Ultrafine graphene oxide–CoFe₂O₄ nanoparticle composite as T₁ and T₂ contrast agent for magnetic resonance imaging

Abstract

Graphene oxide–CoFe₂O₄ nanoparticle composites were synthesized using a two step synthesis method in which graphene oxide was initially synthesized followed by precipitation of CoFe₂O₄ nanoparticles in a reaction mixture containing graphene oxide. Samples were extracted from the reaction mixture at different times at 80 °C. All the extracted samples contained CoFe₂O₄ nanoparticles formed over the graphene oxide. It was observed that the increase in the reflux time significantly increased the saturation magnetization value for the superparamagnetic nanoparticles in the composite. It was also noticed that the size of the nanoparticles increased with increase in the reflux time. Transverse relaxivity of the water protons increased monotonically with increase in the reflux time. Whereas, the longitudinal relaxivity value initially increased and then decreased with the reflux time. Graphene oxide–CoFe₂O₄ nanoparticle composites also exhibit biocompatibility towards the MCF-7 cell line.

---

1. Introduction

The necessary role of image contrast enhancing materials (contrast agents) in bio-imaging techniques such as magnetic resonance imaging (MRI), X-ray computed tomography (CT), positron emission tomography (PET) and fluorescence imaging (FI) has motivated the development of several novel contrast agents.^1,2 These contrast agents are expected to not only enhance the detection sensitivity but also be active simultaneously in several different imaging modalities.^3,4 Both these factors are necessary for desirably reducing the dosage of the contrast agents into the body subjected to bio-imaging. With respect to the MRI technique, researchers have widely explored ferrite based nanoparticles and gadolinium based complexes for the enhancement of the proton relaxivity which in turn is directly related to the improvement in the MRI image contrast and high detection sensitivity.^5,6

There are essentially two types of MRI contrast agents classified on the basis of their effects on the relaxation behavior of the protons in the nuclear magnetic resonance experiment: reduction of longitudinal (T₁) relaxation time which causes bright contrast and, reduction of transverse (T₂) relaxation time which causes dark contrast in the MR image.^5,7 Ferrite based nanoparticles are commonly used for the T₂ weighted MR images.⁶ Resovist and Feridex are two ferrite based MRI contrast agents that have been approved for clinical usage.⁸ The use of T₂ weighted contrast agents in MR imaging is however limited by the blurring effect caused by the high magnetic moment of the nanoparticles which can mislead the correct interpretation of the MR image.⁵ Because of this, use of T₁ weighted agents is preferred in clinical practices. Commercially used T₁ weighted agents are gadolinium based complexes⁹ (Dotarem, Magnetol, Gadavist, Magnevist, Omniscan, DOTA). These agents also possesses undesirable attributes such as short blood circulation time and toxic nature of free Gd³⁺ ions that undesirably resides in the body for long time.^10,11 One way to address this issue is to design biocompatible ferrite nanoparticles based contrast agents which can also enhance the r₁ relaxivity. This can be achieved by manipulating the sizes of the nanoparticles. Thickness of the magnetic dead layer on the surface increases with decrease in the size of the magnetic nanoparticles.¹² Ultrafine ferrite nanoparticles having significant fraction of surface dead layer can affect the T₁ relaxivity of the water protons.¹³ Fe³⁺ has five unpaired electrons in the case of ultrafine ferrite nanoparticles most of the Fe³⁺ will be on the surface which will interact with water protons causing higher longitudinal relaxivity (r₁). There are reports in the literature which show the potential of ultrafine ferrite nanoparticles as T₁ weighted agents.^13,14

It has been demonstrated that graphene oxide can be used as a carrier for both water soluble and insoluble drugs that can be delivered at the infected site by pH and temperature response.¹⁵ Reports on the synthesis and biomedical application of graphene oxide–nanoparticle composites are available in the literature.^16–18 Reports demonstrating the effect of nanoparticles–graphene oxide composites on relaxivity of water protons are also available. It has been demonstrated that the arrangement of nanoparticles on the graphene oxide framework plays a crucial role in deciding the interaction between water protons and magnetic nanoparticles.¹⁸ The nanoparticles–graphene oxide composites can therefore be potential materials for both diagnosis and treatment of cancers.

