Journal of the Korean Physical Society, Vol. 57, No. 6, December 2010, pp. 1545∼1549

Silica-coated Iron-oxide Nanoparticles Synthesized as a T2 Contrast Agent forMagnetic Resonance Imaging by Using the Reverse Micelle Method

Tanveer Ahmad and Ilsu Rhee∗

School of Physics and Energy Sciences, Kyungpook National University, Daegu 702-701

Sungwook Hong

Division of Science Education, Daegu University, Gyeongsan 712-714

Yongmin Chang and Jaejun Lee

Department of Diagnostic Radiology, College of Medicine,Kyungpook National University and Hospital, Daegu 700-721

(Received 12 January 2010, in final form 4 October 2010)

We synthesized iron-oxide (Fe3O4) nanoparticles by using the reverse micelle method and coatedthem with biocompatible silica. The coated nanoparticles were found to be spherical in the TEMimages and showed a uniform size distribution with an average diameter of 10 nm. The T1 and theT2 relaxation times of hydrogen protons in aqueous solutions with various concentrations of silica-coated nanoparticles were determined by using a magnetic resonance (MR) scanner. We found thatthe T2 relaxivity was much larger than the T1 relaxivity for the nanoparticle contrast agent, whichreflected the fact that the T2 relaxation was mainly influenced by outer sphere processes. The T2

relaxivity was found to be 15 times larger than that for the commercial Gd-DTPA-BMA contrastagent. This result demonstrates that silica-coated iron oxide nanoparticles are applicable as a T2

agent in magnetic resonance imaging.

PACS numbers: 81.05.Ni, 76.60.Es, 61.46.Df, 87.61.-cKeywords: Iron-oxide nanoparticle, Silica coating, Relaxivity, Contrast agent, MRIDOI: 10.3938/jkps.57.1545

I. INTRODUCTION

Recently, magnetic nanoparticles have received consid-erable attention because of their potential applicationsin biomedical fields, such as cell labeling [1] and cell sep-aration [2], targeted drug delivery [3,4], and magneticferrofluids hyperthermia [5,6]. The main medical appli-cation of magnetic nanoparticles is in the field of diagnos-tic magnetic resonance imaging, where the superparam-agnetic properties of nanoparticles can provide effectivecontrast in imaging [7-9]. Nanoparticle contrast agentsalter the magnetic relaxation time of hydrogen protonsin tissues [10].

Silica plays a very important and promising role in thedevelopment of coatings for magnetic nanoparticles, inregard to both basic research and technological applica-tions. Silica-coated magnetic nanoparticles are more eas-ily dispersed in liquid media by screening the magneticdipolar attraction among the nanoparticles. The silicacoating also protects the nanoparticles from leaching in

∗E-mail: ilrhee@knu.ac.kr

an acidic environment. Silica-coated magnetic nanopar-ticles are also easier to activate, creating a surface withvarious functional groups. This is due to the existenceof abundant silanol groups on the silica layer. The mostimportant role of the silica coating is to provide a chem-ically inert surface for magnetic nanoparticles in biolog-ical systems [11,12].

In general, two different approaches have beenemployed to synthesize silica coatings on magnetitenanoparticles. The first method relies on the well-knownStober process [13], which is comprised of hydrolysisand polycondensation of tetraethoxysilane under alka-line conditions in ethanol. The other method is basedon microemulsion synthesis, in which micelles or reversemicelles are used as mini-reactors to control the nanopar-ticle size and the silica coating on the magnetic nanopar-ticles [14,15]. Smaller and more uniform particles can besynthesized, with good control over the amount of ironoxide and the resultant magnetic properties, by using themicroemulsion approach [16].

