Journal of the Korean Physical Society, Vol. 52, No. 5, May 2008, pp.

1332$1335

Size Controlled Magnetite Nanoparticles and Their Drug Loading Ability

C. V.

Thach,

N. H.

Hai

and N.

Chau

Center for Materials Science, Hanoi University of Science, 334 Nguyen Trai, Hanoi, Vietnam

(Received 15 July 2007) Magnetite nanoparticles (MNPs) were synthesized by using coprecipitation method with a reaction between the FeCl2 /FeCl3 solution and the ammonia water. The size of the MNPs could be controlled from 10.0 to 14.6 nm by changing the concentration of the solutes. The particles were superparamagnetic at room temperature. The saturation magnetization of the MNPs increased with increasing concentration of reactants. The magnetite nanoparticles were coated with a single layer of oleic acid (OA) to have a hydrophobic surface or with a double layer of oleic acid/sodium dodecyl sulfate (OA/SDS) to have hydrophilic surface. The coated particles could be dispersed in n-Hexane or water. The OA/SDS-coated nanoparticles were used to load an antibiotic drug, chloramphenicol (Cm). Three weight percent of Cm could be loaded onto the OA/SDS coated nanoparticles, which is much higher than amount that can be loaded using the traditional drug loading method.
PACS numbers: 75.50.Mm, 78.67.Bf, 87.83.+a Keywords: Magnetic nanoparticles, Magnetic uids, Magnetite, Drug delivery, Biophysics
I. INTRODUCTION

Magnetic uids consisting of magnetic nanoparticles are the subject of many recent studies, especially in biological applications such as cell separation and sorting, magnetic drug delivery and magnetic resonance imaging [1,2]. On a scale smaller than a critical length, the ferromagnetic behavior of materials disappears and changes to a superparamagnetic (SP) state. The critical length in this case is on the order of few tens of nm at which the thermal agitation dominates the ferromagnetic order. Superparamagnetic materials show strong magnetic properties under an applied magnetic eld, but a nonmagnetic behavior when the eld is removed. The superparamagnetic state is very important in many in-vivo biological applications because it prevents aggregation in the bloodstream. However, in some cases such as hyperthermia, ferromagnetic behavior is required [2]. Magnetite Fe3 O4 NPs are the most commonly used magnetic material due to its high saturation magnetization Ms (90 emu/g), biological compatibility, environmental stability, simple preparation and cost eectiveness. This paper reports the preparation of MNPs by using a coprecipitation method and the use of their superparamagnetism and their large surface areas in applications of drug delivery. Coprecipitation is a simple method to synthesize MNPs, but the particle size is dicult to control. In our studies, the particle size was successfully controlled by changing the concentration of reactants. Magnetite nanoparticles were then functionalized to have
E-mail:

hydrophobic and hydrophilic surfaces to disperse in oil and water and to aim magnetic uids. The drug loading ability of MNPs is also presented.
II. EXPERIMENTS

nhhai@vnu.edu.vn; Fax: +84-4-858-9496

-1332-

Magnetite nanoparticles were synthesized by coprecipitating a Fe2+ /Fe3+ mixed solution (the molar ratio Fe2+ : Fe3+ = 1 : 2) with a NH4 OH solution in air [3]. In a typical reaction, 3.98 g of FeCl2 .4H2 O (molar concentration c = 0.1 M) and 10.82 g of FeCl3 .6H2 O were dissolved in 200 ml of distilled water. An 18 ml solution of NH4 OH 25 % diluted to 100 ml by water, was then added to the Fe2+ /Fe3+ solution at a rate of one drop per second at room temperature with constant stirring. In our study, the concentration of Fe2+ , c (the concentration of Fe3+ ; OH was adjusted accordingly), was changed from 0.002 M to 0.25 M (0.002, 0.005, 0.01, 0.02, 0.05, 0.1, 0.15, 0.2, 0.25) to investigate the in uence of the concentration of the reactants on the size and the magnetic properties of synthesized MNPs. After the reaction, black precipitates of magnetite were washed ve times by using distilled water and magnetic decantation to remove all the soluble substances. 0.5 g of MNPs in 20 ml of water were then vigorously mixed with 10 ml of oleic acid (OA, 9-Octadecenoic acid C18 H34 O2 , CAS number: 112-80-1) in order to coat the MNPs with a single surfactant layer of OA (denoted as OA-coated nanoparticles) and to disperse them into a non-polar solvent n-Hexane. The excess OA was washed o three times by using nHexane and magnetic decantation. For dispersing in water, the OA-coated nanoparticles in 20 ml n-Hexane were

Size Controlled Magnetite Nanoparticles and Their { C. V. Thach et

al.

-1333-

Fig. 2. TEM images of the samples with c = 0.05 M and = 0.25 M.

