Published: August 24, 2011

r 2011 American Chemical Society 12134 dx.doi.org/10.1021/la202096x | Langmuir 2011, 27, 12134–12142

ARTICLE

pubs.acs.org/Langmuir

“Two-in-One” Fabrication of Fe3O4/MePEG-PLA CompositeNanocapsules as a Potential Ultrasonic/MRI Dual Contrast AgentBin Xu,† Hongjing Dou,*,† Ke Tao,† Kang Sun,*,† Jing Ding,‡ Weibin Shi,‡ Xiasheng Guo,§ Jiyu Li,‡

Dong Zhang,§ and Kun Sun‡

†The State Key Lab ofMetal Matrix Composites, School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai,200240, P. R. China‡Department of General Surgery, Xinhua Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, 200092, P. R. China§Institute of Acoustics, Nanjing University, Nanjing, 210093, P. R. China

bS Supporting Information

’ INTRODUCTION

Moving from the macroscale to the nanoscale, superparamag-netic iron oxide (SPIO) nanoparticles demonstrate great poten-tial in various biomedical fields owing to their unique physical,chemical, and thermal properties.1�7 For biomedical applica-tions, a suitable surface coating for SPIO nanoparticles is usuallyrequired to avoid recognition by the mononuclear phagocyticsystem (MPS).1,4,7,8 Macromolecules such as polylactide (PLA)based copolymers, and polyethylene glycol (PEG) based copoly-mers are suitable for surface coating because they are known to bebiocompatible and versatile, while providing a platform for furtherbiological modification.9�13 Accordingly, coated SPIO compositenanoparticles can be classified as nanospheres or nanocapsulesbased on whether the macromolecules are absorbed onto thesurface of the SPIO nanoparticles or act as a vesicular matrix toencapsulate the SPIO nanoparticles, respectively.

Compared to nanospheres, nanocapsules are more useful forbiomedical imaging with ultrasonography due to their capabilityto further encapsulate gaseous bubbles to act as ultrasoniccontrast imaging agents (UCAs).13�17 UCAs enhance ultrasoniccontrast by altering the acoustical properties of the tissues, whichincludes improving, for example, back scattering, nonlinear har-monic, sound attenuation and phase velocity. Considering theT2-weighted MRI enhancement capability of SPIO nanoparticles,

SPIO enveloped composite nanocapsules should be of greatpotential for acting as ultrasonic/MRI dual contrast imagingagents, which is important for medical imageology because ultra-sonic imaging is an ideal complementary diagnostic tool to MRI’sreal-time temporal resolution.15

The extensive applications of SPIO based nanoparticles con-sequentially promoted the development of corresponding synth-esis techniques.18�21 As one of the most widely applied SPIOnanoparticles, Fe3O4 magnetic nanoparticles with hydrophilicsurface modification, are commonly synthesized by copre-cipitation in aqueous mediums.21�23 To improve the magneticproperties and size distribution of the resulting magneticnanoparticles, a facile interfacial coprecipitation to synthesizeFe3O4 nanoparticles was developed.

24,25 More recently, bycombining this interfacial coprecipitation with a double emul-sification (W1/O/W2) technique Fe3O4/PLA magnetic micro-capsules with sizes of 2�4 μm were successfully fabricatedusing a single step method, but the nanosized capsules werestill difficult to be prepared by using this method.26 Theexisting fabrication approaches for Fe3O4-based composite

Received: June 4, 2011Revised: August 18, 2011

ABSTRACT:A newmethod for the fabrication of Fe3O4 nanoparticles envelopedby polymeric nanocapsules is proposed. This method is characterized bycombining a double emulsification with the interfacial coprecipitation of ironsalts to form Fe3O4/polymer composite nanocapsules in a single step. Todemonstrate the viability of this approach, methoxy poly(ethylene glycol)-poly-(lactide) (MePLEG) was chosen as the shell material for Fe3O4/MePLEGnanocapsules. In addition to the versatility offered for fabricating nanocapsuleswith different shell materials, the method was found to be convenient for adjustingthe magnetite content of the nanocapsules from 0 to 43%. In addition to theirconfirmed T2-weighted magnetic resonance imaging (MRI) enhancement, theresultant composite nanocapsules display much more obvious acoustic responsesthan MePLEG nanocapsules in an acoustic investigation. Furthermore, the lowtoxicity of these composite nanocapsules, as confirmed by our study, combinedwith their magnetic and acoustic properties ensure that these composite nanocapsules have great potential in acting as ultrasonic/MRI dual contrast agents.

12135 dx.doi.org/10.1021/la202096x |Langmuir 2011, 27, 12134–12142

Langmuir ARTICLE

magnetic nanocapsules (MNCs) are usually time-consuming andinclude at least two steps, that is, the synthesis of Fe3O4 nano-particles and the incorporation of nanoparticles into a polymernanocapsule by, for example, an emulsion based technique.27�29 Anew approach that allows for the direct one-step synthesis ofMNCs with controllable properties is desirable.

For the fabrication of nanosized capsules by emulsification, theformation and stabilization of the nanoemulsion is critical andrequires careful design of the variables, such as the category ofshell materials, and emulsification methods.27 In the presentwork, MePEG-b-PLA (MePLEG) was chosen to form thebuilding blocks of the polymeric shell of the nanocapsulesbecause of the biocompatibility of both blocks and the efficientprotection PEG offers from the MPS.9,10,12 In addition, high-energy ultrasound has been confirmed to be effective with theW1/O/W2 emulsification methodology when fabricating MeP-LEG nanocapsules,27 and thus is introduced here to develop afacile “two-in-one” approach for producing Fe3O4/MePLEGMNCs. The “two-in-one” approach realizes the combination ofW1/O/W2 emulsification with the interfacial coprecipitationsynthesis of Fe3O4 nanoparticles, and enables the fabrication ofMNCs with a facile one step process. For this approach, both theFe3O4 nanoparticles loading content and the resultant saturationmagnetization of the MNCs can be adjusted conveniently byvarying the dose of the iron salts. Furthermore, the biocompat-ibility, acoustic properties, as well as T2-weighted MRI enhance-ment of MNCs are investigated. The results show that thesemagnetic nanocapsules have great potential for acting as ultra-sonic/MR dual imaging agents.