In this paper, we report the synthesis of silica coat-ings on iron oxide (Fe3O4) nanoparticles fabricated byusing the reverse micelle method, and we evaluate these

-1545-

-1546- Journal of the Korean Physical Society, Vol. 57, No. 6, December 2010

Fig. 1. TEM images of silica-coated iron-oxide nanoparti-cles.

coated particles as a potential T2 contrast agent in mag-netic resonance imaging. We studied the T1 and the T2

relaxations of hydrogen protons in water molecules inan aqueous solution of silica-coated iron oxide nanopar-ticles. We found that the T2 relaxivity for an aqueoussolution of silica-coated oxide nanoparticles was 15 timeslarger than that of the commercial Gd-DTPA-BMA con-trast agent. This result demonstrates that silica-coatediron-oxide nanoparticles are suitable as a T2 agent inmagnetic resonance imaging.

II. EXPERIMENTAL

Silica-coated iron-oxide (Fe3O4) nanoparticles weresynthesized by using the reverse micelle method [17].The emulsion was prepared by adding 3.5 g of sodiumdodecylbenzenesulfonate (NaDBS, Aldrich) to 30 ml of axylene (isomers plus ethylbenzene, 98.5+%) solution andmixing well with sonication for 30 min. Under vigorousstirring at 500 rpm, an iron salt solution, composed of0.1 M FeCl2·4H2O (99% Aldrich), 0.2 M Fe(NO3)3 ·9H2O(98%, Aldrich), and 1.8ml water (ACS, reagent, Aldrich),was added to the emulsion solution. For stabilizationof the reverse micelle solution (water-in-oil phase), theemulsion was stirred continuously for 16 hours at roomtemperature. Then, after the solution had been stirredfor its homogeneity at 500 rpm under continuous argonflow for an hour, the micelle solution was slowly heatedto 90 ◦C, and 1 ml of a hydrazine solution (34 wt-%water solution) was injected into the solution. The re-sulting solution was aged for 3 hours and cooled downto 40 ◦C within 90 min. After cooling, 4 ml of TEOSwas injected into the emulsion. While stirring at 500rpm for 6 hours, silica shells were formed on the sur-faces of the iron-oxide (Fe3O4) nanoparticles by hydrol-ysis of TEOS molecules in the water region of reversemicelles. The coated nanoparticles were separated byusing acetone and subsequent centrifugations. The col-lected nanoparticles were redispersed in water for therelaxivity measurements.

The particle size distribution and the structure of

Fig. 2. MR images for T1 measurements. Each figurecorresponds to a specific TI (time of inversion): TI = 50msec (left) and TI = 1600 msec (right). Circular images inthe picture are MR images for aqueous samples of variousconcentrations. The five samples located to the far right arethe ones relevant to this study.

Fig. 3. MR signal intensity as a function of the time ofinversion (TI). The T1 relaxation in the 696 ppm sample (leftfigure) is faster than that in the 70 ppm sample (right figure).

the silica-coated nanoparticles were checked with aTEM (transmission electron microscope, H-7600, HitachiLtd.). For the relaxivity measurements, aqueous solu-tions of various nanoparticle concentrations were pre-pared. The concentration of nanoparticles in the aqueoussolution was measured with an ICP (inductively coupledplasma) spectrophotometer (Thermo Jarrell Ash IRIS-AP). The T1 and the T2 relaxation times of hydrogenprotons in the aqueous solution of the coated nanoparti-cles were measured using an MR scanner (1.5T Scanner,GE Medical System).

III. RESULTS AND DISCUSSION

Figure 1 shows TEM images of the silica-coated iron-oxide (Fe3O4) nanoparticles. The coated nanoparticlesare spherical, with an average diameter of 10 nm and astandard deviation of 2 nm. The diameter of the ironoxide core was approximately 5 nm.

MR images for the aqueous solutions of variousnanoparticle concentrations were obtained. We usedfive different samples with varying concentrations ofnanoparticles: 70, 140, 280, 420, and 696 ppm.

For the T1 measurements, the inversion recovery pulsesequence was used. MR images for 35 different TI’s

Silica-coated Iron-oxide Nanoparticles Synthesized as a T2 Contrast Agent · · · – Tanveer Ahmad et al. -1547-

Fig. 4. MR signal intensity as a function of the time ofinversion (TI). The T1 relaxation in the 696 ppm sample (leftfigure) is faster than that in the 70 ppm sample (right figure).