This study investigates GO–CoFe₂O₄ ultrafine nanoparticles composite as a potential image contrast enhancing material for both T₁ and T₂ weighted images in MRI. Reports on the investigation of the CoFe₂O₄ nanoparticles as MRI contrast agent material is already available in the literature.¹⁹ These reports have however only demonstrated the suitability of the CoFe₂O₄ nanoparticles as T₂ contrast agent.

2. Experimental

In this work, GO–CoFe₂O₄ nanoparticle composite was synthesized using a two step approach. In the first step graphene oxide (GO) was synthesized and in the second step CoFe₂O₄ nanoparticles were precipitated in the presence of graphene oxide to produce GO–CoFe₂O₄ nanoparticle composites.

2.1 Synthesis of graphene oxide

Graphene oxide was synthesized by the Hummer's method.²⁰ Following procedure was adopted. 50 mL of concentrated H₂SO₄ was poured into a beaker containing 2 g of graphite powder and 1 g of NaNO₃. This reaction mixture was then cooled below 5 °C under constant stirring. At this temperature, 6 g of KMnO₄ was slowly added into the reaction mixture while maintaining its temperature always below 15 °C. After the addition of KMnO₄, temperature of the reaction mixture was increased to 35 °C and was kept at this temperature for 30 min. After 30 min the reaction mixture was diluted with 80 mL of distilled water. Addition of water raised the temperature of the reaction mixture to 80 °C due to exothermic reaction. The reaction mixture was then maintained at this temperature for another 20 min by external heating. After 20 min the reaction mixture was cooled to the room temperature. At room temperature, a mixture of 100 mL of water and 3 mL of 30% H₂O₂ was slowly added into the reaction mixture to sediment the as-synthesized graphene oxide. After sedimentation, the supernatant was discarded and the isolated graphene oxide was washed with 100 mL of water and 100 mL 30% HCl. In this article, pure graphene oxide synthesized in the first step is referred as ‘GO’ in the figures.

2.2 Synthesis of GO–CoFe₂O₄ nanoparticle composite

To synthesize the GO–CoFe₂O₄ nanoparticle composite, 200 mg of initially synthesized graphene oxide was dispersed in 200 mL of distilled water by sonication for 30 min. 0.0457 g of FeCl₃ and 0.0251 g of CoCl₂·6H₂O were then added into this dispersion under continuous stirring. pH of the reaction mixture was adjusted to 11 by drop wise addition of methyl amine. This reaction mixture was then heated to 80 °C in an inert atmosphere. At 80 °C, samples were extracted from the reaction mixture after 0, 5, 10, 20, 40 and 60 min of reflux using a syringe. The extracted reaction mixtures were then cooled to the room temperature and centrifuged to isolate the synthesis product. The synthesis reaction was performed under inert atmosphere.

3. Characterization techniques

X-ray diffraction (XRD) profiles were obtained from the as-synthesized samples using the X-Pert PAN Analytical machine employing Cu K-alpha radiation source. A 300 keV field emission FEI Tecnai F-30 transmission electron microscope (TEM) was used for obtaining TEM bright field images and selected area electron diffraction (SAD) patterns from as-synthesised samples. Samples for the TEM based analysis were prepared by drop-drying a highly dilute dispersion of the as-synthesised sample onto an electron transparent carbon coated Cu grid. Magnetic measurement data from the as-synthesized samples was obtained by using the Lakeshore vibrating sample magnetometer (VSM). Mass of graphene oxide in the GO–CoFe₂O₄ composite nanoparticles was determined by the thermal gravimetric analysis (TGA) measurement conducted using the TGA NETZSCN STA 403 PC machine. The concentration of iron in dispersions used in the MRI experiment was calculated by atomic absorption spectroscopy (AAS) technique conducted using the Thermo Electron Corporation M-series machine. X-ray photoelectron spectroscopy (XPS) profiles were obtained from the as-synthesized samples using an AXIS Ultra DLD (KRATOS ANALYTICAL) instrument. Raman spectra from as-synthesized samples were obtained using a microscope setup (HORIBA JOBIN YVON, Lab RAM HR) consisting of a diode-pumped solid-state laser operating at 532 nm with a charge coupled detector.

