., 2010, 4, 217.
29. K. Woo, J. Hong, S. Choi, H. W. Lee, J. P. Ahn, C. S. Kim and S. W. Lee, Chem. Mater., 2004, 16, 2814–2818.
30. N. Z. Bao, L. M. Shen, Y. S. Wang, P. Padhan and A. Gupta, J. Am. Chem. Soc., 2004, 126, 273–279.
31. A. K. Sarmah, M. T. Meyer and A. B. A. Boxall, Chemosphere, 2006, 65, 725.
32. W. S. Hummers, R. E. Offeman and J. Am, J. Am. Chem. Soc., 1958, 80, 1339.
33. N. I. Kovtyukhova, P. J. Ollivier, B. R. Martin, T. E. Mallouk, S. A. Chizhik, E. V. Buzaneva and A. D. Gorchinskiy, Chem. Mater., 1999, 11, 771.

---

Footnotes

† Electronic supplementary information (ESI) available: TEM, UV-vis spectra, Raman spectra, FTIR, and TGA. See DOI: 10.1039/c0nr00776e

‡ These authors contributed equally to this work.

---