’EXPERIMENTAL SECTION

Materials. Ferric chloride hexahydrate (FeCl3 3 6H2O), ferrouschloride tetrahydrate (FeCl2 3 4H2O), di-n-propylamine and methylenedichloride (DCM) were purchased from Sinopharm Chemical ReagentCo., Ltd. Methoxy poly(ethylene glycol)-poly(lactic acid) (MePEG-PLA, Mw = 20 000, MePEG/PLA = 1:9, wt %) was purchased fromDaigang Biological Technology Co., Ltd. Poly(vinyl alcohol) (PVA,molecular weight 13 000�23 000 Da, alcoholysis degree 87�89%) waspurchased from Sigma-Aldrich, Inc. Sodium periodate (NaIO4) andhydrated ruthenium dioxide (RuO2 3 xH2O) were purchased fromAladdin chemical reagent Co., Ltd. All of the reagents were used asreceived.Fabrication of NC0, Fe3O4/MePEG-PLA Magnetic Nano-

capsules (MNCs) and Fe3O4 Nanoparticles (MNPs). TheFe3O4/MePEG-PLA magnetic nanocapsules were fabricated by a mod-ified “two-in-one”method. Briefly, FeCl3 3 6H2O and FeCl2 3 4H2Oweredissolved in 5 mL of deionized water (W1). 0.5 g of MePEG-PLA and

3mL of di-n-propylamine were dissolved in 20mL ofDCM(O). A seriesof Fe3+ aqueous solutionwith concentrations of 0.045mol/L, 0.15mol/L,0.37 mol/L, and 0.57 mol/L were used, and the resultant nanocapsuleswere named as MNC5, MNC15, MNC30, and MNC40, respectively.The molar ratio of Fe2+: Fe3+ was fixed to 1: 2. The aqueous solution ofiron salts was added dropwise into the DCM solution under ultrasonicvibration (300 W, continuous mode) in a nitrogen atmosphere, asshown in Scheme 1. The process of dropping was completed in about 5min, and ultrasonic oscillation continued for another 5 min to completethe reaction. The obtainedW1/O emulsion was then poured into 40 mLof 1 wt % PVA aqueous solution (W2). The obtained mixture wasultrasonicated for 1 min (180 W, continuous mode). The secondemulsification process was carried out in an ice�water bath. Theresultant W1/O/W2 double emulsion was poured into 100 mL of 2%(v/v) isopropanol aqueous solution and violently stirred at roomtemperature for 3 h to evaporate the DCM. The formed nanocapsuleswere separated by centrifugation (4 �C, 10 000 rpm, 30 min). Theprecipitate was resuspended in deionized water, the solution wascentrifuged and this process was repeated once more. Thereafter, thenanocapsules were purified by magnetic separation (the procedure isdescribed in the Supporting Information). The precipitate was thenresuspended in 25 mL of deionized water for an acoustic test. The drypowder sample of the nanocapsules was collected by lyophilization inorder to analyze its other characteristics.

For the preparation ofNC0 in whichwithout Fe3O4 encapsulated, theprocess is similar to that for fabricating MNCs, except that neither ironsalts nor di-n-propylamine were added during the fabrication. After thefabrication, The formed nanocapsules were separated by centrifugation(4 �C, 10 000 rpm, 30 min).

The Fe3O4 nanoparticles enveloped in MNCs, named MNP5,MNP15, MNP30, and MNP40, were separated from the MNCs bydissolving the composite nanocapsules in chloroform, a good solvent forMePLEG. The samples were collected by a magnetic separation.

A sample of Fe3O4 nanoparticles was fabricated by an interfacialcoprecipitation progressing to the interface between W1 and O phasewithout the addition of MePLEG and is termed as MNP. FeCl3 3 6H2Oand FeCl2 3 4H2O were dissolved in 5 mL of deionized water (W1), and3 mL of di-n-propylamine were dissolved in 20 mL of DCM (O). Theconcentration of Fe3+ aqueous solution was 0.045 mol/L and the molarratio of Fe2+/Fe3+ was fixed to 1:2. The aqueous solution of iron saltswas added dropwise into the DCM solution under ultrasonic vibration(300 W, continuous mode) in a nitrogen atmosphere. The process ofdropping was completed in about 5 min, and ultrasonic oscillationcontinued for another 5 min to complete the reaction. Thereafter, theMNP nanoparticles were purified by magnetic separation.Size Characterization of Nanocapsules. The particle size and

size distributions (mean diameters and PDI) were determined byphoton correlation spectroscopy (PCS) using a particle size analyzer(Zetasizer Nano ZS 90, Malvern Instruments).

Scheme 1. Schematic Process for the Fabrication of Fe3O4/MePEG-PLA Nanocapsules

12136 dx.doi.org/10.1021/la202096x |Langmuir 2011, 27, 12134–12142

Langmuir ARTICLE

Transmission Electron Microscopy Observation. Transmis-sion electron microscopy (TEM, 200 kV, JEM-2010, JEOL, Japan) wasused to observe the morphology of the nanocapsules. For TEMobservation, an aqueous solution of the nanocapsules was sprayed ontoa carbon film coated copper grid device, and after evaporating the water,the sample was stained by a 1 wt % ruthenium tetroxide (RuO4) aqueoussolution. The RuO4 aqueous solution was prepared freshly according tothe literature.30 Briefly, NaIO4 (0.128 g) was dissolved in 10 mL ofdeionized water at room temperature. RuO2 3 xH2O (0.06 g) was thenadded to the aqueous solution of NaIO4. RuO4 was synthesized whenRuO2 3 xH2O began to dissolve, and the unreacted RuO2 3 xH2O settledto the bottom of the bottle.X-ray Diffraction Analysis. Crystallographic analysis was per-

formed by an X-ray diffraction system (XRD, D/max-2200/PC, RigakuCorporation, Japan) with Cu radiation to identify the dominant phase ofthe samples. The 2-theta angle range of the measurements was from10�70�. The phase was determined by using standard powder diffrac-tion files from the Joint Committee for Powder Diffraction Studies(JCPDS).