T₁ and T₂ relaxivity of water protons in the presence of as-synthesized composites was measured using a Siemens Magneto Avanto 1.5 T MRI scanner. An in vitro phantom containing GO–CoFe₂O₄ composite nanoparticles at 0.1, 0.2, 0.3 and 0.4 mM of iron was prepared to quantify the relaxivity values. All curve fitting routines used to determine the relaxation rate maps were performed using Matlab, The MathworksInc, MA.

3.1 R₁ mapping

A spin echo pulse sequence with the following parameters were used to obtain a total of 10 images (one image for each TR value): slice thickness – 5 mm, matrix size – 256×256, number of slices – 1, echo time (TE) – 8.7 ms, repetition time (TR) values – 100, 200, 300, 400, 600, 800, 1000, 1250, 1500 and 2000 ms were used. R₁ map was generated using the saturation recovery equation²¹ given below:

S = S₀ (1 − e^(TRR₁)) + c (1)

where S₀ is magnetization at equilibrium and c is the compensation term for noise in measurement of data.

3.2 R₂ mapping

Fast Spin Echo (FSE) sequence with an echo train length of 16 with TR value of 3000 ms, matrix size of 512 × 512, slice thickness of 5 mm, and number of slices as 1 was used. The TE was varied from 22 ms to 352 ms with a difference of 22 ms resulting in 16 images with varied T₂ weighting. Using these images, the R₂ map was generated based on equation:²²

S = S₀ (1 − e^(TRR₂)) + c (2)

where, S₀ is magnetization at equilibrium and c is compensation noise in measurement relaxivity is determined by the relaxation rate R as a function of concentration C given by:²³

3.3 Cytotoxicity analysis

For the cytotoxicity analysis MCF-7 (human breast cancer cells) cell line was cultured in Dulbecco's Modified Eagle's Medium (DMEM) with 10% fetal bovine serum and incubated in a humidified atmosphere at 37 °C with 5% CO₂. MTT assay was performed to evaluate the cytotoxicity of the six extracted GO–CoFe₂O₄ nanoparticle composites and graphene oxide. Cells were seeded at optimum density into each well of 96-well plate and exposed to varied concentrations of GO–CoFe₂O₄ composite nanoparticles and GO. The cells were then incubated with the GO–CoFe₂O₄ composite nanoparticles and GO for 24 h. 5 mg mL⁻¹ MTT dye was then added into each well followed by further incubation for 3–4 h. The MTT formazan crystals formed are metabolically reduced by the mitochondria in viable cells to a colored formazan product, the intensity of which was measured spectrophotometrically in a plate reader at 570 nm.

4. Results and discussion

SEM-EDS profile obtained from all the extracted samples reveled peaks corresponding to Fe and Co. One representative EDS profile obtained from sample extracted after 0 min of reflux (at 80 °C) is shown in Fig. 1. XRD profiles obtained from the as-synthesized graphene oxide and from GO–CoFe₂O₄ nanoparticle composites extracted from the reaction mixture at different times are shown in Fig. 2. It can be seen that the XRD profile obtained from as-synthesized graphene oxide reveals a peak at the 2θ value of 10.48° corresponding to the typical graphene oxide interlayer spacing value of 8.45 Å.¹⁸ Intensity of this peak is negligible in the XRD profiles that were obtained from the extracted samples indicating complete exfoliation of the graphene oxide in the second step of the synthesis process. All the XRD profiles obtained from the extracted samples revealed the (311) peak of the CoFe₂O₄ phase (indicated by arrow in Fig. 2). Representative TEM bright field micrograph obtained from the GO–CoFe₂O₄ nanoparticle composites extracted from the reaction mixture at different times is shown in Fig. 3. It can be observed in Fig. 3 that nanoparticles have formed over the graphene oxide sheet for all the extracted samples. Representative SAD patterns obtained from the GO–CoFe₂O₄ composites extracted after different times are also shown in Fig. 3. The SAD patterns reveal diffraction signals corresponding only to the CoFe₂O₄ phase indicating that the synthesis process in the second step successfully produced CoFe₂O₄ nanoparticles.

