Solvent-free atom transfer radical polymerization for the preparation of poly(poly(ethyleneglycol) monomethacrylate)-grafted Fe3O4 nanoparticles: Synthesis, characterization and cellular uptake

ARTICLE IN PRESS

Biomaterials 28 (2007) 5426–5436 www.elsevier.com/locate/biomaterials

Solvent-free atom transfer radical polymerization for the preparation of poly(poly(ethyleneglycol) monomethacrylate)-grafted Fe3O4 nanoparticles: Synthesis, characterization and cellular uptake Qu-Li Fana,b,c, Koon-Gee Neohb,, En-Tang Kangb, Borys Shutera, Shih-Chang Wanga a

Department of Diagnostic Radiology, National University of Singapore, Kent Ridge, Singapore 119260, Singapore Department of Chemical and Biomolecular Engineering, National University of Singapore, Kent Ridge, Singapore 119260, Singapore c Institute of Advanced Materials (IAM), Nanjing University of Posts and Telecommunications, 66 XinMoFan Road, Nanjing 210003, China b

Received 9 June 2007; accepted 29 August 2007 Available online 24 September 2007

Abstract Poly(poly(ethyleneglycol) monomethacrylate) (P(PEGMA))-grafted magnetic nanoparticles (MNPs) were successfully prepared via a solvent-free atom transfer radical polymerization (ATRP) method. The macroinitiators were immobilized on the surface of 6.470.8 nm Fe3O4 nanoparticles via effective ligand exchange of oleic acid with 3-chloropropionic acid (CPA), which rendered the nanoparticles soluble in the PEGMA monomer. The so-obtained P(PEGMA)-grafted MNPs have a uniform hydrodynamic particle size of 36.071.2 nm. The successful grafting of P(PEGMA) on the MNP surface was ascertained from FTIR and XPS analyses. The uptake of the MNPs by macrophage cells is reduced by two-orders of magnitude to o2 pg Fe/cell after surface grafting with P(PEGMA). Furthermore, the morphology and viability of the macrophage cells cultured in a medium containing 0.2 mg/mL of P(PEGMA)-grafted MNPs were found similar to those of cells cultured without nanoparticles, indicating an absence of significant cytotoxicity effects. T2weighted magnetic resonance imaging (MRI) of P(PEGMA)-grafted MNPs showed that the magnetic resonance signal is enhanced significantly with increasing nanoparticle concentration in water. The R1 and R2 values per millimole Fe, and R2/R1 value of the P(PEGMA)-grafted MNPs were calculated to be 8.8 mM1 s1, 140 mM1 s1, and 16, respectively. These results indicate that the P(PEGMA)-grafted MNPs have great potential for application in MRI of specific biotargets. r 2007 Elsevier Ltd. All rights reserved. Keywords: ATRP; Solvent-free; PEGMA; Magnetic nanoparticle; MRI

1. Introduction Recently, superparamagnetic nanoparticles of iron oxides have shown great potential in bioapplications, including magnetic resonance imaging (MRI) contrast enhancement [1,2], drug delivery [3,4], bioseparation [5,6], tissue repair [5,6], hyperthermia [1,5], and magnetofection [6]. In these fields, the preparation of the monodispersed magnetic nanoparticles (MNPs) and the introduction of functionalities on the surface of these MNPs through grafting of water soluble, biocompatible groups and targeting ligands are much desired. Corresponding author. Tel.: +65 6516 2176; fax: +65 6779 1936.

E-mail address: [email protected] (K.-G. Neoh). 0142-9612/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2007.08.039

Polymer coating of MNPs is now one of the most attractive methods to realize such functionalities because the polymer shell offers flexibility in controlling the chemical composition and functional groups on the nanoparticle surface [6]. Most of such MNP surface modification approaches utilize the ‘‘grafting to’’ strategy, which involves the coating of MNPs with pre-existing polymers via either hydrophobic interaction [7–16] or deposition [17–19] through electrostatic affinity. Although this strategy is flexible, the polymers tend to coat more than one nanoparticle to form nanoparticle clusters. On the other hand, the ‘‘grafting from’’ method, which propagates the polymer chains from the nanoparticle surface, is a good candidate to form small polymer-coated single nanoparticles. Furthermore, with this method, surface modification

ARTICLE IN PRESS Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

with high graft density and high stability of the polymer shell can be achieved. As such, surface-initiated atom transfer radical polymerization (ATRP) appears a promising ‘‘grafting from’’ technique which offers many advantages including good control over molecular weight and monodispersity, and therefore thickness of the polymer shell [20–24]. Another advantage of ATRP is that the endfunctionalized polymers or block copolymers grafted onto the nanoparticle surface can offer a variety of active sites for further multi-biofunctionalization for specific targeting. Despite these advantages, most of the research on MNP surface-initiated ATRP focused on the formation of polystyrene or poly(methyl methacrylate) hydrophobic shells on the MNPs, which have limited potential for bioapplication. Compared with the more widely reported bioapplications of hydrophilic polymer-coated MNPs via the ‘‘grafting to’’ method, the preparation of hydrophilic polymer-grafted MNPs through ATRP and more importantly their bioapplications still need further investigation. Recently several hydrophilic polymer-coated MNPs through surface-initiated ATRP in polar solvents were successfully developed by our group and Hatton’s group [25–27]. In addition, their possible bioapplications were also investigated [25,26]. However, their potential bioapplication for MRI which is one of the most important application fields for MNPs has not been explored. Among the hydrophilic monomers, poly(ethyleneglycol) monomethacrylate) (PEGMA), a poly(ethyleneglycol) (PEG) derivative which contains one pendant hydroxyl group in every monomer, is one of the most attractive monomers for future bioapplication [25]. PEG has been well proven to be nonimmunogenic, nonantigenic, and protein resistant [28– 30]. Consequently, PEG has been introduced on the MNPs surface via the ‘‘grafting to’’ method to enhance their biocompatibility [31–34]. Recently grafting poly (PEGMA) or P(PEGMA), instead of PEG, via ‘‘grafting from’’ method showed that it will not only prevent rapid clearing by macrophages, but also provide more sites for further biofunctionalization [25]. However, their further bioapplication was impeded by the poor control over the rate and extent of the ATRP process in solvent due to the formation of flash polymerization when higher monomer concentration is used for ATRP. We herein present a relative simple and scalable approach for preparing P(PEGMA)-grafted Fe3O4 core– shell nanoparticles with well-controlled properties using a solvent-free ATRP. Yang and co-workers first reported the solvent-free ATRP of styrene on Fe3O4 nanoparticles and showed that with this method the desorption of the initiator molecules from the particle surfaces could be reduced, and the polymerization occurred favorably on the macroinitiator surfaces [21], giving rise to monodispersed polymer-coated nanoparticles. In our present work, the initiator, 3-chloropropionic acid (CPA), used for ATRP [20] at inorganic surfaces was introduced through ligand exchange [20–23,27] with the oleic acid on the surface of the pristine nanoparticles prepared from the high-tempera-