(times of inversion), ranging from 50 to 1750 msec, wereobtained. Figure 2 shows two sets of MR images for dif-ferent TI’s of 50 msec (left) and 1600 msec (right). Inthis figure, the five samples located at the far right ofeach figure are the samples of silica-coated nanoparticlespertaining to this study. The signal intensities for 35different TI’s could be obtained from these MR images.Applying the data to the intensity function of the MRsignal, I ∼ M0(1− 2e−t/T1), we could obtain the T1 re-laxation time. Figure 3 shows the plot for two differentsamples of nanoparticles at 70 and 696 ppm.

The CPMG (Carr-Purcell-Meiboon-Gill) pulse se-quence with multiple spin echo was used for the T2

measurements. MR images for 22 different TE’s (timesof echo) ranging from 10 to 850 msec were obtained.The signal intensity function for T2 relaxation, I ∼Mo

xye−t/T2 , was used to determine the T2 relaxationtimes. Figure 4 shows the T2 relaxation times for twosamples of nanoparticles, 70 and 696 ppm, respectively.

The relaxivity is a measure of the ability of MRI con-trast agents to increase the relaxation of the surroundingnuclear spins (hydrogen protons), which can then be usedto improve the contrast in MR images. The relaxivity isexpressed in units of s−1 per ppm of nanoparticles. Thecontribution of paramagnetic contrast agents to the re-laxation of the nuclear spins is due to both the inner andthe outer sphere processes. The inner sphere process isfrom the chemical interchange interaction between thebound water of the paramagnetic agents and the sur-rounding free water, which eventually increases the relax-ation (larger effect on T1) of nuclear spins. On the otherhand, the outer sphere process occurs when the param-agnetic agents diffuse through free water. In this process,random fluctuations of paramagnetic agents create an in-homogeneity of the local magnetic field, thus increasingthe relaxation (larger effect on T2) of nuclear spins [18].In the clinically-used gadolinium-based contrast agents,gadolinium ions are formed as chelates. Thus, the boundwater of the chelates can continuously interact with thesurrounding free water and increase the T1 relaxation ofnuclear spins. Most gadolinium chelate agents have aninner sphere effect that is larger than the outer sphereeffect; therefore, they are used as T1 contrast agents.On the other hand, coated ferrite nanoparticle agents are

Fig. 5. Plot of 1/T1 versus particle concentration forsilica-coated iron-oxide nanoparticles. The same plot for Gd-DTPA-BMA is shown here for reference.

completely surrounded by their coating material, and thechemical interchange interaction (inner sphere process)does not occur. However, the ferrite noaoparticles havea much larger magnetic moment than gadolinium ionsand produce larger magnetic field fluctuations (inhomo-geneity). Due to this property of magnetic nanoparticles,they are considered to be ideal T2 contrast agents.

The relaxivities of nuclear spins in an aqueous solutionof magnetic nanoparticles can be expressed as [19]

1Tim

=1Ti

+ RiC, (1)

where i = 1 or 2, and 1/Ti represents the relaxivity ofnuclear spins with no nanoparticle contrast agent. Ri isthe relaxivity of nuclear spins per ppm of nanoparticles,and C represents the concentration of nanoparticles inthe aqueous solution.

Figure 5 is a plot of 1/T1 against particle concen-tration for aqueous solutions of silica-coated iron oxidenanoparticles and the commercial Gd-DTPA-BMA con-trast agent [20]. The slopes of the lines were 0.003 and0.0068 ppm−1sec−1, respectively. The T1 relaxivity forthe coated iron-oxide sample was 2.3 times smaller thanthat for the Gd-DTPA-BMA sample. This result demon-strates that for the nanoparticle contrast agent the T1

relaxation was more influenced by the inner sphere pro-cess than by the outer sphere process.