Fig. 1 SEM EDS profile obtained from the 0 minute refluxed GO–CoFe₂O₄ nanoparticle composite.

Fig. 2 XRD profiles obtained from pure graphene oxide and GO–CoFe₂O₄ nanoparticle composites extracted after different reflux times.

Fig. 3 TEM bright field image and SAD patterns obtained from GO–CoFe₂O₄ nanoparticle composites refluxed for different times.

Raman spectrum obtained from the graphene oxide produced in the first step of the synthesis process and the GO–CoFe₂O₄ nanoparticle composites extracted at different times at 80 °C are shown in Fig. 4. All the spectrums in Fig. 4 reveal two prominent peaks at ∼1355 and ∼1605 cm⁻¹ corresponding respectively to the D and G bands of graphene oxide.²⁴ Presence of the peaks corresponding to graphene oxide in the Raman spectrums confirmed that the absence of peak corresponding to graphene oxide in the XRD profiles is due to the complete exfoliation of the graphene oxide during the second stage of the synthesis process. The ratios of the peak intensities corresponding to the D and G band (I_D/I_G) for all the samples are listed in Table 1. Negligible variation in the (I_D/I_G) ratio between all the GO–CoFe₂O₄ nanoparticle composites indicated that the graphene oxide in all the six composites has similar defect structure.

Fig. 4 Raman spectrums obtained from pure GO and GO–CoFe₂O₄ nanoparticle composites.

Table 1 List of I_D/I_G values for GO–CoFe₂O₄ nanoparticle composites refluxed for different times

Reflux time (min)	ID/IG
0	1.04
5	1.07
10	1.06
20	1.08
40	1.05
60	1.04

---

Data obtained from the XPS based analysis of as-synthesized graphene oxide and GO–CoFe₂O₄ nanoparticle composites are presented in Fig. 5(a–c). The Co 2p spectra shown in Fig. 5(a) reveal peaks at ∼780.1 and ∼796.3 eV indicating the presence of Co²⁺.²⁵ The Fe 2p spectra shown in Fig. 5(b) reveal peaks at ∼710 and ∼723.3 eV with the separation of 13.3 eV indicating the presence of Fe³⁺. Fig. 5(c) reveal peaks at 284.6 and 286.9 eV corresponding to the binding energies of sp² C–C and C–O bonds (C 1s spectrum) in the graphene oxide of the composites.²⁶ A variation in the relative intensity of C–O and C–C peaks showing a decrease in the intensity of the C–O peak with reflux time can be observed in Fig. 5(c). This indicates reduction in graphene during the precipitation of CoFe₂O₄ nanoparticles.

Fig. 5 XPS spectrums obtained from Go–CoFe₂O₄ nanoparticle composites showing (a) Co 2p, (b) Fe 2p and (c) C 1s peaks spectrums.

Magnetic hysteresis curves and saturation magnetization values, obtained from the GO–CoFe₂O₄ nanoparticle composites extracted at different times are shown in Fig. 6(a) and (b) respectively. Magnetic hysteresis curves were obtained at the room temperature using applied magnetic field in the range of 0–2 tesla. The magnetic hysteresis curves for all the extracted GO–CoFe₂O₄ nanoparticle composites revealed negligible magnetic coercivity and no magnetic saturation till 2 T applied field indicating superparamagnetic nature for the CoFe₂O₄ nanoparticles in all the composites. Interestingly, it can be seen that the saturation magnetization values increased significantly with increase in the reflux time (Fig. 6(b)). Superparamagnetic nature of the nanoparticles was confirmed from the occurrence of a broad peak in the Zero Field Cooled (ZFC) curve obtained using 100 G applied field. The ZFC curves shown in Fig. 6(c) illustrate blocking of CoFe₂O₄ nanoparticle magnetic moment in a certain temperature range which is an indication of the superparamagnetic nature of the CoFe₂O₄ nanoparticles in all the samples extracted at different times form the reaction mixture. Additionally, the shift in blocking temperature (denoted by the peak of the ZFC curve) towards higher temperature values with increase in the reflux time as seen in Fig. 6(c) essentially reveals that the average sizes of the nanoparticles increases with increase in the reflux time. Increase in the value of saturation magnetization with increase in the reflux time can be because of the increase in the nanoparticle size and a corresponding decrease in the volume fraction of the surface dead layer over the nanoparticles. The size dependence of magnetization of the nanoparticles (m_s) is illustrated by the following relation:

m_s = M_s[(r − d)/r]³ (4)

where, M_s is the saturation magnetization of the bulk; r is the radius of the magnetic nanoparticle; d is the thickness of the disordered surface spin layer.

Fig. 6 (a) Magnetic hysteresis curves for GO–CoFe₂O₄ nanoparticle composites refluxed for different times, (b) saturation magnetization vs. reflux time, (c) zero-field-cooled (ZFC) curves measured at 100 G for GO–CoFe₂O₄ nanoparticle composites, (d) experimental (black) and calculated (red) magnetic hysteresis loops for 60 minute refluxed sample and (e) size distributions calculated by Langevin fit.

Since the CoFe₂O₄ nanoparticles in all the extracted samples were superparamagnetic, the M–H curve (Fig. 6(a)) was used to calculate size distributions using the Langevin fit.²⁷ The Langevin function given below describes the dependence of magnetization M on the magnetic field.

where, M_s is the saturation magnetization, μ₀ is the magnetic permeability of vacuum, μ is the magnetic dipole moment of a nanoparticle, H is the magnetic field, and k_BT is the thermal energy. For superparamagnetic nanoparticles, the magnetic moments were assumed to follow the lognormal distribution under this condition the M and H are related by²⁸

The above equation was used to fit the experimentally obtained data to derive the value of the magnetic moment of the nanoparticles. P(μ) is the probability density function of magnetic moment. The fitted and measured data for 60 minute refluxed sample is provided in Fig. 6(c).

The calculated magnetic moment value was then used to obtain the diameter d of the superparamagnetic nanoparticles by the relation.

where m_s is the material-dependent saturation magnetization per unit volume. The probability density function is defined by the following equation:²⁸where, μ* and d* are respectively the dipole moment and magnetic diameter at the maximum of the distribution, and the width of the distribution is described by σ_μ = 3σ_d.

The above formulations were used to obtain the plot of normalized distribution density versus size of nanoparticles for nanoparticles in the GO–CoFe₂O₄ composites extracted after different times of reflux. It was assumed that the nanoparticle contained single magnetic domain. Size distributions calculated for extracted samples are shown in Fig. 6(e). It can be observed in Fig. 6(e) that the increase in the reflux time increased the average size of the nanoparticles. Increase in the reflux time also increases the distribution of nanoparticle sizes. A shift of the peak to the higher blocking temperature values with increase in the reflux time (Fig. 6(c)) also indicates an increase in the size of the CoFe₂O₄ nanoparticles with increasing reflux.

The transverse and longitudinal relaxivity values obtained from the dispersion of extracted samples are tabulated in Table 2. 1/T₂ and 1/T₁ vs. concentration of iron are shown in Fig. 7(a) and (c) respectively. It can be observed that transverse relaxivity (r₂) values increased with increase in the reflux time (Fig. 7(b)). This increase is primarily because of the increase in magnetization value and nanoparticle size. The r₂ shows a linear dependence on reflux time as shown in Fig. 7(b). Whereas, the longitudinal relaxivity (r₁) value initially increases and then decreases with the reflux time as shown in Fig. 7(d). The longitudinal relaxivity is due to the paramagnetic dead layer with free spins on the nanoparticle surface. These surface spins which are not coupled interact with water and exchange energy to facilitate faster relaxation. It should be noted that the highest r₁ value 4.73 mM⁻¹ s⁻¹ is obtained at the reflux time of 5 min. This value of the r₁ is comparable to the relaxivity value obtained from the commercially used Gd–DOTA.²⁹ The decrease in the relaxivity value with increase in the reflux time can be due to the increase in the nanoparticle size and a corresponding decrease in the volume fraction of the paramagnetic surface layer.