5427

ture decomposition of iron acetylacetonate [35]. The ligand exchange method enables the solubility of the MNP to be changed to match the requirement of solvent-free ATRP since surface capping agents can be exchanged in a controllable fashion, depending on the functional groups and concentrations of the surfactants [20–23]. The physical properties, cytotoxicity and MRI of the P(PEGMA)grafted MNPs from solvent-free ATRP were investigated and the results showed that these MNPs may be a good candidate for bioapplications utilizing MRI. 2. Experimental 2.1. Materials Benzyl ether, 1,2-hexadecanediol, oleic acid, oleylamine, iron(III) acetylacetonate, 3-chloropropionic acid, 2,20 -bipyridyl (Bpy), and copper(I) chloride were purchased from Aldrich Chemical Co. and used as received. Poly(ethyleneglycol) monomethacrylate macromonomer (Mn360) was passed through a silica gel column to remove the inhibitor and stored under an argon atmosphere at 10 1C. Mouse macrophage cells (RAW 264.7) and 3T3 fibroblasts were purchased from ATCC. RPMI-1640 medium, Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum, L-glutamine, penicillin, and streptomycin were purchased from Sigma. All other solvents and chemicals were purchased from either Fisher Scientific or Aldrich and used as received.

2.2. Preparation of magnetic nanoparticles The Fe3O4 MNPs were prepared according to a previously reported method [35]: iron(III) acetylacetonate (1.766 g, 0.337 mL, 5 mmol), 1,2hexadecanediol (6.461 g, 25 mmol), oleic acid (4.237 g, 4.761 mL, 15 mmol), oleylamine (4.012 g, 4.935 mL, 15 mmol), and benzyl ether (50 mL) were mixed and magnetically stirred under a flow of nitrogen. The mixture was heated to 200 1C for 2 h and then, under a blanket of nitrogen, heated to reflux (300 1C) for another 1 h. The black mixture was cooled to room temperature after removal of the heat source. Under ambient conditions, ethanol (100 mL) was added to the mixture, and a black material was precipitated and separated via centrifugation. The black product was dissolved in 40 mL of hexane in the presence of oleic acid (1 mL) and oleylamine (1 mL). Centrifugation (6000 rpm, 10 min) was applied to remove any undispersed residue. The product, Fe3O4 nanoparticles, was then precipitated with ethanol, and collected by centrifugation (6000 rpm, 10 min). The nanoparticles were then dried under reduced pressure and stored at 0–4 1C.

2.3. Ligand exchange of magnetic nanoparticles A 100 mg of oleic acid-stabilized nanoparticles were dispersed in 100 mL of 0.25 M CPA in hexane and stirred for 24 h at room temperature under the protection of argon. The resulting black precipitate was separated using a centrifuge at 6000 rpm for 4 min and washed three times with hexane to remove the excess initiator. The nanoparticles were then dried under reduced pressure and stored at 0–4 1C.

2.4. Solvent-free atom transfer radical polymerization (Fig. 1(a)) The CPA-stabilized nanoparticles (60.0 mg) were first dissolved in PEGMA monomer (6 mL in a Pyrex tube containing a magnetic stir bar) to form a transparent brownish solution. The solution was purged with argon for 15 min, and CuCl (5.1 mg) and Bpy (24 mg) were then added. After purging with argon for another 10 min, the Pyrex tube was sealed and kept in a 30 1C water bath for 18 h under stirring. After the reaction was complete, the mixture was diluted with tetrahydrofuran (THF) at a

ARTICLE IN PRESS Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

5428

reaction mixture/THF volume ratio of 1:5. The dilute solution was first centrifuged at 4000 rpm for 4 min to remove the Cu(II) precipitate formed in the ATRP process and then dripped into 200 mL of diethyl ether to precipitate the final deep-brownish products. The collected nanoparticles were re-dispersed in ethanol and centrifuged at a speed of 2000 rpm for 4 min to further remove any agglomerates and remaining Cu(II) precipitate. The nanoparticles were reprecipitated three times with diethyl ether and dried under vacuum before further characterization.

2.5. Cell culture Mouse macrophage cells (RAW 264.7) were used to assess the intracellular uptake of the MNPs. The cells were routinely cultured in RPMI-1640 medium, supplemented with 10% fetal bovine serum, 2 mM Lglutamine, 100 IU/mL penicillin, and 100 mg/mL streptomycin at 37 1C in 5% CO2 atmosphere. The cultured cells were washed with PBS and detached with trypsin-EDTA solution. The cells were then collected by centrifugation and resuspended with the medium containing nanoparticles at a concentration of 0.2 mg/mL to achieve a cell concentration of 105 cells/mL. The cells were then seeded in 24-well culture plates. In the control experiment, the cells were cultured in the medium at the same cell concentration but without MNPs. After incubation at 37 1C for prescribed time periods, the cells were washed three times with PBS to remove the nanoparticles in the medium and then detached with trypsin-EDTA solution. After counting with a hemocytometer, the cells were collected by centrifugation, and the cell pellet was dissolved in 37% HCl at 60 1C for 2 h. The iron concentration was determined using Thermal Jarrell Ash Duo Iris Inductively Coupled Plasma-Optical Emission Spectrometer (ICP-OES). The cytotoxicity of the P(PEGMA)-grafted MNPs was evaluated by determining the viability of 3T3 fibroblasts after incubation with DMEM (supplemented with 10% fetal bovine serum, 1 mM L-glutamine, 100 IU/ mL penicillin) containing the nanoparticles at a concentration of 0.2 mg/ mL. Cell viability testing was carried out via the reduction of the MTT reagent (3-[4,5-dimethyl-thiazol-2-yl]-2,5-diphenyltetrazolium bromide, Sigma). The MTT assay was performed in a 96-well plate following the standard procedure with minor modifications. The nanoparticles were sterilized with UV irradiation for 30 min before use. Control experiments were carried out using the complete growth culture media only (nontoxic control), and with 1% Triton X-100 (Sigma) (toxic control). 3T3

fibroblasts were seeded at a density of 104 cells per well for 24 h before the medium was replaced with one containing the MNPs at 0.2 mg/mL. The 3T3 fibroblasts were incubated at 37 1C and 5% CO2 for 24 h. The culture media from each well was then removed and 90 mL of media and 10 mL MTT solution (5 mg/mL in PBS) were then added to each well. After 2 h of incubation at 37 1C and 5% CO2, the media were removed and the formazan crystals were solubilized with 100 mL DMSO for 15 min. The optical absorbance was then measured at 560 nm on a microplate reader (Tecan GENios). The results were expressed as percentages relative to the results obtained with the nontoxic control. The differences in the results obtained from P(PEGMA)-grafted MNPs and the controls were analyzed statistically using the two-sample t-test. The differences observed between samples were considered significant for Po0.05.