Fig. 1. XRD pattern of a typical sample with c = 0.2 M. The powder diraction intensities of Fe3 O4 (pdf: 1-1111) are presented by the bars.

mixed with 40 ml of an aqueous solution containing 1 g of sodium dodecyl sulfate (SDS, Sodium lauryl sulfate C12 H25 OSO3 Na, CAS number: 151-21-3) to obtain double layer OA/SDS nanoparticles (denoted as SDS/OA nanoparticles). The excess SDS was washed o three times by using water with the help of magnetic decantation. To load antibiotic into the SDS/OA nanoparticles, 96 mg of Chloramphenicol (Cm) was dissolved in 2.5 ml of ethanol and then diluted to 10 ml by using distilled water. Then, the obtained Cm solution was vigorously mixed with 100 mg of OA/SDS-coated MNPs in 20 ml of water for 15 hours. The unloaded Cm was washed o ve times by distilled water with the aid of magnetic decantation. The achieved sample was then stored at low temperature (5  C) to limit the diusion of Cm from OA/SDS coated MNPs. The structural analysis of a dried powder of MNPs was conducted by using a D5005 X-ray diractometer. The particles sized were determined by using a JEOL JEM 1010 transmission electron microscope (TEM). A cooper grid was dipped in a solution containing OA-coated NPs and then dried before being submitting for TEM. Magnetic properties were measured by using a DMS 880 vibrating sample magnetometer. The thermogravimetric analysis (TGA) was carried out by using a SDT 2960 TA Instrument. Fourier Transformed Infrared (FTIR) Spectra were recorded in the transmission mode on a Nicolet Impact 410 spectrometer.
III. RESULTS AND DISCUSSION

XRD patterns having all the characteristic peaks of magnetite Fe3 O4 conrmed that all samples were magnetite with the inverse spinel structure as in the case of

the bulk sample. Typical XRD patterns of a sample with c = 0.2 M and re ections of standard Fe3 O4 powder are shown in Figure 1. The XRD results do not show the presence of impurities or of other phases. All peaks are broadened due to the small sizes of nanoparticles. The mechanism of MNP formation is understood as a short single burst of nucleation that occurs when the concentration of constituent species reaches critical supersaturation. Then, the nuclei so obtained are allowed to grow uniformly by diusion of solutes from the solution and/or aggregation of other nuclei to their surface until the nal size is attained [4]. In conventional coprecipitation, the size (d) can be controlled by using the pH and the ionic strength [5]. Sizes of 10 { 100 nm can be made by using this technique. A smaller size is dicult to obtain. Changing the concentration of Fe2+ can change the particle size. Transmission electron microscope images of two typical samples with c = 0.05 M and c = 0.25 M are shown in Figure 2. When the solution is diluted (c = 0.05 M), the diusion of solutes is limited, resulting in a limited growth process for the particles. Particles are spherical with a diameter, as determined from the TEM image, of 10:0 2:8 nm. As the concentration increases (c = 0.25 M), the growth process happens much more strongly, which creates a particle size of 14:6 3:2 nm. Nanoparticles should be well separated due to the OA coating layer, but the clusters of NPs can be explained by the drying process in the preparation for TEM measurements. For a small particle size of about 10 nm, a dilution of the solution is needed, but we don't need to decrease c lower than 0.05 M because 10 nm is considered to be the smallest particle size that can be obtained by using a conventional coprecipitation process. The particle size can be even greater than 14.6 nm if c is greater than 0.25 M. However, particles of greater sizes may lose their superparamagnetic behavior, which is unexpected in further applications in drug delivery. Owing to the particle size being smaller than the critical length (superparamagnetic radius), all samples ex-

-1334-

Journal of the Korean Physical Society, Vol. 52, No. 5, May 2008

Fig. 3. Dependence of the saturation magnetization on the concentration and a typical magnetization curve of the sample with c = 0.2 M (the inset).

hibit superparamagnetic properties at room temperature, with their saturation magnetizations depending on the concentration as shown in Figure 3. The magnetization curve of a typical sample with c = 0.2 M (the inset in Figure 3) shows no remanence (the remained magnetic magnetization after removing the applied magnetic eld) and coercivity (the applied magnetic eld required to demagnetize the MNPs). The saturation magnetization increases with the increasing concentration of solutes and reaches 74 emu/g. That value is smaller than that of bulk samples, which is attributed to the core-shell structure of the particles and the possible oxidation by the oxygen diusing into the solutes. In both cases, the higher the concentration, the greater the particle size and/or the lower the oxidation, making the saturation magnetization higher. Bare MNPs can be quickly aggregated to make clusters of particles, which results in unexpected properties. This problem was solved by coating MNPs with a single surfactant layer (OA) or with a double surfactant layer (OA/SDS). TGA measurements of the bare MNPs and of the OA-coated and the OA/SDS-coated MNPs were performed (not shown). The bare MNPs show a small mass loss of 1 % over the whole range of temperature from room tempearture to 500  C, which can be ascribed to the remaining water trapped in the samples. OA-coated NPs show a loss of 7 % in mass in the range of 220  C { 420  C, which can be explained by the evaporation of OA. However, when compared with the boiling temperature of OA 150  C { 250  C, this higher range suggests bonds between OA and the MNP surface, which make the OA molecule more dicult to evaporate. This fact is also supported by the FTIR spectra. The OA/SDScoated NPs shows two mass losses: 3 % in the range of 200  C { 300  C (OA evaporation dominated) and 5 %