Table 2 List of r₁ and r₂ for the GO–CoFe₂O₄ nanoparticle composite refluxed for different times

Reflux time (min)	r1 (mM^−1 s^−1)	r2 (mM^−1 s^−1)
0	3.21	4.1
5	4.72	23.8
10	1.62	59.6
20	0.75	84.7
40		137.1
60		188.2

---

Fig. 7 (a) 1/T₂ vs. concentration of iron plot, (b) r₂ vs. reflux time, (c) 1/T₁ vs. concentration of iron plot and (d) r₁ vs. reflux time.

To investigate the cytotoxicity of the GO–CoFe₂O₄ nanoparticle composites and graphene oxide, MTT assay was performed using MCF-7 (human breast cancer cells). Results from the cytotoxicity experiment are shown in Fig. 8. The GO–CoFe₂O₄ nanoparticle composites with different concentrations (100, 50, 25, 12.5 and 6.25 μg mL⁻¹) showed biocompatibility towards the MCF-7 cell line.

Fig. 8 MTT assay of GO–CoFe₂O₄ nanoparticle composites refluxed for different times.

5. Conclusions

Graphene oxide–CoFe₂O₄ nanoparticle composites were also synthesized using the two step synthesis method in which graphene oxide was initially synthesized by the Hummer's method followed by precipitation of CoFe₂O₄ nanoparticles in a reaction mixture that contained graphene oxide. In this study, the effect of solution reflux time on the size and magnetic property evolution of nanoparticle and the effect of these attributes on the proton relaxivity values obtained in the dispersion of these composites were investigated. Samples were extracted from the reaction mixture at different times (0, 5, 10, 20, 40 and 60 minute) at 80 °C. Extensive characterization experiments revealed that all the extracted samples contained CoFe₂O₄ nanoparticles formed over the graphene oxide (graphene oxide–CoFe₂O₄ nanoparticle composite). Increase in the reflux time significantly increased the saturation magnetization value for the superparamagnetic nanoparticle–GO composite. It was also expectedly noticed that the increase in the reflux time increased the size of the nanoparticles. MRI measurements revealed that both the transverse and longitudinal relaxivity values depended on the reflux time. The transverse relaxivity of the water protons increased monotonically with increase in the reflux time. Whereas, the longitudinal relaxivity value initially increased and then decreased with the reflux time. Cytotoxicity analysis revealed that the graphene oxide–CoFe₂O₄ nanoparticle composites were biocompatible towards the MCF-7 cell line.

Acknowledgements

Authors acknowledge the electron microscopy facilities available at the Advanced Centre for Microscopy and Microanalysis (AFMM) IISc Bangalore. The MCF-7 cell line were provided by Professor Paturu Kondaiah. Cell toxicity analysis was done using the facilities available in Professor Paturu Kondaiah's laboratory in IISc, Bangalore. C. Srivastava acknowledges the research grant received from Department of Science and Technology DST-Nano Mission and SERB, Govt. of India.