2.6. Characterization FT-IR spectra were obtained in a transmission mode on a Bio-Rad FTIR spectrophotometer (Model FTS135) under ambient conditions. Samples of pristine and functionalized MNPs were ground with KBr and then compressed into pellets. The spectrum was taken from 400 to 4000 cm1. Typically, 64 scans at a resolution of 8 cm1 were accumulated to obtain one spectrum. The chemical composition of the pristine and the functionalized MNPs was determined with X-ray photoelectron spectroscopy (XPS) on an AXIS HSi spectrometer (Kratos Analytical Ltd.) using a monochromatized Al Ka X-ray source (1486.6 eV photons) at a constant dwell time of 100 ms and a pass energy of 40 eV. The anode voltage was 15 kV, and the anode current was 10 mA. The pressure in the analysis chamber was maintained at 6.7  106 Pa or lower during each measurement. The MNPs were mounted on standard sample studs by means of double-sided adhesive tape. The core-level signals were obtained at a photoelectron takeoff angle of 901 (with respect to the sample surface). To compensate for surface charging effect, all core-level spectra were referenced to the C 1s hydrocarbon peak at 284.6 eV. In spectral deconvolution, the line width (full-width at half-maximum) of the Gaussian peaks was maintained constant for all components in a particular spectrum. The peak ratios for various elements were corrected using experimentally determined instrumental sensitivity factors. Thermogravimetric analysis (TGA) data were obtained with a TGA 2050 thermogravimetric analyzer (TA Instruments). Samples weighing between 5 and 15 mg were heated from 30 to 700 1C at a heating rate of

CPA Hexane

b CPA-stabilized MNP

Oleic acid-stabilized MNP CuCl, Bpy PEGMA Oleic Acid CPA O C nCl O H(OCH2CH2)mOOC P(PEGMA)-grafted MNP Fig. 1. (a) Preparation route of P(PEGMA)-grafted magnetic nanoparticles by solvent-free ATRP. (b) Photograph of P(PEGMA)-grafted Fe3O4 nanoparticles dispersed in water.

ARTICLE IN PRESS Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436 10 1C/min in air. Transmission electron microscopy (TEM) images were recorded on a JEOL 2010 transmission electron microscope at an accelerating voltage of 200 kV. The TEM specimens were made by placing a drop of the nanoparticle suspension on a carbon-coated copper grid. Measurement of magnetization was carried out with a vibrating sample magnetometer (VSM) (Model 1600, DMS). The hydrodynamic size of the P(PEGMA)-grafted MNPs was determined by dynamic light scattering (DLS) using a 90 Plus particle size analyzer (Brookhaven Instruments). The molecular weight of the P(PEGMA) brushes grafted on the MNP surface was determined by gel permeation chromatography (GPC). The P(PEGMA)-grafted nanoparticles sample was first dissolved in 10% HCl for 5 h. The HCl was then neutralized with 1 M NaOH. After freeze-drying, the polymer was redissolved in THF and centrifuged to remove the salts. The polymer was recovered after removal of the THF under reduced pressure and the molecular weight measurement was performed on a Waters Breeze GPC equipped with a Waters 1515 HPLC pump and a Waters 2414 refractive index detector, using poly(ethylene glycol) as the calibration standard. Water was used as the eluent at a flow rate of 1 mL/min. MRI experiments were performed at 25 1C in a clinical magnetic resonance (MR) scanner (Siemens Symphony; 1.5T). To demonstrate the T1 and T2 effects in an aqueous solution, P(PEGMA)-grafted MNP were suspended in tubes of water (20 mL) with iron concentration at 0.03125, 0.0625, 0.125, 0.25, and 0.625 mM Fe, respectively. The tubes were placed into the MR scanner and a number of MR sequences were run, spin-echo for R2 (32 echoes; TR: 1600 ms; TE: 15–480 ms) and R1 (7 TRs; TR: 100– 6400 ms; TE: 15 ms) determination. Spin-echo R2 sequences (n ¼ 8) were also obtained over a period of 6.25 h while the samples remained in the scanner. The relaxation rates for each sample were computed using inhouse software (MATLAB V7) by fitting appropriate exponential functions.

3. Results and discussion

5429

polymerization time of 18 h and the polydispersity is 1.22. This low polydispersity indicates that the ATRP of PEGMA was successfully accomplished on the surface of the Fe3O4 nanoparticles. Compared with the molecular weight (14,100) of P(PEGMA) from ATRP in aqueous solution after 2 h [25], the molecular weight of P(PEGMA) from solvent-free ATRP after 18 h is relatively low. We also found that unlike the significant increase in viscosity of the PEGMA aqueous solution after 2 h ATRP, the viscosity of the PEGMA monomer solution after 18 h solvent-free ATRP still remained low. These results indicate that the polymerization rate and the resulting molecular weight are easier to be controlled by solvent-free ATRP than by ATRP in aqueous solution. It is well known that the high polarity of water may enhance the reaction rate of ATRP [36]. Thus, the solvent-free method is anticipated to decrease the solvent effect on ATRP and provide better control of the chain propagation of P(PEGMA). Furthermore, the solvent-free ATRP showed higher scalability than ATRP in water. With ATRP in water, when the MNP concentration was increased from 10 to 50 mg/mL, the ATRP cannot be controlled and gelation of the reaction medium occurred within 30 min. On the other hand, no obvious difference was found with a similar concentration change when solvent-free ATRP was used. This is attributed to the difference in polymerization rate of the two methods. Therefore, solvent-free ATRP has a greater potential for preparing P(PEGMA)-grafted MNPs via a scalable process.

3.1. Physical properties of the magnetic nanoparticles 3.2. FT-IR characterization FT-IR spectra of oleic acid-stabilized MNPs before and after ligand exchange with CPA are shown in Fig. 2(a) and (b). It is obvious that samples after ligand exchange had relatively weaker absorption bands of CH2 groups and

(a)

Transmittance (a. u.)