Fig. 4. FTIR spectrum of bare MNPs, OA-coated MNPs and pure OA (upper) and a schematic illustration (lower left) of a single layer OA-coated and (lower right) of a double layer OA/SDS-coated MNPs. The inward and the outward circles represent the hydrophilic heads of the OA and the SDS surfactants, respectively. The black zigzag curves represent the hydrophobic tails.

in the range of 300  C { 400  C. FTIR measurements of bare MNPs, OA-coated MNPs and OA are shown in Figure 4. In the FTIR spectra of pure OA, there are absorbance peaks of the C=O stretch dimer H-bonded (1707 cm1 ), the O-H stretch dimer H-bonded (broad peak at 3000 cm1 ), the symmetric and asymmetric CH2 stretching modes (2861 cm1 , 2932 cm1 ), the C=O band (1280 cm1 ), the O-H inplane band (1460 cm1 ) and O-H out-of-plane band (910 cm1 ). The spectra of OA-coated NPs, however, reveal the absence of the all peaks for C=O and O-H whereas two peaks of the CH2 stretching modes are still present, suggesting binding of the carboxylic group of OA to the iron oxide nanoparticles [6]. This indicates chemisorption of OA hydrophilic heads at the NP surface and of hydrophobic tails of the hydrocarbon chain at the carrier liquid (the lower left in Figure 4). With the presence of SDS molecules, the hydrophobic tail of this molecule will tend inward to the particle, as shown in Figure 4 (the

Size Controlled Magnetite Nanoparticles and Their { C. V. Thach et

al.

-1335IV. CONCLUSIONS

Using a coprecipitation method, we can synthesize superparamagnetic MNPs with their size (d) being controlled by changing the concentration of reactants (c): d = 10:0 2:8 nm for c = 0.05 M and d = 14:6 3:2 nm for c = 0.25 M. The saturation magnetization of MNPs increases with increasing concentration and can reach 74 emu/g. OA molecules bond by chemisorption with MNPs and contribute 7 % of the mass of OA-coated MNPs. Another surfactant, SDS, surrounds the OAcoated MNPs, creating a double layer OA/SDS-coated MNPs double layer occupies 8 % of mass. Chloramphenicol is loaded into the space between the two surfactant layers thanks to the hydrophobic anity and contributes 3 % of the mass of whole system.
Fig. 5. Temperature dependence of the normalized mass of Cm-loaded OA/SDS-coated MNPs and OA/SDS-coated MNPs.
ACKNOWLEDGMENTS

lower right) and leave a hydrophobic space between two layers. We use this feature for drug loading, as presented in the experiment section. Chloramphenicol (Cm), a hydrophobic antibiotic drug, was loaded into the hydrophobic space between the two surfactant layers. The mechanism of this process is the hydrophobic anity [6]. The amount of the loaded Cm was estimated by using TGA measurements on the OA/SDS-coated MNPs and the Cm-loaded OA/SDScoated MNPs (Figure 5). From the total mass loss of the two samples, 10 % and 13 %, respectively, we claim that the loaded Cm occupies 3 % of the mass. This value is smaller than recently achieved result [6], but very large in comparison with that obtained by using traditional methods, which are normally less than 1 %. It is reasonable to expect our system to be a good candidate for applications in magnetic drug delivery. That, Cm was successfully loaded into OA/SDS-coated MNPs is also evident in the antibiotic behavior test of Cm-loaded OA/SDS-coated MNPs [7]. The mechanism of loading Cm into OA/SDS MNPs is quite simple, but versatile, because it is applicable for all the hydrophobic drugs frequently used in pharmacy.

The project is nancially supported by the Vietnamese Fundamental Research Program for Natural Sciences (Project 406506).
REFERENCES

[1] D. L. Leslie-Pelecky, V. Labhasetwar and R. H. Kraus, Jr., Nanobiomagnetics, in Advanced Magnetic Nanostructures, edited by D. J. Sellmyer and R. S. Skomski (Kluwer, New York, 2005). [2] Q. A. Pankkhurst, J. Connolly, S. K. Jones and J. Dobson, J. Phys. D: Appl. Phys. 36, R167 (2003). [3] P. Berger, N. B. Adelman, K. J. Beckman, D. J. Campbell, A. B. Ellis and G. C. Lisensky, J. Chem. Edu. 76, 943 (1999). [4] P. Tartaj, M. P. Morales, S. Veintemillas-Verdaguer, T. Gonzalez-Carreno and C. J. Serna, J. Phys. D: Appl. Phys. 36, R182 (2003). [5] J. P. Jolivet, Metal Oxide Chemistry and Synthesis: From Solutions to Solid State (Willey, New York, 2000). [6] T. K. Jain, M. A. Morales, S. K. Sahoo, D. L. LesliePelecky and V. Labhasetwar, Mol.Pharm. 2, 194 (2005). [7] N. T. Khuat, C. V. Thach, V. A. T. Nguyen, N. H. Hai, T. N. Phan and N. Chau, J. Korean Phys. Soc., in this issue.