The crystallite size is calculated from the full width at half-maximum(fwhm) of the diffraction peaks of the samples using the Scherrerequation

τ ¼ kλβ cosðθÞ ð1Þ

Relative crystallinity was determined from the integral intensity of thediffraction peaks of the samples.Thermogravimetric Analysis. The thermal behavior of the

nanocapsules was measured by a thermogravimetry analyzer (TGA,Q500, TA Instruments, USA). Samples were heated from roomtemperature to 600 �C at a rate of 10 �C/min. The measurements werecarried out under a nitrogen atmosphere.Magnetization Measurements. The magnetization properties

of the Fe3O4/MePEG-PLA nanocapsules at 300 Kwere studied by usinga vibrating sample magnetometer (VSM, Model 7407, Lake ShoreCryotronics Inc., USA). Saturation magnetization, coercive force andremnant magnetization were obtained from the hysteresis loops.In Vitro Acoustic Experiments. Degassed water and hybrid

nanocapsules (5 mg/mL) were imaged using the ultrasonic imagingsystem of a GE LOGIQ Book XP Enhanced scanner (GE MedicalSystems, USA) where a 4 MHz or an 11 MHz ultrasound transduceracted as a transmitter as well as a receiver. All images were acquired withthe same instrument parameters (mechanical index (MI) = 0.5). Theschematic diagram of the measurement was shown in Scheme S1 inSupporting Information, that is, SI.

An ultrasound spectroscopy method31 was used to measure theacoustic attenuation spectrum. As shown in Scheme 2, a broadbandacoustic pulse was excited and the signals prior to and during insertion of

the sample were detected and analyzed by FFT software. The soundattenuation of a sample (αs) can be expressed as

as ¼ aw1d lnjPWðωÞTðωÞj

jPSðωÞj� �

ð2Þ

where αw is the sound attenuation in water; Pw(ω) and Ps(ω) are theamplitude spectra before and after the insertion of specimen respec-tively;T(ω) is the spectrum of the overall transmission coefficient; and dis the thickness of the specimen.In Vitro MR Imaging Experiments. The relaxivity of magnetic

resonance imaging was obtained by a Siemens 3.0 T scanner (MagnetomTrio, Siemens, Munich, Germany) with a wrist coil. Phantom MRI wascarried out at various iron concentrations of MNCs from 0 to 0.4 mM inan agarose gel. The spin echo sequence was used. The imagingparameters were: repetition time (TR) 3000 ms; field of view (FOV)106 � 180 (mm � mm); matrix size 576 � 342 (mm � mm); slicethickness 5.0mm. Then, the resulting change in the transverse relaxationtime (T2) of the nanocapsules suspension were continuously measuredby recording the above-mentioned single-slice gradient-echo signal. Nophase or frequency encoding was used. According to the monoexpo-nential signal decay as the function of echo time (TE), the transverserelaxation time of well-mixed nanocapsules suspension can be estimated.Cytotoxicity Assays. To investigate the cytotoxicity of the mag-

netic nanocapsules, an in vitro experiment was performed using rat livercell BRL-3A and rat kidney cell NRK cultures. The number of viable cellswas determined by the estimation of their mitochondrial reductaseactivity using the tetrazolium-based colorimetric method (MTTmethod)and two-color flow cytometry (FCM method).

A MTT assay depends on the cell’s reductive capacity to metabolizethe yellow tetrazolium salt into a highly colored formazan product.

BRL-3A cells in the log phase of growth were seeded in RPMI-1640culture media with 100 U/mL penicillin, 100 U/mL streptomycin, and10% fetal bovine serum (FBS) at 10 000 cells/mL in an incubator at37 �C for 12 h. NRK cells in the log phase of growth were seeded in0.1 mL DMEM/high glucose media containing 100 U/mL penicillin,100U/mL streptomycin, and 10% heat-inactivated FBS at 10 000 cells/mLin an incubator at 37 �C for 12 h. The cytotoxicity of the MNCs wasevaluated by determining the viability of the cells after coincubation withdifferent concentrations of MNCs (from 0.050 to 1 mg/mL) with 5%CO2 at 37 �C for 24 h. At the end of the incubation period with theMNCs, cells were incubated with a 200 μL sample of 10% 3-(4,5-dime-thylthiazol-2-yl)-2,5-diphenyltetrazolium bromide dye (MTT) solutionat 37 �C for 4 h. Afterward, 100 μL of DMSO was added in order todissolve the formazan crystals. The UV absorbance of the solubilizedformazan crystals was measured at 492 and 630 nm. Cell viability wasexpressed as the ratio between the amount of formazan determined forcells treated with the different MNCs suspensions and for controlnontreated cells. Each point was performed five times so that thestandard deviations were calculated.

For the FCM assay, NRK cells in the log phase of growth were seededin 0.1 mL DMEM/high glucose media containing 100 U/mL penicillin,100 U/mL streptomycin, and 10% heat-inactivated FBS. The cells weremaintained in at 10 000 cells/mL in an incubator at 37 �C for 12 h. Thencells were coincubated with MNCs in 5% CO2 at 37 �C for 24 h. Theconcentration of theMNCswas controlledwith concentrations of 25, 100,and 400μg/mLused. Cells were stainedwith Annexin-V (25μg/mL) andPI (50 μg/mL) and analyzed with a flow cytometer (Cytomics FC500,Beckman Counter Inc.); 10 000 events were analyzed for each sample.

’RESULTS AND DISCUSSION

Fabrication and Composition of Composite Nanocap-sules. Interfacial coprecipitation, in which the coprecipitationreaction is confined only to the interface between the water and

Scheme 2. Schematic Diagram of the Device for SoundAttenuation Spectrum Measurement

12137 dx.doi.org/10.1021/la202096x |Langmuir 2011, 27, 12134–12142

Langmuir ARTICLE

the oil phase, has been confirmed to be an effective approach tocreate hydrophilic magnetite nanoparticles capped with aminegroups.24,25 A previous study validated the efficacy of high-energyultrasound in the fabrication of MePLEG nanocapsules byW1/O/W2 emulsification.