The preparation route of the P(PEGMA)-grafted MNPs by solvent-free ATRP is shown schematically in Fig. 1(a). Both oleic acid and CPA-stabilized nanoparticles are black powders. The oleic acid-stabilized nanoparticles can disperse very well in nonpolar solvents, such as hexane, toluene, etc. However, after these nanoparticles were stirred in a hexane solution of CPA for 12 h, the collected nanoparticles cannot be redissolved in hexane. After this treatment, these nanoparticles show good solubility in polar solvents, such as THF, DMF and ethanol. This confirms that these particles have very different surface chemistry from those stabilized by oleic acid, indicating the successful ligand exchange with CPA. It is intriguing to find that such CPA-stabilized nanoparticles, i.e. the nanoparticles with macroinitiators, dissolved very well in PEGMA monomer without the addition of any solvent, which makes the solvent-free ATRP of PEGMA on the MNP surface realizable. After polymerization, the collected nanoparticles can be dissolved in water to form a clear deep brown solution (Fig. 1(b)), indicating the successful polymerization of the PEGMA monomer. Powder X-ray diffraction (XRD) analysis confirmed that the crystalline property of the P(PEGMA)-grafted MNPs corresponds to magnetite (Fe3O4) when compared to those reported [34]. The molecular weight (Mn) of the P(PEGMA) was determined by GPC to be 8600 after a

2852 2923

(b)

2963 (c)

1548

1444

1726 3500

3000

2500

2000

1500

1110 1000

500

Wavenumber (cm-1)

Fig. 2. FT-IR spectra of oleic acid-stabilized Fe3O4 nanoparticles (a) before and (b) after ligand exchange with CPA, and (c) P(PEGMA)grafted Fe3O4 nanoparticles by solvent-free ATRP.

ARTICLE IN PRESS 5430

Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

stronger bands of carboxylate groups than the sample before ligand exchange, judging by the characteristic symmetric (sym, 2852 cm1) and asymmetric (asym, 2923 cm1) stretch in the methylene groups and symmetric (sym, 1440 cm1) and asymmetric (asym, 1550 cm1) CQO vibrations of carboxylate groups. This observation clearly indicates that the oleic acid on the surface of nanoparticles was successfully replaced by CPA, since CPA contains a much higher ratio of carboxylate group to CH2 group than oleic acid. It is noteworthy that the alkyl C–H vibration bands of CH3 asymmetric stretch (2963 cm1) mode could be observed in both samples [21], suggesting that an equilibrium exists in this ligand exchange process and not all oleic acid molecules were replaced from the iron oxide surfaces since the CH3 group is only present in oleic acid and not in CPA. In Fig. 2(c), the FT-IR spectrum of the MNPs modified with P(PEGMA) shows a strong band at 1110 cm1 attributed to the C–O–C ether stretch, and a band at 1726 cm1 assigned to the CQO stretch. The spectrum also displays a strong band around 2900 cm1 corresponding to the CH2 stretching vibrations. The CH2, C–O–C, and CQO absorption bands provide strong evidence that the MNPs have been modified with PEGMA. No strong absorption band at around 1630 cm1, which is the characteristic band for CQC stretching vibration of the PEGMA monomer, was found in Fig. 2(c). The absence of CQC stretching vibration further confirms the successful polymerization of the PEGMA monomer on the MNPs. 3.3. XPS Ligand exchange of oleic acid by CPA on the surface of MNPs was also confirmed with XPS. The photoelectron lines at binding energy (BE) of about 55, 284, 529, and 710 eV which are attributed to Fe 3p, C 1s, O 1s, and Fe 2p, respectively, are observed in the wide scan spectrum of the pristine MNPs (Fig. 3(a1)). Two peak components are observed in the C 1s core-level spectrum of the oleic acidstabilized MNPs (Fig. 3(a2)). The peak at about 284.6 eV is attributed to C–C and C–H, and the peak at 288.3 eV is attributed to COOH, indicating that the MNPs are well covered with the oleic acid layer. In addition to the photoelectron lines in Fig. 3(a1), another photoelectron line at BE about 200 eV which is attributed to Cl 2p is observed in the wide scan spectrum of the CPA-stabilized nanoparticles (Fig. 3(b1)). The C 1s core-level spectrum of the CPA-stabilized nanoparticles (Fig. 3(b2)) can be curvefitted with three peak components having BE at about 288.5, 285.9, and 284.5 eV, attributable to the COOH, C– Cl, and C–C/C–H species, respectively. The appearance of the C–Cl species confirms the presence of CPA-stabilized on the nanoparticles surface. Fig. 3(b3) shows the Cl 2p spectrum of CPA-stabilized nanoparticles surface, where the peak components at about 199.9 and 201.5 eV are assigned to C–Cl 2p3/2 and C–Cl 2p1/2, respectively. All the XPS results are in accordance with the FT-IR results,

indicating the successful ligand exchange of oleic acid by CPA on the surface of the MNPs. The XPS wide scan spectrum of the P(PEGMA)-grafted MNPs via solvent-free ATRP is almost identical to that reported earlier for P(PEGMA)-grafting on MNPs via ATRP in a polar solvent [25]. It has two prominent photoelectron lines at BE of about 284 and 530 eV which are attributed to C 1s and O 1s, respectively. A very weak photoelectron line at BE of 200 eV is also discernible and that is assigned to the preserved active chlorine groups from ATRP. The C 1s core-level spectrum of the P(PEGMA)-grafted nanoparticles can be curve-fitted with three peak components having BE at about 284.6, 286.2, and 288.5 eV, attributable to the C–C/C–H, C–O, and OQC–O species, respectively. The area ratio of the three peaks is 4:11:1, which is close to 3:12:1 expected of P(PEGMA). Similarly, the photoelectron line of Fe 2p at 710 eV is not discernible, indicating the P(PEGMA) layer on the surface of the MNPs is thicker than the XPS probing depth (5 nm) [25].