Figure 6 shows 1/T2 as a function of the particle con-centration for the silica-coated iron-oxide sample andthe commercial Gd-DTPA-BMA contrast agent. Theslopes of the lines for these samples were 0.15 and 0.0097ppm−1sec−1, respectively. The T2 relaxivity for thenanoparticle sample was 15 times larger than that forthe Gd-DTPA-BMA sample, which was expected. Thisresult shows that the iron-oxide nanoparticles are moreeffective as a T2 agent than the gadolinium-based con-

-1548- Journal of the Korean Physical Society, Vol. 57, No. 6, December 2010

Fig. 6. Plot of 1/T2 versus particle concentration forsilica-coated iron-oxide nanoparticles. The same plot for Gd-DTPA-BMA is shown here for reference. We can see thatthe T2 relaxivity for the coated iron-oxide sample was 15-fold larger than that for the Gd-DTPA-BMA sample. Seethe text for details.

Fig. 7. T2-weighted MR images of the abdomen of a mouse(a) before and (b) 20 minutes after the injection of 0.21 mgof iron-oxide nanoparticle agents into the mouse vein. Theratio of signal intensities at the marked position (©) before(a) and after (b) injection was measured as Ib/Ia = 0.53. Inother words, the presence of the particles resulted in a 47%decrease in the signal intensity.

trast agents.Compared to the previously reported relaxivity of a

dextran-coated iron-oxide sample [21, 22], we obtaineda 208% increase in T2 relaxivity with the silica-coatediron-oxide sample. In that paper, we also reported theeffect of the dextran coating on the T2 relaxivity. TheT2 relaxivity for the dextran-coated nanoparticles was 52times smaller than that of the bare iron-oxide nanopar-ticle, which was 3.75 ppm−1sec−1. The decrease in theT2 relaxivity for silica-coated nanoparticle is much lessthan this, but is still 25 times smaller than that for bareiron-oxide nanoparticles. The thickness of the dextrancoating was 128 nm, which is about thirteen times largerthan that for silica coating, 10 nm. This thick coating ofdextran might screen the dipolar interaction of nanopar-ticles with free water.

The T2 relaxation enhancement effect of our coatedsample was also observed in an animal model. We ob-tained abdominal MR images of a mouse (150 g weight)

both with and without the injection of an aqueous solu-tion of silica-coated nanoparticles. 0.3 ml of the aqueoussolution was given to the mouse, which corresponds to0.21 mg of iron oxide. The MR images were obtainedusing an MRI scanner (1.5T MR Scanner, GE MedicalSystem). A T2 image with no contrast agent was takenfor reference. Then, after the injection of the contrastagent, the T2 images were taken every 5 minutes. Fig-ure 7(a) shows an abdominal MR image prior to injec-tion of the agent. Most parts of this image correspondto the liver. Figure 7(b) shows an image taken 20 min-utes after the injection of the agent. In this figure, withagent in the system, the liver portion of the image wasrendered visibly darker. This can be attributed to thefaster T2 relaxation of nuclear spins in the liver due tothe uptake of iron-oxide nanoparticles by Kupffer’s cellsin the liver. The signal intensity after the injection of theagent at the marked position (© in Fig. 7(b)) became47% smaller than the signal intensity at the same posi-tion before the injection (© in Fig. 7(a)). The results ofanimal experimentation show that our contrast agent ofsilica-coated nanoparticles can be used as a T2 agent inMRI.

IV. CONCLUSION

We synthesized iron-oxide (Fe3O4) nanoparticles byusing the reverse micelle method and coated them withbiocompatible silica. The coated nanoparticles werefound to be spherical in the TEM images and showed auniform size distribution with an average diameter of 10nm. The T1 and the T2 relaxation times of the hydrogenprotons in aqueous solutions of varying concentrationswere determined by using a magnetic resonance scanner.We found that the T2 relaxivity was much larger than theT1 relaxivity for the nanoparticle contrast agent, whichreflected the fact that the T2 relaxation was mainly influ-enced by outer sphere processes. The T2 relaxivity wasfound to be 15 times larger than that for the commercialGd-DTPA-BMA contrast agent. This result shows thatsilica-coated iron-oxide nanoparticles are applicable as apotential T2 agent in magnetic resonance imaging.