32 Therefore, according to the design,the combination of interfacial coprecipitation into W1/O/W2emulsification would ideally realize a “two-in-one” fabrication ofencapsulated iron oxide nanoparticles, or nanocapsules, as de-scribed in Scheme 3, and provide a methodology to simplify thefabrication of composite magnetic nanocapsules (MNCs). Thecoprecipitation was accomplished during the first emulsificationand occurred only at the interface between the inner water phase(W1) and the outer MePLEG methylene dichloride solution.To adjust the iron oxide content inside the MNCs so that themagnetic property of resultant MNCs could be controlled theratio of ferric salt/MePLEG was varied and the resultant MNCswere named NC0, MNC5, MNC15, MNC30, and MNC40,respectively.The properties of iron oxide nanoparticles relate closely to

their composition and crystalline structure.33 To elucidate thecomposition of the iron oxide nanoparticles inside the MNCs,the nanoparticles named MNP were synthesized by a W1/O emul-sification in which pure methylene dichloride withoutMePLEGwas

used as oil phase. The diameter of the MNPs was determined byTEM observation as around 10 nm (as shown in the Figure S1 ofSI). In addition, as shown in Figure 1, the XRD patterns of MNPand MNC5 confirm their structure as Fe3O4 because the posi-tions and relative intensities of the main peaks match well tothose from the JCPDS card (19�0629) for Fe3O4.34 Moreover,compared with that of MNP, the peaks of the XRD pattern forMNC5 are unobvious and broad, which might be because themain composition of the MNC are a MePLEG polymer matrix.The samples fabricated by this “two-in-one” approach were

characterized by various measurements for the purpose ofvalidating the methodology. Shown in Figure 2 are the TGAcurves of five samples prepared with different ferric salt/MeP-LEG ratios. The thermograms of the composite magneticnanocapsules (MNC5, MNC15, MNC30, and MNC40) exhibitdifferent decomposition stages when compared with nanocap-sules without Fe3O4 nanoparticles (NC0), the decompositionstage around 300�400 �C in the thermograms of MNCs may beattributed to the interaction between the polymer and the Fe3O4nanoparticles which postpones the decomposition of thepolymer.35 The residue above 600 �C is ascribed to the inorganiccontent of the MNCs and thus, the weight percentages of theinorganic iron oxide nanoparticles in the MNCs were calculatedfrom the weight residue of the composite MNCs at 600 �C.The inorganic content for a series of samples are listed in

Table 1, for which the theoretical content was calculated from theiron salt dosed during coprecipitation and is compared with theFe3O4 content calculated from the residue weight by the TGAcurves shown in Figure 2. It is worth noting that during theinterfacial coprecipitation and double emulsification, inevitably asmall quantity of MePLEG solid nanoparticles are formed with-out Fe3O4 inside because pf the possible escape of inner water

Scheme 3. Mechanism of the Fabrication of Fe3O4/MePEG-PLA Nanocapsules

Figure 1. X-ray diffractogram patterns of MNC5 and MNP separatedfrom the first emulsification. The inset is the selected area electrondiffraction pattern of MNC5.

Figure 2. Thermogravimetric curves of Fe3O4/MePEG-PLA nanocap-sules with different Fe3O4 contents (wt %) (samples are all measuredafter magnetic separation).

Table 1. Fe3O4 Content in Fe3O4/MePLEG MNCs

sample

theoretical

Fe3O4 content

Fe3O4 content (before

separation) (%)

Fe3O4 content (after

separation) (%)

NC0 0.0%

MNC5 5.0% 4.3 10.8

MNC15 15.0% 10.4 22.1

MNC30 30.0% 18.6 38.7

MNC40 40.0% 27.9 43.5

12138 dx.doi.org/10.1021/la202096x |Langmuir 2011, 27, 12134–12142

Langmuir ARTICLE

phase from the oil phase during the second emulsification.27 Toremove these impurities from the products, the crude productswere all purified by magnetic separation as specified in theExperimental Section, and the residue after magnetic separationwas confirmed to contain all organic substances which degradecompletely before 600 �C (Figure S2 in SI). The Fe3O4 contentof the crude products and their TGA results were also examined(Figure S3 in SI). As displayed by Table 1, the Fe3O4 content inthe crude products are all lower than the theoretical Fe3O4content, and the disparity from the theoretical value increases athigher iron salt doses. Considering that both the concentration ofdi-n-propylamine in methylene dichloride and the W1/O inter-face area were kept constant with increasing doses of iron salt, theexistence of a greater disparity at higher iron salt doses isreasonable because of the inefficient reaction between excessiron salts and hydroxyl ions. Additionally, the real Fe3O4 contentafter magnetic separation calculated from the TGA results werelisted in Table 1. After removing the MePLEG solid nanoparti-cles from the crude products, the Fe3O4 content increased due tothe decrease of the total content of MePLEG.Morphology and Size Distributions of Composite Nano-

capsules. The morphology of composite MNCs were observedby TEM, as shown in Figure 3a, which is a typical TEM image forthe magnification of one capsule fromMNC5. As is presented inFigure 3a, the MNCs possess a spherical shape and the Fe3O4nanoparticles are encapsulated mostly inside the internal portionof the nanocapsules. The light gray halo around the MNCs isspeculated to be the PEG corona on the periphery of thenanocapsules and the PVA emulsifier adsorbed on the surfaceof the nanocapsules, both of which can be stained by RuO4.