3.4. TEM Fig. 4 shows representative TEM images of as-synthesized Fe3O4 nanoparticles before and after ligand exchange and after P(PEGMA) grafting. The oleic acid-stabilized Fe3O4 nanoparticles (Fig. 4(a)) are quite uniform with a particle size of 6.470.8 nm. From Fig. 4(b), it is clear that there is a little change in the particle size after ligand exchange although some particles assume a more cubic rather than spherical morphology. A previous report has shown that ligand exchange will cause some morphological change in the nanoparticles [21]. Thus, the observed morphological change of our prepared MNPs also provides an indication that oleic acid molecules on the surfaces of Fe3O4 nanoparticles have been replaced by the initiator. Fig. 4(c) shows the TEM image of the nanoparticles after P(PEGMA) grafting. In the TEM image, only the core of Fe3O4 can be observed. The P(PEGMA) shell is not discernible due to the lack of contrast with the background. It can be observed clearly from Fig. 4(c) that the P(PEGMA)-grafted nanoparticles with single Fe3O4 core disperse very well in aqueous solution, and their morphologies and sizes appeared to be comparable to those CPAstabilized nanoparticles before the solvent-free ATRP. To complement the TEM images which only provide information on the size of the Fe3O4 core, DLS measurements were also carried out to determine the hydrodynamic particle size of whole nanoparticles, (i.e. inclusive of the polymer shell and the magnetite core). The DLS measurements revealed that the P(PEGMA)-grafted MNPs obtained from the ‘‘grafting from’’ method possess a relatively narrow size distribution with a mean size of 36.071.2 nm. These results show that the solvent-free ATRP of PEGMA on MNP surface is an excellent technique for preparing

ARTICLE IN PRESS Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

(a1 )

5431

(a2 )

C 1s

C 1s C-C C-H

Fe 2p O 1s

Fe 3p

0

200

COO H

400

600

800

1000

Intensity (a. u.)

(b1 )

282

284

(b 2 )

286

200

C-Cl

400

600

800

1000

282 (b3 )

284

286

290 C 1s

C-C C-H

Cl 2p

0

288

COO H

288

290 Cl 2p

Cl 2p3/2

Cl 2p1/2

196 Binding Energy (eV)

198

200

202

204

Fig. 3. (a1) XPS wide scan and (a2) C 1s core level spectra of oleic acid-stabilized Fe3O4 nanoparticles before ligand exchange; (b1) XPS wide scan, (b2) C 1s, and (b3) Cl 2p core level spectra of Fe3O4 nanoparticles after ligand exchange with CPA.

monodispersed water-soluble nanoparticles with single iron oxide core. 3.5. TGA TGA studies were carried out for the MNPs modified with oleic acid, CPA, and P(PEGMA). The MNPs after each stage of modification give their distinctive TGA curves, which provide indications of the amount of initiator and P(PEGMA) on the MNPs. In Fig. 5(a), the weight loss from 205 to 400 1C is attributed to the oxidation of oleic acid. The TGA curve of the pristine MNPs shows a weight loss of about 29% after the MNPs

were heated to 700 1C. Assuming that all Fe3O4 has been oxidized to Fe2O3, the weight of the organic fraction in oleic acid-stabilized MNPs is calculated to account for 32% of the total weight. The TGA curve of the MNPs stabilized with CPA (Fig. 5(b)) shows that the weight loss is initiated at a lower temperature than in Fig. 5(a) but the total weight loss (33%) is comparable to that of the pristine MNPs. Since the FT-IR results have indicated that CPA was successfully introduced on the MNP surface through the ligand exchange process, the first weight loss step from 160 to 220 1C is attributed to the oxidation of the CPA initiator. The second weight loss step from 220 to 400 1C would be due to the residual oleic acid. From the weight

ARTICLE IN PRESS 5432

Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

100 (b)

(a)

Weight (%)

80

60

(c)

40

0

100

200

300 400 500 Temperature (°C)

600

700

800

Fig. 5. TGA curves of (a) oleic acid-stabilized Fe3O4 nanoparticles, (b) Fe3O4 nanoparticles after ligand exchange with CPA, and (c) P(PEGMA)grafted Fe3O4 nanoparticles by solvent-free ATRP.

loss curve, it is estimated that the CPA initiator accounts for about 15% of the weight of the CPA-stabilized nanoparticles. Fig. 5(c) shows the TGA curve of P(PEGMA)-grafted MNPs. This curve shows that there are two predominant weight loss steps (completed at 340 and 420 1C respectively) upon heating in air. The weight loss of the P(PEGMA)-grafted nanoparticles at 700 1C increases to 67% (compared to 33% for the MNPs after modification with CPA), consistent with the increase in the organic content. 3.6. Cell uptake

Fig. 4. TEM images of (a) oleic acid-stabilized Fe3O4 nanoparticles before ligand exchange, (b) Fe3O4 nanoparticles after ligand exchange with CPA, and (c) P(PEGMA)-grafted Fe3O4 nanoparticles by solvent-free ATRP.

Intravenous injection is the most useful method for MNPs to reach target organs and tissues, because all vital cells receive supplies by means of blood circulation. When particles with a largely hydrophobic surface are injected into the bloodstream, they are prone to be quickly coated by components of the circulation, such as plasma proteins (opsonization process). These particles are endocytosed/ phagocytosed by the circulating monocytes or the fixed macrophages, leading to their rapid elimination from blood circulation. Thus, for nanoparticles to achieve a plasma half-life as long as possible to increase their probability of attaining the desired target, avoidance of macrophage recognition is necessary. This can be achieved with particles that are surface-protected by a layer of hydrophilic groups. Thus, our MNPs grafted by hydrophilic P(PEGMA) chains were anticipated to decrease the possibility of macrophage recognition, and their uptake by mouse macrophage cells was investigated. Fig. 6 shows the optical microscopy images of macrophage cells incubated in medium without and with CPAstabilized MNPs and P(PEGMA)-grafted MNPs at a concentration of 0.2 mg/mL for up to 48 h. It is clear that the CPA-stabilized MNPs have low solubility in the medium and the MNPs form agglomerations outside the

ARTICLE IN PRESS Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

5433

Fig. 6. Optical microscopy images of RAW 264.7 cells in control culture (without any nanoparticles) for (a1) 0 h and (a2) 48 h; cells after culturing in medium containing CPA-stabilized nanoparticles (0.2 mg/mL) for (b1) 0 h and (b2) 48 h; and cells after culturing in medium containing P(PEGMA)grafted nanoparticles (0.2 mg/mL) for (c1) 0 h and (c2) 48 h. Scale bar ¼ 50 mm.

macrophage cells (compare Fig. 6(b1) with (a1)). On the other hand, the P(PEGMA)-grafted MNPs dissolve well and no aggregation can be seen in the medium (Fig. 6(c1)). After incubation of the cells in the medium with either CPA-stabilized or P(PEGMA)-grafted MNPs for 48 h, the number of cells increases and the cells retain a similar morphology (Fig. 6(b2) and (c2)) as that of the control cells (Fig. 6(a2)). Comparing Fig. 6(b1) with (b2), it can be seen that in the latter, the area around each macrophage cell is devoid of the nanoparticles, indicating that most of the CPAstabilized MNPs surrounding the cells have been taken up by the macrophage cells. For the P(PEGMA)-grafted MNPs, because of their ultrasmall particle size, the extent of uptake of the nanoparticles by the cells would not be evident from optical microscopy. Thus, the ICP-OES was used to quantify the amount of iron internalized by the cells.

