Superparamagnetic plasmonic nanocomposites: Synthesis and characterization studies

Chemical Engineering Journal 264 (2015) 66–76

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier .com/locate /ce j

Superparamagnetic plasmonic nanocomposites: Synthesisand characterization studies

http://dx.doi.org/10.1016/j.cej.2014.11.0591385-8947/� 2014 Elsevier B.V. All rights reserved.

⇑ Corresponding author. Tel.: +98 21 64542978.E-mail address: [email protected] (S.N.A. Keivani).

Seyedeh Narjes Abdollahi Keivani a,⇑, Malek Naderi a, Ghassem Amoabediny b,c

a Department of Mining and Metallurgical Engineering, Amirkabir University of Technology, Tehran, Iranb Department of Chemical Engineering, School of Engineering, University of Tehran, Tehran, Iranc Research Center for New Technologies in Life Science Engineering, University of Tehran, Tehran, Iran

h i g h l i g h t s

� Long mixing time leads toagglomeration of modified magnetiteparticles.� The intensity of XRD peaks depends

on gold mass ratio in nanostructures.� Diamagnetic gold shell causes

reduction in magnetic saturation ofnanocomposites.� SPR peak shows a red shift to longer

wavelengths upon forming magneticshell.� Attachment of magnetite particles to

APTES-modified nanoshells formssmooth shell.

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 13 August 2014Received in revised form 7 November 2014Accepted 10 November 2014Available online 18 November 2014

Keywords:Superparamagnetic nanoparticlesPlasmonic nanostructuresGold nanoshellsCore–shell nanostructuresMultifunctional nanocomposites

a b s t r a c t

Bifunctional magneto-optical nanocomposites have recently received significant attention because oftheir simultaneous magnetic and optical properties, and they can be considered as one of the most prom-ising candidates for future medical diagnosis and therapy. In this paper, multilayer magnetic gold nano-shells were prepared by coating a magnetic shell around outer surface of silica-gold nanoshells (GNs) viausing three procedures based on surface modification of gold nanoshells or magnetite nanoparticles(MNs) in order to increase the affinity between these nanostructures. The procedures used in this workinclude attachment of MNs onto the surface of 3-aminopropyltriethoxysilane (APTES) modified GNs (P1),attachment of APTES-modified MNs onto the surface of GNs (P2) and attachment of APTES-modified MNsonto the surface of 3-mercaptopropionic acid (MPA) modified GNs (P3). From the microscopic character-izations, it is inferred that the surface of the nanocomposites obtained from P1 is smooth. So, P1 is morepreferable to produce spherical smooth SiO2@Au@Fe3O4 nanostructures. Besides, the spectroscopic andmagnetization data of obtained nanocomposites reveal that all the synthesized nanocomposites havesuperparamagnetic plasmonic nature. Therefore, they can be considered as possible options for multi-modal applications requiring bifunctional magnetic plasmonic structures such as simultaneous MRIimaging and photothermal therapy.

� 2014 Elsevier B.V. All rights reserved.

1. Introduction

Nanotechnology as an interdisciplinary research field providesnotable potential for a wide range of applications including cataly-sis, biosensing, optical devices, nanomedicine due to the unique

S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76 67

optical, chemical, magnetic, catalytic and photoelectrochemicalproperties of nanostructures [1,2]. Currently, using the nanomateri-als for medical diagnosis and therapy has aroused more attractionamong the diverse applications of nanotechnology and variouskinds of nanostructures have been introduced and employed fornumerous biomedical applications [3]. For example PEG-functional-ized silica nanoparticles were used to locate the damages in theinjured spinal cord tissue [4], gold nanocages were utilized as thecontrast agent of photoacoustic tomography (PAT) [5] and hollow/rattle-type mesoporous Au@SiO2 nanoparticles were used asin vitro cancer-targeting fluorescent imaging agent [6].

Among different structures, gold-based nanostructures is one ofthe most studied ones owing to their optical properties originatingfrom the collective oscillation of the conduction electrons inresponse to the optical excitation [7,8] as well as their biocompati-bility, low toxicity and convenient preparation and delivery [9,10].The optical properties of gold-based nanostructures depend on theirstructure, shape, size, interparticle interactions, dielectric proper-ties of the surrounding environment of the nanoparticle and theoptical response can be controlled by changing these parameters[7,8,10,11]. For instance, gold nanoshells consisting of a dielectriccore such as silica coated with a thin layer of gold exhibit strongoptical response to different wavelengths from the visible regionto the near infrared (NIR) upon varying the core diameter to shellthickness ratio and changing the dielectric constant of the core orthe embedded medium [12,13]. Tunability of gold nanoshells toNIR region makes them more preferable among gold-based nano-structures for medical application requiring nanostructures withSPR band in NIR region (also called the therapeutic window region);particularly using gold nanoshells for photothermal ablation ther-apy (PTA) has gained lots of attention lately [14,15].