36

Dynamic light scattering measurement is a versatile approachfor determining the average hydrodynamic diameter (ÆDhæ) ofnanosized colloidal particles. As shown in Figure 3(b), the ÆDhævalues of composite MNCs are higher than that of NC0, whichillustrates that the encapsulation of Fe3O4 nanoparticles resultsin the increase of the diameter of the nanocapsules. There are twopossible reasons for this result: the addition of di-n-propylamineincreases the interface energy of the O/W2 interface during thesecond emulsification or, the encapsulation of Fe3O4 nanoparticles

in the inner water phase results in the increase of the size of theW1nanodroplets, and therefore increases the size of the oil dropletsduring the second emulsification. In addition, the ÆDhæ of MNC5from DLS measurements are slightly larger than the 50�200 nmdiameters observed from the TEM image, whichmight be becauseof the shrinkage of the capsules during the preparation of the TEMsamples.37�39

Magnetic Properties and In Vitro MR Imaging of Compo-site Nanocapsules. Shown in Figure 4a are the room tempera-ture hysteresis loops of MNCs fabricated by different doses ofiron salts, which displays almost immeasurable coercivity andremanence, suggesting that individual magnetic nanoparticlesinside MNCs are of a single domain. Since the saturationmagnetization of MNCs increase with increasing Fe3O4 content,this indicates that the magnetic properties of MNCs could beadjusted by controlling the reagent ratio during the first emulsi-fication, a conveniently realized methodology. The inset ofFigure 4a demonstrates that MNC5, which possess the lowestsaturation magnetization of the MNCs, still displays magneticresponsibility. Additionally, to explore the effects of polymeraddition on the magnetic properties of the resultant magneticnanoparticles inside theMNCs, the room temperature hysteresisloop ofMNP nanoparticles was also measured and is displayed inFigure 4a. The saturation magnetization of the MNP synthesizedby a W1/O interfacial coprecipitation was found to be as high as78 emu/g, a value higher than that of magnetite nanoparticlessynthesized by both aqueous coprecipitation21�23 and interfacialcoprecipitation proceeding under stirring.25 The higher satura-tion magnetization of MNP can be attributed to the ultrasonicprocess used here. We suggest that the introduction of ultra-sonics in the homogeneous process increased the area of theW1/O interface, and thus improved the opportunity for thereaction between the iron salts and the hydroxide ions. With thismethodology, an ultrasonic process might play a crucial role inimproving the magnetic properties of resultant magnetitenanoparticles.A previous study disclosed that in aqueous coprecipitation, the

addition of a polymer into the medium will affect the crystallinestructure and therefore the magnetic properties of the resultant

Figure 3. (a) TEM image of MNC5 stained by RuO4. The inset is the high-magnification TEM image of MNC5. (b) Size distribution of Fe3O4/MePEG-PLA nanocapsules with different Fe3O4 contents (wt %).

12139 dx.doi.org/10.1021/la202096x |Langmuir 2011, 27, 12134–12142

Langmuir ARTICLE

magnetite nanoparticles.40 For the purpose of elucidating thispoint, the saturation magnetization of the MNCs and MNPs(the Fe3O4 nanoparticles inside the MNCs after removing thepolymer) are included in Figure 4b. Utilizing themagnetite weightcontent of the MNCs (WM%) determined by TGA and thesaturation magnetization of the MNP (SMMNP), the theoreticalsaturation magnetization of MNCs (TSMMNCs) were calculatedaccording to eq 3. Notably, the calculation of TSMMNCs by eq 3is based on the hypothesis that the properties of the magne-tite nanoparticles inside the MNCs are the same as that ofMNP. From Figure 4b, the saturation magnetization of MNCs(ASMMNCs) is lower than the corresponding TSMMNCs, whichindicates that the addition of MePLEG into the oil phase duringthe W1/O interface coprecipitation affected the crystalline struc-ture, and thus, the magnetic properties of the resultant magnetitenanoparticles.

TSMMNCs ¼ SMMNP � ðWM%Þ ð3ÞThe XRD measurements were then carried out to investigate

the Fe3O4 crystallization, and the resultant XRD patterns of theMNPs are shown in Figure 5. By using the Scherrer equation(eq 1), τ = kλ/(β cosθ) the crystallite size of the MNPs werecalculated from the fwhm of (311), (440), and (220) reflections.Listed in Table 2 are the crystallite size of the MNPs, the relative

crystallinity of the MNPs relative to the MNP, and the saturationmagnetization of the MNPs (ASMMNPs) calculated according toeq 4. The ASMMNPs are all lower than the 78 emu/g of ASMMNPsuggesting that the crystallinity of the MNPs inside MNCs arenot as high as that of MNP because there is a correlation betweensaturation magnetization and crystallinity.41 As listed in Table 2,the lower relative crystallinity of MNPs is also verified by thecalculation from the XRD patterns.

ASMMNPs ¼ ASMMNCs=ðWM%Þ ð4Þ

To evaluate the T2 enhancing capability, agarose gel withvarious concentrations of MNCs were investigated by the T2-weighted MRI. As shown in Figure 6, the signal intensity of MRIdecreased with the increase of the concentration of MNCs.Figure 6b indicated that the T2 relaxation time decreased asthe concentrations of iron increasing and the trend is well fit by alinear line within the analyzed range of iron concentrations,exhibiting the typical properties of Fe3O4 nanoparticles withshortening T2 relaxation time. For example, the T2 relaxationtime of MNC5 decreased from 73.3 to 17.7 ms as the concentra-tions of iron increasing from 0.05 to 0.5 mM. Specifically, therelaxivity, that is, the slope of T2

�1 versus Fe concentration, ofMNC5 was calculated to be 107.09 mM�1 s�1, which is higherthan that of MNP (85.05 mM�1 s�1). This may due to theenhanced susceptibility effect of assembled MNP inside thenanocapsules, for the case of MNC, the number of MNP pervolume in the inner cavity is much higher than that for the freeMNP case.15

Figure 4. (a) Magnetization curves of Fe3O4/MePEG-PLA nanocapsules with different Fe3O4 contents and pure MNPs separated from the firstemulsification. The inset shows the magnetic responsibility of MNC5 dispersed in water. (b) The theoretical saturation magnetization (TSMMNCs) andactual saturation magnetization (ASMMNCs) of Fe3O4/MePEG-PLA magnetic nanocapsules (MNCs) with different Fe3O4 contents.