Fig. 7 shows the uptake of the oleic acid-stabilized, CPAstabillized and P(PEGMA)-grafted MNPs by macrophage cells. The oleic acid MNPs and CPA-stabilized MNPs were quickly internalized into the cells in the first day, resulting in an uptake of 154 and 95 pg Fe/cell, respectively. The amount taken in by the cells decreased over the following several days. This may be due to the rapid growth and division of the macrophage cells [31]. After the MNPs were grafted by P(PEGMA), the nanoparticle uptake by macrophage cells decreased dramatically. Only 1 pg Fe/cell was reached in the first day, which was much lower than what was observed with oleic acid or CPA-stabilized MNPs. The amount of P(PEGMA)-grafted MNPs taken in by the cells remained below 2 pg Fe/cell over the following 2 days. The very low uptake of P(PEGMA)grafted MNPs could be attributed to the fact that the P(PEGMA) grafting enhances the surface hydrophilicity of MNPs, which in turn will lower the adsorption of proteins

ARTICLE IN PRESS Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

5434

oleic acid-stabilized MNP CPA-stabilized MNP

Iron Concentration (pg/cell)

150

P(PEGMA)-grafted MNP

100

50

0 1

2

3

Growth Time (day)

Fig. 7. Uptake of oleic acid-stabilized MNPs (the data were reproduced from Ref. [25]), CPA-stabilized MNPs, and P(PEGMA)-grafted MNPs by RAW 264.7 cells as a function of growth time.

60

(a)

Magnetization (emu/g)

40

(b) (c)

20 0 -20 -40 -60 -9

-6

-3 0 3 Applied Field (KOe)

6

9

Fig. 8. Field dependent magnetization at 25 1C for (a) oleic acid-stabilized Fe3O4 nanoparticles, (b) Fe3O4 nanoparticles after ligand exchange with CPA, and (c) P(PEGMA)-grafted Fe3O4 nanoparticles by solvent-free ATRP.

from the culture medium and decrease the possibility of macrophage recognition. In view of the potential biomedical applications of the P(PEGMA)-grafted MNPs, the cytotoxicity of these MNPs was evaluated by determining the viability of 3T3 fibroblasts after incubation with medium containing the MNPs using the MTT assay. The viability of 3T3 fibroblasts remains high (95–105% as compared to the nontoxic control) after 24 h of incubation in the medium containing 0.2 mg/mL of the P(PEGMA)-grafted MNPs, indicating the cyto-compatibility of these MNPs. 3.7. Analysis of magnetic properties Fig. 8 shows the magnetization curves of as-synthesized Fe3O4 nanoparticles before and after ligand exchange, and

after P(PEGMA)-grafting as determined by VSM. The saturation magnetization (Ms) value of oleic acid-stabilized MNPs is 54 emu/g at 25 1C. After correcting for the oleic acid content (30%), the saturation magnetization value is 80 emu/g of Fe3O4, which is a little lower than the bulk magnetite value of 89 emu/g. After ligand exchange, there is little change in the Ms (50 emu/g), which is expected since the Fe3O4 content in the nanoparticles before and after ligand exchange is similar. After P(PEGMA) grafting, the Ms decreased to 28 emu/g due to the decrease in Fe3O4 content of the nanoparticles (31%). This level of Ms is deemed sufficient for bioapplications where Ms of 7– 22 emu/g is usually adopted [37–40]. No hysteresis curve was observed in Fig. 8 which indicates the characteristic superparamagnetic behavior of the nanoparticles at room temperature. It is noteworthy that an obvious ‘‘knee’’ appears earlier on the magnetization curve of P(PEGMA)grafted MNPs (Fig. 8(c)) than those of the MNPs before ATRP (Fig. 8(a) and (b)). Similar field-dependent magnetization results at room temperature after coating with nonmagnetic polymers have been reported previously [20,41]. It was considered that introduction of polymers with their own polarizability would modify inter-particle interactions and change the total magnetic anisotropy which is one of the most fundamental properties that determines the shape of the magnetization curve [41]. Since superparamagnetic Fe3O4 nanoparticles are good T2-type contrast agents in MRI and P(PEGMA) is a biocompatible and nonmagnetic macromolecule, it is interesting to know if P(PEGMA)-grafted Fe3O4 nanoparticles have a MR signal-enhancing property. A phantom test was carried out to investigate this. Fig. 9 shows T2weighted MR images of various concentrations of P(PEGMA)-grafted MNPs in water (spin-echo technique with TR ¼ 1600 and TE ¼ 80 ms). It can be seen that the MR signal intensity (related to the T2 relaxation time in T2weighted image) for the samples of different concentrations is not identical. With increasing nanoparticle concentration in water, the MR signal is enhanced significantly (negative in brightness in T2-weighted image). This result indicates that the nanoparticles can generate high magnetic field gradients near the surface of the P(PEGMA)-grafted MNPs. Fig. 10(a) and (b) show the inverse relaxation times 1/T1 and 1/T2 as a function of the iron molar concentration, [Fe], for the P(PEGMA)-grafted MNPs. In Fig. 10(a) and (b), the inverse relaxation times were found to vary linearly with the iron concentration, according to the following equations [42]: 1=T 1 ¼ 1=T 1 ð½Fe ¼ 0Þ þ R1 ½Fe,

(1)

1=T 2 ¼ 1=T 2 ð½Fe ¼ 0Þ þ R2 ½Fe,

(2)

where R1 and R2 are the longitudinal and transverse relaxivities, respectively. The intercepts 1/T1([Fe] ¼ 0) and 1/T2([Fe] ¼ 0) are the proton inverse relaxation times in pure water. The R1 value per millimole Fe of the

ARTICLE IN PRESS Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

5435

Fe Concentration (µM) 250

625

125

62.5

31.25

P(PEGMA)Grafted MNPs

100

10

80

8 (b)

60

6 4

40

1/T1 (s-1)

1/T2 (s-1)

Fig. 9. T2-weighted MRI images (1.5T, spin-echo sequence: repetition time TR ¼ 1600 ms, echo time TE ¼ 80 ms) of the P(PEGMA)-grafted Fe3O4 nanoparticles.