ACKNOWLEDGMENTS

This work was supported by the National ResearchFoundation of Korea (2010-0021315). This work wasalso supported by Basic Science Research Programthrough the National Research Foundation of Korea(NRF) grant funded by the Korea government (MEST)(2009-0072413) and Nuclear R&D Program (grant code:20090081817) of NRF funded by MEST.

Silica-coated Iron-oxide Nanoparticles Synthesized as a T2 Contrast Agent · · · – Tanveer Ahmad et al. -1549-

REFERENCES

[1] E. Farrell, P. Wielopolski, P. Pavljasevic, N. Kops, H.Weinans, M. R. Bernsen and G. J. V. M. van Osch, Os-teoarthr. Cartilage 17, 961 (2009).

[2] J. Ahn, W.-J. Chung, I. Pinnau and M. D. Guiver, J.Membr. Sci. 314, 123 (2008).

[3] D. F. Emerich and C. G. Thanos, Biomol. Eng. 23, 171(2006).

[4] A. S. Kearney, Adv. Drug Delivery Rev. 19, 225 (1996).[5] R. Hergt and S. Dutz, J. Magn. Magn. Mater. 311, 187

(2007).[6] R. Hergt, R. Hiergeist, I. Hilger, W. A. Kaiser, Y. Lapat-

nikov, S. Margel and U. Richter, J. Magn. Magn. Mater.270, 345 (2004).

[7] Y. Anzai, K. E. Blackwell, S. L. Hirschowitz, J. W.Rogers, Y. Sato, W. T. C. Yuh, V. M. Runge, M. R.Morris, S. J. McLachlan and R. B. Lufkin, Radiology192, 709 (1994).

[8] R. Weissleder, A. Bogdanov, E. A. Neuwelt and M. Pa-piosv, Adv. Drug Delivery Rev. 16, 321 (1995).

[9] R. W. Briggs, Z. Wu, C. R. J. Mladinich, C. Stoupis,J. Gauger, T. Liebig, P. R. Ros, J. R. Ballinger and P.Kubilis, J. Magn. Reson. Imaging 15, 559 (1997).

[10] S. Hong, I. Rhee and Y. Chang, J. Korean Phys. Soc.

51, 1453 (2007).[11] Y.-H. Deng, C.-C. Wang, J.-H. Hu, W.-L. Yang and S.-

K. Fu, Colloid. Surface. A 262, 87 (2005).[12] F. Mazaleyrat, M. Ammar, M. LoBue, J.-P. Bonnet, P.

Audebert, G.-Y. Wang, Y. Champion, M. Hytch and E.Snoeck, J. Alloys Compd. 483, 473 (2009).

[13] W. Stober, A. Fink and E. Bohn, J. Colloid InterfaceSci. 26, 62 (1968).

[14] S. Santra, R. Tapee, N. Theodoropoulou, J. Dobson, A.Hebard and W. Tan, Langmuir 17, 2900 (2001).

[15] P. Tartaj and C. J. Serna, J. Am. Chem. Soc. 125, 15754(2003).

[16] G. Mobe, K. Kon-No, K. Kyori and A. Kitharan, J. Col-loid Interface Sci. 93, 293 (1983).

[17] J. Lee, Y. Lee, J.-K. Youn, H.-B. Na, T. Yu, H. Kim,S.-M. Lee, Y.-M. Koo, J.-H, Kwak, H.-G. Park, H.-N.Chang, M. Hwang, J.-G. Park, J. Kim and T. Hyeon,Small 4, 143 (2008).

[18] R. B. Lauffer, Chem. Rev. 87, 901 (1987).[19] Y. Okuhata, Adv. Drug Delivery Rev. 37, 121 (1999).[20] S. Hong, Ph. D. Dissertation, Kyungpook National Uni-

versity, 2007.[21] S. Hong and I. Rhee, J. Korean Phys. Soc. 51, 1453

(2007).[22] S. Hong and I. Rhee, Int. JMRI 1, 15 (2007).