Figure 5. X-ray diffractogram patterns of Fe3O4 nanoparticles inFe3O4/MePEG-PLA nanocapsules with different Fe3O4 contents.

Table 2. Crystalline Size of Fe3O4 in Fe3O4/MePEG-PLANanocapsules with Different Fe3O4 Contents and Its Corre-sponding Saturation Magnetization

sample size (d 3m)saturation magnetizationa

(emu/g)

relative crystallinity

(%)

MNP5 14.5 25.9 61.4

MNP15 15.5 44.0 66.1

MNP30 17.2 33.2 89.5

MNP40 16.8 36.4 64.3aThe saturation magnetization of magnetite nanoparticles were calcu-lated from the saturation magnetization ofMNCs as shown in Figure 4b.

12140 dx.doi.org/10.1021/la202096x |Langmuir 2011, 27, 12134–12142

Langmuir ARTICLE

Acoustic Properties of Composite Nanocapsules. Ultra-sonic imaging from sound attenuation is a new technique, whichis reported to be effective in helping tissue characterization anddiagnosis of early nonoccupancy canceration.42 Sound attenua-tion may be caused by various acoustical responses, including thescattering and absorbing of ultrasonic waves produced as itcrosses an inhomogeneous medium. In addition, on the condi-tion of the incident frequency being consistent with the naturalresonance frequency of the ultrasonic contrast imaging agents(UCAs), all of the incident ultrasound energy would be absorbedby the UCAs. As a result, the sound attenuation would reach apeak value and the corresponding video intensity of the ultra-sound contrasted image would be enhanced.43 Recent reportsconcluded that the sound attenuation of soft microbubbleUCAs are directly proportional to their concentration, and the

resonance frequency is inversely proportional to their particlesize.44 However, in the present case, the comparatively harderMePLEG shell, the nanometer sized capsules, as well as the Fe3O4nanoparticles inside the nanocapsules may result in entirelydifferent acoustical responses from those previously reported.To investigate the contrast efficiency of MNCs, in vitro

ultrasonography was performed with an emission frequency of4 and 11 MHz, two commonly used frequency in medicaldiagnostics. As shown in Figure 7, the video intensity is enhancedat both frequencies in the presence of MNCs when compared tothat of degassed normal saline and NC0, and the intensityshowed a tiny increase with increasing Fe3O4 content. Accordingto the mechanism of 2D imaging, the video intensity is directlyproportional to the strength of sound attenuation.45 Thereforethe enhancement of the video intensity in in vitro ultrasonogra-phy results from the higher sound attenuation provided byMNCs. The enhancement of the acoustic response resultingfrom the encapsulation of Fe3O4 nanoparticles may provide anew approach to improve the enhancement of UCAs in ultra-sound imaging.For investigating the effect of Fe3O4 nanoparticles on the

acoustical properties of MNCs, the MNCs are water-filled toexclude the effect of gas so that the variation of acousticalcharacteristics of MNCs depends only on the Fe3O4 content.It was reported that the Fe3O4 nanoparticles embedded in thepolymeric shell of microcapsules can enhance the contrast ofultrasound signals and that this may be due to the additionalacoustic impedance provided by Fe3O4 nanoparticles duringultrasound imaging.15,46,47 The unique characteristics of thepresent case are that the Fe3O4 nanoparticles are located in theinternal cavity instead of the shell and that the core is filled withwater instead of gas. As disclosed by the sound attenuationspectra for NC0 and MNCs shown in the Figure S5 of SI, thesound attenuation is actually enhanced by the presence of Fe3O4nanoparticles embedded inside the nanocapsules. In addition,with increasing Fe3O4 content, the acoustic attenuation increasesaccordingly and the resonance frequency shifts to a higher value.Cytotoxicity of Composite Nanocapsules. For biomedical

materials, low cell cytotoxicity is a prerequisite for their applica-tion. Although poly(lactic acid)-based polymers are generallybiocompatible,48 the cytotoxicity of Fe3O4/MePEG-PLA com-posite MNCs nanocapsules were investigated by using variousapproaches. A classical MTT assay was performed for twometabolic cell cultures, NRK and BRL-3A cells, to evaluate thenanocapsules’ cytotoxicity. The cell viability obtained by theMTT assay was expressed as a fraction of viable cells andnormalized to that of cells incubated without MNCs (blank

Figure 6. (a) T2-weighted MR images of MNC5, MNC15, MNC30,MNC40, and MNP. (b) T2 relaxation rate (1/T2) as a function of Feconcentration (mM), TE is 60 ms. (MNC5,9; MNC15,b; MNC30,2;MNC40, 1; MNP, [).

Figure 7. In vitro ultrasonography of degassed normal saline, NC0, MNC5,MNC15, MNC30, andMNC40 with emission frequency of 4 and 11MHz.The concentration of samples are all 5 mg/mL.

12141 dx.doi.org/10.1021/la202096x |Langmuir 2011, 27, 12134–12142

Langmuir ARTICLE

control). The results are shown in Figure 8. For NRK samples, asshown in Figure 8a, the viability of the cells incubated with bothNC0 and MNC5 are usually close to 100% and even higher than100% at higher nanocapsule concentrations, indicating that theNC0 nanocapsules show almost no cytotoxicity, and can evenstimulate cell proliferation. As for BRL-3A cells shown inFigure 8b, the results are similar to that for NRK cells, whichindicates that Fe3O4/MePEG-PLA nanocapsules impart littlecytotoxic effects to metabolic cells.A flow cytometry fluorescent-activated cell sorting technique

(FACS) utilizes a laser beam that differentiates cells based on theirsize and density to determine the genotoxic potential of MNCs byexamining the extent ofDNAdamage. By usingDNA intercalatingdyes, the cellular DNA content can be used to determine theproportion of cells undergoing apoptosis.49 The cytotoxicity ofMNCs at various concentrations on NRK cells was evaluated byusing FACS to detect their apoptosis via a fluorescein annexin-V-FITC/PI double labeling. The rates of apoptosis in the early stage(as defined in the SI) ofNRKcells coincubatedwith 0, 25, 100, and400 μg/mL MNCs nanocapsules are presented in Figure 9 (theoriginal flow cytometry results for NRK cells are displayed in theSI), the results of which are a direct indicator of the cytotoxicity ofMNCs.49 When compared with the control experiment (withoutcoincubation with MNCs), MNCs did not show an effect on the

apoptosis of NRK cells, with the rate of apoptosis being lower than2%. Consistent with MTT results, the results of FACS shows thatboth NC0 and MNC5 have no obvious cytotoxic effects to NRKcells at the three concentrations examined.