(a) 2

20 0 0.0

0.1

0.2

0.3 0.4 [Fe] (mmol/L)

0.5

0.6

0 0.7

Fig. 10. (a) T1 and (b) T2 relaxation rates (1/T1 and 1/T2, s1) as a function of iron concentration (mM) of P(PEGMA)-grafted Fe3O4 nanoparticles in aqueous solution (1.5 T, 25 1C).

P(PEGMA)-grafted MNPs at 8.870.8 mM1 s1 is smaller than those of conventional dextran-coated MNPs (20– 30 mM1 s1) [43]. In contrast, the R2 value per millimole Fe of the P(PEGMA)-grafted MNPs, which is 140.27 3.2 mM1 s1, is significantly larger than those for dextrancoated MNPs (30–50 mM1 s1) [43]. It is well known that the relaxivity ratio, R2/R1, is usually an important parameter to estimate the efficiency of T2-contrast agents. In our work, the R2/R1 is calculated to be 16, which is much larger than that of the dextrane-coated MNPs [43]. Furthermore, it is noteworthy that the T2, R2 and the particle size are very stable over a period of more than 6 h in the MR scanner, demonstrating the P(PEGMA)-grafted MNPs should perform well as T2-contrast agents in bioapplications. 4. Conclusion In summary, P(PEGMA)-grafted Fe3O4 nanoparticles were successfully synthesized via a relative simple and scalable approach, solvent-free ATRP method which is one of the ‘‘grafting from’’ methods. The macroinitiators on the surface of 6.470.8 nm Fe3O4 nanoparticles were introduced through effective ligand exchange of long alkane chain surfactants (oleic acid) by CPA, which rendered them soluble in the PEGMA monomer. After the solvent-free ATRP, monodispersed P(PEGMA)grafted Fe3O4 nanoparticles with a hydrodynamic particle

size of about 36 nm were obtained. These nanoparticles are superparamagnetic with Ms of 28 emu/g and possess good solubility and stability in water. Macrophage cells cultured in a medium containing 0.2 mg/mL of these nanoparticles exhibit similar morphology and viability as those of the cells cultured in the absence of nanoparticles. MTT assay of 3T3 fibroblasts cultured in the corresponding medium containing 0.2 mg/mL of these nanoparticles also confirmed the lack of significant cytotoxicity. The uptake of the MNPs by macrophage cells is reduced by two-orders of magnitude to o2 pg Fe/cell after surface grafting with P(PEGMA), indicating the good biocompatibility of the P(PEGMA)-grafted MNPs. MRI of the nanoparticles in water confirmed its contrast enhancement effect in T2weighted sequences. Compared with conventional dextrancoated MNPs, the P(PEGMA)-grafted nanoparticles exhibit relatively lower R1 value, higher R2 value and higher R2/R1 value, indicating that the P(PEGMA)-grafted nanoparticles may be a good candidate as T2-contrast agent. Furthermore, the preserved chlorine end groups and the pendant hydroxyl groups of the P(PEGMA) can serve as reactive sites for further biofunctionalization. These results indicate the vast potential future applications of the solvent-free ATRP method for preparing monodispersed iron oxide nanoparticles for MRI of specific biotargets. Acknowledgments This work was financially supported by the Singapore Bioimaging Consortium under SBIC Grant no. 023/2005. The authors gratefully acknowledge the assistance of Dr. Bao-Yu Zong and Mr. Fei-Xiong Hu, Department of Chemical and Biomolecular Engineering, National University of Singapore, in the VSM measurements and other experimental aspects. References [1] Mornet S, Vasseur S, Grasset F, Veverka P, Goglio G, Demourgues A, et al. Magnetic nanoparticle design for medical applications. Prog Solid State Chem 2006;34:237–47. [2] Thorek LJD, Chen AK, Czupryna J, Tsourkas A. Superparamagnetic iron oxide nanoparticle probes for molecular imaging. Ann Biomed Eng 2006;34:23–38. [3] Neuberger T, Scho¨pf B, Hofmann H, Hofmann M, Rechenberg BV. Superparamagnetic nanoparticles for biomedical applications: possibilities and limitations of a new drug delivery system. J Magn Magn Mater 2005;293:483–96.

ARTICLE IN PRESS 5436

Q.-L. Fan et al. / Biomaterials 28 (2007) 5426–5436

[4] Berry CC, Curtis ASG. Functionalisation of magnetic nanoparticles for applications in biomedicine. J Phys D Appl Phys 2003;36: R198–206. [5] Ito A, Shinkai M, Honda H, Kobayashi T. Functionalisation of magnetic nanoparticles for applications in biomedicine. J Biosci Bioeng 2005;100:1–11. [6] Gupta AK, Gupta M. Synthesis and surface engineering of iron oxide nanoparticles for biomedical applications. Biomaterials 2005;26: 3995–4021. [7] Anzai Y, Prince MR, Chenevert TL, Maki JH, Londy F, London M, et al. MR angiography with an ultrasmall superparamagnetic iron oxide blood pool agent. J Magn Reson Imaging 1997;7:75–81. [8] Molday RS. Magnetic iron-dextran microspheres. US patent 4452773 1984. [9] Petri-Fink A, Chastellain M, Juillerat-Jeanneret L, Ferrari A, Hofmann H. Development of functionalized superparamagnetic iron oxide nanoparticles for interaction with human cancer cells. Biomaterials 2005;26:2685–94. [10] Kim DK, Voit W, Zapka W, Bjelke B, Muhammed M, Rao KV. Biomedical application of ferrofluids containing magnetite nanoparticles. Mater Res Soc Symp Proc 2001;676. y8.32.31.31–36. [11] Yu S, Chow GM. Carboxyl group (–CO2H) functionalized ferrimagnetic iron oxide nanoparticles for potential bioapplications. J Mater Chem 2004;14:2781–6. [12] Lee SJ, Leong JR, Shin SC, Kim JC, Chang YH, Chang YM, et al. Nanoparticles of magnetic ferric oxides encapsulated with poly(D,L latide-co-glycolide) and their applications to magnetic resonance imaging contrast agent. J Magn Magn Mater 2004;272–276:2432–3. [13] Moffat BA, Reddy GR, McConville P, Hall DE, Chenevert TL, Kopelman RR, et al. A novel polyacrylamide magnetic nanoparticle contrast agent for molecular imaging using MRI. Mol Imaging 2003;2:324–32. [14] Gupta AK, Wells S. Surface-modified superparamagnetic nanoparticles for drug delivery: preparation, characterization, and cytotoxicity studies. IEEE Trans Nanobiosci 2004;3:66–73. [15] Saeed M, Wendland MF, Engelbrecht M, Sakuma H, Higgins CB. Value of blood pool contrast agents in magnetic resonance angiography of the pelvis and lower extremities. Eur Radiol 1998;8:1047–53. [16] Nitin N, LaConte LE, Zurkiya O, Hu X, Bao G. Functionalization and peptide-based delivery of magnetic nanoparticles as an intracellular MRI contrast agent. J Biol Inorg Chem 2004;9:706–12. [17] Caruso F. Nanoengineering of particle surfaces. Adv Mater 2001; 13:11–22. [18] Thu¨nemann AF, Schu¨tt D, Kaufner L, Pison U, Mo¨hwald H. Maghemite nanoparticles protectively coated with poly(ethyleneimine) and poly(ethylene oxide)-block-poly(glutamic acid). Langmuir 2006;22:2351–7. [19] Berret JF, Schonbeck N, Gazeau F, Kharrat DE, Sandre O, Vacher A, et al. Controlled clustering of superparamagnetic nanoparticles using block copolymers: design of new contrast agents for magnetic resonance imaging. J Am Chem Soc 2006;128:1755–61. [20] Vestal C, Zhang ZJ. Atom transfer radical polymerization synthesis and magnetic characterization of MnFe2O4/polystyrene core/shell nanoparticles. J Am Chem Soc 2002;124:14312–3. [21] Wang Y, Teng X, Wang JS, Yang H. Solvent-free atom transfer radical polymerization in the synthesis of Fe2O3@polystyrene core– shell nanoparticles. Nano Lett 2003;3:789–93. [22] Li G, Fan J, Jiang R, Gao Y. Cross-linking the linear polymeric chains in the ATRP synthesis of iron oxide/polystyrene core/shell nanoparticles. Chem Mater 2004;16:1835–7. [23] Gravano SM, Dumas R, Liu K, Patten TE. Methods for the surface functionalization of g-Fe2O3 nanoparticles with initiators for atom transfer radical polymerization and the formation of core–shell inorganic–polymer structures. J Polym Sci Part A: Polym Chem 2005;43:3675–88.