On the other hand, nowadays studying the iron oxide nanopar-ticles especially magnetite nanoparticles is the attractive subjectfor many researchers due to their interesting superparamagneticproperties. Their ability to be easily directed toward a certain loca-tion through using an external magnetic field, their biocompatibil-ity and easy removal from body through natural elimination orexcretion processes by liver, spleen or kidney make them usefulfor many biomedical applications including localized hyperthermiatreatment of cancer cells, using as contrast agent for magneticresonance imaging (MRI), magnetic separation, RNA and DNApurification, cell labeling and targeted drug delivery [16–19].

Recently, some researchers have attempted to integrate varietiesof functional nanoparticles into a single nanostructures in order tointroduce multifunctional nanoparticles capable of performingmultiple tasks simultaneously [3,20,21]. Multilayer magnetic goldnanoshells are a class of bifunctional nanoparticles with combinedmagnetic and optical properties. In fact, these nanostructures notonly are capable to absorb or emit different wavelengths just sameas gold-based nanostructures, but also are sensitive to externalmagnetic field. The unique properties of the magneto-optical nano-composites make them considerable candidates for multimodaltherapeutic treatments such as diagnostic imaging and drug deliv-ery using magnetic properties along with simultaneous photother-mal therapy using optical properties [22–25].

However, most of up to now reported multilayer magnetic goldnanoshells are consisted of the silica-coated magnetite nanoparti-cles as cores which are encapsulated within a thin gold layer[9,23,26–28]. Besides, a few papers reported multilayer nanocom-posites obtained through formation gold shell around magnetite-seeded silica particles [21,22,29]. But, coating magnetic particleswith gold layer leads to significant decreasing in the magneticmoment and superparamagnetic behavior of MNs due to diamag-netic nature of gold [30].

Herein, we introduce a new strategy to produce multilayermagnetic nanoshells and describe synthesis and characterization

of SiO2@Au@Fe3O4 nanocomposites. The silica spheres were syn-thesized and were coated with gold layer resulting in preparationof silica-gold nanoshells. The magnetic shell was obtained aroundGNs through using three new and facile strategies via formingcovalent or non-covalent binding between the surface of gold shelland magnetite particles. The magnetic characterization of thesynthesized nanocomposites reveals that although the magneticsaturation of multifunctional particles is lower than the one foras-prepared MNs due to presence of gold layer, they still showsuperparamagnetic behavior like synthesized MNs. Furthermore,the spectroscopy results show that forming magnetic layer aroundGNs does not diminish the optical properties of the nanostructuresand they still show proper optical response to visible wavelengthsjust same as as-prepared GNs. Hence, within this work throughintroducing three new and facile procedures, we could preparemultifunctional superparamagnetic nanoshells without sacrificingoptical properties of nanostructures.

2. Experimental

2.1. Materials

Ammonium hydroxide (25%), ethanol absolute, ferric chloridehexahydrate (P 99%), ferrous chloride tetrahydrate (P 99%),hydrochloric acid (37%), potassium carbonate, trisodium citratedihydrate and tetraethylorthosilicate (TEOS) (P 99%) were pur-chased from Merck. 3-Aminopropyltriethoxysilane, 3-mercapto-propionic acid (P 99%) and nitric acid (65%) were purchasedfrom Sigma–Aldrich. Sodium borohydride (98%) and formaldehydesolution (37%) were purchased from Acros and Applichem, respec-tively. All chemicals were used as received.

2.2. Synthesis of amine-functionalized magnetite nanoparticles

Magnetite nanoparticles were produced through co-precipita-tion method [31] via the following procedure. 500 ml of aqueoussolution of ammonia (1 M) was deoxidized for 30 min under theargon gas flow with vigorous stirring at room temperature. By mix-ing 40 ml of aqueous solution of FeCl3�6H2O (1 M) and anothersolution made of dissolving FeCl2�4H2O (3.97 g) in 10 ml of hydro-chloric acid solution (2 M), a yellow solution was prepared. Then, itwas added to the ammonia solution. As these two solutions weremixed, a black suspension was produced indicating formation ofMNs. The obtained magnetic particles were rinsed for several timesby using a magnet and were dispersed in 500 ml deionized water.

In order to modify magnetic particles, the MNs were sonicatedfor 5 min. 10 ml of nanoparticles was stirred with 0.1 ml of APTESover night. Then, it was centrifuged for 15 min and the depositedsolid was dispersed in deionized water.

2.3. Synthesis of modified silica-gold nanoshells

Silica-gold nanoshells were fabricated as described previously[32,33]. Concisely, silica spheres were prepared by stirring ethanol,deionized water, ammonium hydroxide and tetraethylorthosilicateat 40 �C for 3 h. Silica nanoparticles were modified by APTES andthe functionalized particles were centrifuged and were dispersedin deionized water. Gold nanoparticles were prepared throughmixing HAuCl4, trisodium citrate, NaBH4 and room temperaturedeionized water.

In order to synthesize GNs, modified silica spheres and goldnanocolloids were mixed together. Then, the solution was centri-fuged and the deposited pellet was redispersed in deionized water.At the gold shell growth step, gold seeded silica solution was added

Fig. 1. (a) TEM image and (b) XRD pattern of synthesized magnetite particles. (c) Magnetization as a function of applied magnetic field for nanoparticles at 300 K.