’CONCLUSION

A novel “two-in-one” approach was developed to fabricateFe3O4/MePEG-PLA nanocapsules with magnetic/ultrasonicdual responses; this approach is characterized by the combina-tion of a double emulsification coupled with an interfacialcoprecipitation. Specifically, magnetite nanoparticles weresynthesized by the interfacial coprecipitation that proceededon the W1/O interface during the first emulsification and theresultant magnetite nanoparticles with hydrophilic surfaces thusentered the W1 phase after their formation. The Fe3O4 nanopar-ticles encapsulated composite nanocapsules were fabricated uponthe progress of the second emulsification and the evaporation of oilphase. The sizes, structures, aswell as themagnetite contents of theresultant MNCs were investigated by various methods; it is foundthat the magnetite content of MNCs can be adjusted convenientlyfrom 0% to 43% by controlling the dosage of the iron salts. Inaddition, the MNCs were confirmed to exhibit magnetic/ultrasonic dual responses and low cytotoxicity. These character-istics enable the MNCs to be used as a potential biomedicalmaterial for enhancing the contrast of MR/ultrasonic images.

’ASSOCIATED CONTENT

bS Supporting Information. Complementary TGA, acous-tic attenuation, and FCM results of the composite nanocapsules.This information is available free of charge via the Internet athttp://pubs.acs.org

’AUTHOR INFORMATION

Corresponding Author*Tel.: +86-21-34202743. Fax: +86-21-34202745. E-mail: hjdou@sjtu.edu.cn (H.D.); ksun@sjtu.edu.cn (K.S.).

’ACKNOWLEDGMENT

This work was financially supported by the National Nat-ural Science Foundation of China (No. 20904032, 50902093,

Figure 8. Cytotoxicity profiles of (a) NRK and (b) BRL-3A cell, when incubated with various concentrations of Fe3O4/MePEG-PLA nanocapsules asdetermined by MTT assay. Data represent mean ( SD, n = 5.

Figure 9. Apoptosis of renal cell NRK coincubated with Fe3O4/MePEG-PLA nanocapsules at various concentrations according to theresults from FCM.

12142 dx.doi.org/10.1021/la202096x |Langmuir 2011, 27, 12134–12142

Langmuir ARTICLE

10974093), Science and Technology Committee of Shanghai(Project No. 05XD14015) and Shanghai Education Committee(Project No. 09YZ103), State Key Lab of Metal Matrix Compo-sites and the Shanghai Jiao Tong University Foundation ofMedicine-Engineering Cross-Disciplinary Research (ProjectNo. YG2009ZD202). We thank Instrumental Analysis Centerof SJTU for the assistance on measurements. We also thankShanghai Sunny New Technology Development Co. Ltd. fortheir support.

’REFERENCES

(1) Pankhurst, Q. A.; Connolly, J.; Jones, S. K.; Dobson, J. J. Phys. D:Appl. Phys. 2003, 36 (13), R167–R181.(2) Bonnemain, B. J. Drug Targeting 1998, 6 (3), 167–174.(3) Jun, Y. W.; Seo, J. W.; Cheon, A. Acc. Chem. Res. 2008, 41 (2),

179–189.(4) Sun, C.; Lee, J. S. H.; Zhang, M. Q. Adv. Drug Delivery Rev. 2008,

60 (11), 1252–1265.(5) Casula, M. F.; Floris, P.; Innocenti, C.; Lascialfari, A.; Marinone,

M.; Corti, M.; Sperling, R. A.; Parak, W. J.; Sangregorio, C. Chem. Mater.2010, 22 (5), 1739–1748.(6) Kedziorek, D. A.; Kraitchman, D. L. Superparamagnetic Iron

Oxide Labeling of Stem Cells for MRI Tracking and Delivery inCardiovascular Disease. In Stem Cells for Myocardial Regeneration:Methods and Protocols; Lee, R. J., Ed.; Humana Press Inc: Baltimore,MD, 2010; pp 171�183.(7) Liu, G.; Wang, Z. Y.; Lu, J.; Xia, C. C.; Gao, F. B.; Gong, Q. Y.;

Song, B.; Zhao, X. N.; Shuai, X. T.; Chen, X. Y.; Ai, H.; Gu, Z. W.Biomaterials 2011, 32 (2), 528–537.(8) Lee, H.; Lee, E.; Kim,D. K.; Jang, N. K.; Jeong, Y. Y.; Jon, S. J. Am.

Chem. Soc. 2006, 128 (22), 7383–7389.(9) Cole, A. J.; David, A. E.; Wang, J.; Galban, C. J.; Hill, H. L.; Yang,

V. C. Biomaterials 2011, 32 (8), 2183–2193.(10) Sun, C.; Sze, R.; Zhang, M. Q. J. Biomed. Mater. Res., Part A

2006, 78A (3), 550–557.(11) Kohler, N.; Fryxell, G. E.; Zhang, M. Q. J. Am. Chem. Soc. 2004,

126 (23), 7206–7211.(12) Vila, A.; Gill, H.; McCallion, O.; Alonso, M. J. J. Controlled

Release 2004, 98 (2), 231–244.(13) Mishra, B.; Patel, B. B.; Tiwari, S. Nanomed.: Nanotechnol., Biol.

Med. 2010, 6 (1), e9–e24.(14) Bae, K. H.; Ha, Y. J.; Kim, C.; Lee, K. R.; Park, T. G. J. Biomater.

Sci., Polym. Ed. 2008, 19 (12), 1571–1583.(15) Yang, F.; Li, Y. X.; Chen, Z. P.; Zhang, Y.; Wu, J. R.; Gu, N.