Notes and references

1. K. Glunde, A. P. Pathak and Z. M. Bhujwalla, Trends Mol. Med., 2007, 13, 287–297.
2. L. Fass, Mol. Oncol., 2008, 2, 115–152.
3. A. R. Kherlopian, T. Song, Q. Duan, M. A. Neimark, M. J. Po, J. K. Gohagan and A. F. Laine, BMC Syst. Biol., 2008, 2, 74–92.
4. P. Sharma, A. Singh, S. C. Brown, N. Bengtsson, G. A. Walter, S. R. Grobmyer, N. Iwakuma, S. Santra, E. W. Scott and B. M. Moudgil, Methods Mol. Biol., 2010, 624, 67–81.
5. A. S. Merbach, L. Helm and E. Toth, The Chemistry of Contrast Agents in Medical Magnetic Resonance Imaging, John Wiley & Sons, 2013.
6. H. B. Na, I. C. Song and T. Hyeon, Adv. Mater., 2009, 21, 2133–2148.
7. S. Laurent, D. Forge, M. Port, A. Roch, C. Robic, L. V. Elst and R. N. Muller, Chem. Rev., 2008, 108, 2064–2110.
8. W. Cai and J. Wan, J. Colloid Interface Sci., 2007, 305, 366–370.
9. S. Laurent, L. V. Elst and R. N. Muller, Contrast Media Mol. Imaging, 2006, 1, 128–137.
10. H. Ersoy and F. J. Rybicki, J. Magn. Reson. Imaging, 2007, 26, 1190–1197.
11. Z. Zhou and Z. R. Lu, Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol., 2013, 5, 1–18.
12. B. Issa, I. M. Obaidat, B. A. Albiss and Y. Haik, Int. J. Mol. Sci., 2013, 14, 21266–21305.
13. U. I. Tromsdorf, O. T. Bruns, S. C. Salmen, U. Beisiegel and H. Weller, Nano Lett., 2009, 9, 4434–4440.
14. Z. Li, S. X. Wang, Q. Sun, H. L. Zhao, H. Lei, M. B. Lan, Z. X. Cheng, X. L. Wang, S. X. Dou and G. Q. Lu, Adv. Healthcare Mater., 2013, 2, 958–964.
15. C. Chung, Y. K. Kim, D. Shin, S. R. Ryoo, B. H. Hong and D. H. Min, Acc. Chem. Res., 2013, 46, 2211–2224.
16. M. Balcioglu, M. Rana and M. V. Yigit, J. Mater. Chem. B, 2013, 1, 6187–6193.
17. X. Fan, G. Jiao, W. Zhao, P. Jin and X. Li, Nanoscale, 2013, 5, 1143–1152.
18. N. Venkatesha, P. Poojar, Y. Qurishi, S. Geethanath and C. Srivastava, J. Appl. Phys., 2015, 117, 154702–154710.
19. N. Venkatesha, C. Srivastava and V. Hegde, IET Nanobiotechnol., 2013, 8, 184–189.
20. D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii, Z. Sun, A. Slesarev, L. B. Alemany, W. Lu and J. M. Tour, ACS Nano, 2010, 4, 4806–4814.
21. X. Li, E. B. Welch, A. B. Chakravarthy, L. Xu, L. R. Arlinghaus, J. Farley, I. A. Mayer, M. C. Kelley, I. M. Meszoely, J. Means-Powell, V. G. Abramson, A. M. Grau, J. C. Gore and T. E. Yankeelov, Magn. Reson. Med., 2012, 68, 261–271.
22. A. A. O. Carneiro, G. R. Vilela, D. B. Araujo and O. Baffa, Braz. J. Phys., 2006, 36, 9–15.
23. S. H. Koenig and K. E. Kellar, Magn. Reson. Med., 1995, 34, 227–233.
24. K. N. Kudin, B. Ozbas, H. C. Schniepp, R. K. Prudhomme, I. A. Aksay and R. Car, Nano Lett., 2008, 8, 36–41.
25. K. Maaz, A. Mumtaz, S. K. Hasanain and A. Ceylan, J. Magn. Magn. Mater., 2007, 308, 289–295.
26. D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. Soc. Rev., 2009, 39, 228–240.
27. A. Kakay, M. W. Gutowski, L. Takacs, V. Franco and L. K. Varga, J. Phys. A: Math. Gen., 2004, 37, 6027–6041.
28. S. P. Yu, Y. H. Liu, A. C. Sun and J. H. Hsu, J. Appl. Phys., 2009, 106, 103905–103910.
29. J. Y. Park, M. J. Baek, E. S. Choi, S. Woo, J. H. Kim, T. J. Kim, J. C. Jung, K. S. Chae, Y. Chang and G. H. Lee, ACS Nano, 2009, 3, 3663–3669.

---