[24] Marutani E, Yamamoto S, Ninjbadgar T, Tsujii Y, Fukuda T, Takano M. Surface-initiated atom transfer radical polymerization of methyl methacrylate on magnetite nanoparticles. Polymer 2004;45: 2231–5. [25] Hu F, Neoh KG, Cen L, Kang ET. Cellular response to magnetic nanoparticles ‘‘PEGylated’’ via surface-initiated atom transfer radical polymerization. Biomacromolecules 2006;7:809–16. [26] Wuang SC, Neoh KG, Kang ET, Pack DW, Leckband DE. Heparinized magnetic nanoparticles: in-vitro assessment for biomedical applications. Adv Funct Mater 2006;16:1723–30. [27] Lattuada M, Hatton TA. Functionalization of monodisperse magnetic nanoparticles. Langmuir 2007;23:2158–68. [28] Gupta AK, Curtis ASG. Surface modified superparamagnetic nanoparticles for drug delivery: interaction studies with human fibroblasts in culture. J Mater Sci Mater Med 2004;15:493–6. [29] Moghimi SM, Hunter AC, Murray JC. Long-circulating and targetspecific nanoparticles: theory to practice. Pharmacol Rev 2001;53: 283–318. [30] Tiefenauer LX, Tschirky A, Kuhne G, Andres RY. In vivo evaluation of magnetite nanoparticles for use as a tumor contrast agent in MRI. Magn Reson Imaging 1996;14:391–402. [31] Zhang Y, Kohler N, Zhang M. Surface modification of superparamagnetic magnetite nanoparticles and their intracellular uptake. Biomaterials 2002;23:1553–61. [32] Kohler N, Fryxell GE, Zhang M. A bifunctional poly(ethylene glycol) silane immobilized on metallic oxide-based nanoparticles for conjugation with cell targeting agents. J Am Chem Soc 2004;126: 7206–11. [33] Kohler N, Sun C, Fichtenholtz A, Gunn J, Fang C, Zhang M. Methotrexate-immobilized poly(ethylene glycol) magnetic nanoparticles for MR imaging and drug delivery. Small 2006;2:785–92. [34] Lee H, Lee E, Kim DK, Jang NK, Jeong YY, Jon S. Antibiofouling polymer-coated superparamagnetic iron oxide nanoparticles as potential magnetic resonance contrast agents for in vivo cancer imaging. J Am Chem Soc 2006;128:7383–9. [35] Sun SH, Zeng H, Robinson DB, Raoux S, Rice PM, Wang SX, et al. Monodisperse MFe2O4 (M ¼ Fe, Co, Mn) nanoparticles. J Am Chem Soc 2004;126:273–9. [36] Matyjaszewski K, Xia J. Atom transfer radical polymerization. Chem Rev 2001;101:2921–90. [37] Lu HC, Yi GS, Zhao SY, Chen DP, Guo LH, Cheng J. Synthesis and characterization of multi-functional nanoparticles possessing magnetic, up-conversion fluorescence and bio-affinity properties. J Mater Chem 2004;14:1336–41. [38] Xu CJ, Xu KM, Gu HW, Zhong XF, Guo ZH, Zheng RK, et al. Nitrilotriacetic acid-modified magnetic nanoparticles as a general agent to bind histidine-tagged proteins. J Am Chem Soc 2004;126: 3392–3. [39] Tartaj P, Serna CJ. Synthesis of monodisperse superparamagnetic Fe/silica nanospherical composites. J Am Chem Soc 2003;125: 15754–5. [40] Brusentsov NA, Gogosov VV, Brusentsova TN, Sergeev AV, Jurchenko NY, Kuznetsov AA, et al. Evaluation of ferromagnetic fluids and suspensions for the site-specific radiofrequency-induced hyperthermia of MX11 sarcoma cells in vitro. J Magn Magn Mater 2001;225:113–7. [41] Srikanth H, Hajndl R, Chirinos C, Sanders J, Sampath A, Sudarshan TS. Magnetic studies of polymer-coated Fe nanoparticles synthesized by microwave plasma polymerization. Appl Phys Lett 2001;79: 3503–5. [42] Gillis P, Koenig SH. Transverse relaxation of solvent protons induced by magnetized spheres-application to ferritin, erythrocytes, and magnetite. Magn Reson Med 1987;5:323–45. [43] Wang YJ, Hussain SM, Krestin GP. Superparamagnetic iron oxide contrast agents: physicochemical characteristics and applications in MR imaging. Eur J Radiol 2001;11:2319–31.