Biomaterials 2009, 30 (23�24), 3882–3890.(16) Pisani, E.; Tsapis, N.; Galaz, B.; Santin, M.; Berti, R.; Taulier,

N.; Kurtisovski, E.; Lucidarme, O.; Ourevitch, M.; Doan, B. T.; Beloeil,J. C.; Gillet, B.; Urbach, W.; Bridal, S. L.; Fattal, E. Adv. Funct. Mater.2008, 18 (19), 2963–2971.(17) Diaz-Lopez, R.; Tsapis, N.; Libong, D.; Chaminade, P.; Connan,

C.; Chehimi, M. M.; Berti, R.; Taulier, N.; Urbach, W.; Nicolas, V.; Fattal,E. Biomaterials 2009, 30 (8), 1462–1472.(18) Sun, S. H.; Zeng, H. J. Am. Chem. Soc. 2002, 124 (28), 8204–

8205.(19) Peng, S.; Wang, C.; Xie, J.; Sun, S. H. J. Am. Chem. Soc. 2006,

128 (33), 10676–10677.(20) Hyeon, T.; Lee, S. S.; Park, J.; Chung, Y.; Bin Na, H. J. Am.

Chem. Soc. 2001, 123 (51), 12798–12801.(21) Cheng, F. Y.; Su, C. H.; Yang, Y. S.; Yeh, C. S.; Tsai, C. Y.; Wu,

C. L.; Wu, M. T.; Shieh, D. B. Biomaterials 2005, 26 (7), 729–738.(22) Wan, S. R.; Huang, J. S.; Yan, H. S.; Liu, K. L. J. Mater. Chem.

2006, 16 (3), 298–303.(23) Kim, D. K.; Mikhaylova, M.; Zhang, Y.; Muhammed, M. Chem.

Mater. 2003, 15 (8), 1617–1627.(24) Tao, K.; Dou, H. J.; Sun, K. Colloids Surf., A 2008, 320 (1�3),

115–122.

(25) Tao, K.; Dou, H. J.; Sun, K. Chem. Mater. 2006, 18 (22),5273–5278.

(26) Lu, R.; Tao, K.; Sun, K.; Dou, H. J.; Xu, B. Colloid Polym. Sci.2010, 288 (3), 353–357.

(27) Anton, N.; Benoit, J. P.; Saulnier, P. J. Controlled Release 2008,128 (3), 185–199.

(28) Koo, H. Y.; Chang, S. T.; Choi, W. S.; Park, J. H.; Kim, D. Y.;Velev, O. D. Chem. Mater. 2006, 18 (14), 3308–3313.

(29) Hickey, R. J.; Haynes, A. S.; Kikkawa, J. M.; Park, S. J. J. Am.Chem. Soc. 2010, 133 (5), 1517–1525.

(30) Trent, J. S. Macromolecules 1984, 17 (12), 2930–2931.(31) Gong, Y. J.; Zhang, D.; Gong, X. F.; Tan, K. B.; Liu, Z. Chin.

Phys. 2006, 15 (7), 1526–1531.(32) Xu, B.; Dou, H. J.; Tao, K.; Sun, K.; Lu, R.; Shi, W. B. J. Polym.

Res. 2011, 18 (1), 131–137.(33) Gupta, A. K.; Gupta, M. Biomaterials 2005, 26 (18), 3995–

4021.(34) McClune, W. F. J. JCPDS: Swarthmore 1980.(35) Hamoudeh, M.; Faraj, A. A.; Canet-Soulas, E.; Bessueille, F.;

L�eonard, D.; Fessi, H. Int. J. Pharm. 2007, 338 (1�2), 248–257.(36) Trent, J. S.; Scheinbeim, J. I.; Couchman, P. R.Macromolecules

1983, 16 (4), 589–598.(37) Liu, M.; Dong, J.; Yang, Y.; Yang, X.; Xu, H. Eur. Polym. J. 2005,

41 (2), 375–382.(38) Wassel, R. A.; Grady, B.; Kopke, R. D.; Dormer, K. J. Colloids

Surf., A 2007, 292 (2�3), 125–130.(39) Cong, Y.; Wang, G.; Xiong, M.; Huang, Y.; Hong, Z.; Wang, D.;

Li, J.; Li, L. Langmuir 2008, 24 (13), 6624–6629.(40) Tao, K.; Dou, H. J.; Sun, K. Colloids Surf., A 2006, 290 (1�3),

70–76.(41) Varanda, L. C.; M. Jafelicci, J.; Tartaj, P.; Grady, K. O.;

Gonzalez-Carreno, T.; Morales, M. P.; Munoz, T.; Serna, C. J. J. Appl.Phys. 2002, 92 (4), 2079–2085.

(42) Zhang, D.; Gong, X. F.; Liu, J. H.; Shao, L. Z.; Li, X. R.; Zhang,Q. L. Ultrasound Med. Biol. 2000, 26 (2), 347–351.

(43) Asanuma, T.; Belohlavek, M.; Viggen, K.; Seward, J. B. J. Am.Coll. Cardiol. 2001, 37 (2), 416A–417A.

(44) Greis, C. Eur. J. Radiol. 2004, 14, P11–P15.(45) Lindner, J. R. Nat. Rev. Drug Discovery 2004, 3 (6), 527–532.(46) Yang, F.; Li, L.; Li, Y. X.; Chen, Z. P.; Wu, J. R.; Gu, N. Phys.

Med. Biol. 2008, 53 (21), 6129–6141.(47) Park, J. I.; Jagadeesan, D.; Williams, R.; Oakden, W.; Chung,

S. Y.; Stanisz, G. J.; Kumacheva, E. ACS Nano 2010, 4 (11), 6579–6586.(48) Basarkar, A.; Devineni, D.; Palaniappan, R.; Singh, J. Int. J.

Pharm. 2007, 343 (1�2), 247–254.(49) Yin, H.; Casey, P. S.; McCall, M. J. J. Nanosci. Nanotechnol.

2010, 10 (11), 7565–7570.