68 S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76

to gold hydroxide solution and it was stirred in presence offormaldehyde.

The obtained GNs were sonicated for 5 min. Then, the nanopar-ticles were modified through stirring of 100 ml of GNs with 0.1 mlof APTES over night. Next, after separation step via using centri-fuge, the obtained particles were dispersed in deionized water.The procedure was repeated for 0.1 ml of MPA as modifier, too.

2.4. Synthesis of magnetic gold nanocomplexes

Here, it should be noticed that all the nanostructures were son-icated for 5 min before usage. In this research, three differentmethods were used in order to prepare magnetic nanoshells. First,for P1, 200 ml of APTES-functionalized GNs were mixed mechani-cally with 10 ml magnetite particles for 24 h. For P2, 200 ml ofGNs and 10 ml of APTES-functionalized MNs were mixed for24 h. Finally, for P3, 200 ml of MPA-functionalized GNs and 10 mlof functionalized MNs were stirred for one day, too. After mixingstep, the mixture was centrifuged and the particles were dispersedin deionized water.

2.5. Characterization

FTIR spectroscopy was performed by using a Thermo NicoletNexus 670 FTIR spectrometer (Thermo Electron Corporation, United

States). The samples were prepared by dropping several droplets ofthe suspension on a watch glass and evaporation of dispersant in anoven at 35 �C. The UV–visible absorption spectra were recorded by aT80+ double beam spectrophotometer (PG Instrument, United King-dom). The zeta potential of the nanoparticles was recorded by a ZEN3600 zetasizer (Malvern, United Kingdom). An Equinox 3000 X-raydiffractometer (Inel, France) operating at 40 kV and 30 mA usingCuKa radiation (k = 1.5418 Å) used in order to obtain the X-raydiffraction patterns of the samples. Magnetization curves weremeasured on a MDK 6 magnetometer (Meghnatis Daghigh KavirCo., Iran) by varying the field up to 14 kOe at 300 K. The magneticparticles in suspension were collected with a magnet and weredried at room temperature in a seal desiccator. S-4160 fieldemission scanning electron microscope (Hitachi, Japan) paired withan energy dispersive X-ray spectrometer (Oxford Instrument,United Kingdom) was used to study the surface morphology andthe chemical composition of the nanostructures. Moreover, someFE-SEM images were taken by a MIRA3 field emission scanning elec-tron microscope (Tescan, Czech). The samples were dried in an ovenat 30 �C, were placed on a cylinder and were coated with a thin layerof gold. Transmission electron microscopy measurements were per-formed with an EM 208 electron microscope (Philips, Netherlands)operating at a bias voltage of 100 kV. Furthermore, a CM 30 electronmicroscope (Philips, Netherlands) working at 150 kV used to takeHR-TEM images and the selected area electron diffraction (SAED)

Fig. 2. (a) TEM image of silica nanoparticles. (b,c) FE-SEM and TEM images of gold seeded silica particles. (d,e) FE-SEM and TEM images of silica-gold nanoshells.

S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76 69

pattern of samples. The sonicated nanostructures were droppedonto a carbon-coated copper grid and were dried at ambientatmosphere, then, they were characterized with TEM and HR-TEMmicroscope. The particle sizes of nanoparticles were measuredby Digimizer image analysis software (version 4.1.1.0, MedCalcSoftware, Belgium).

3. Results and discussion

3.1. Characterization of magnetite nanoparticles

Magnetite nanoparticles with an average particle size of12.6 nm with a standard deviation about 18% were synthesized

as mentioned before. It can be inferred from Fig. 1a presentingTEM image of MNs that they are almost spherical. The X-ray dif-fraction pattern of the particles is represented in Fig. 1b. The posi-tion and the relative intensity of the peaks are well-matched withthe (220), (311), (222), (400), (422), (511), (440), (531), (620),(533), (622) and (444) planes of the standard XRD date for mag-netite (JCPDS file No. 19-0629 [34,35]) confirming the cubic inversespinel crystalline structure of synthesized MNs. Although, the stan-dard XRD diffraction pattern of maghemite (c-Fe2O3, JCPDS file No.04-0755) is similar to the magnetite [36], however it can be inter-preted that the resulting nanoparticles are pure magnetite due toabsence of other diffraction peaks of maghemite particularly thepeak rises up at 2h = 32.17� corresponding to (300) plane of

Fig. 3. FTIR spectra of (i) SiO2@Au nanoshells and (ii) APTES modified GNs.

70 S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76

maghemite (Fig. S1 demonstrates the comparison between diffrac-tion pattern of prepared MNs and the standard JCPDS pattern ofmaghemite). The average particle size of the magnetite particleswas calculated 13.5 nm for the (311) peak by using Scherrer equa-tion [37], which it is too close to the particle size obtained fromTEM image shown in Fig. 1a. The magnetic hysteresis loop of theMNs was measured at 300 K and the result is depicted in Fig. 1c

Fig. 4. (a) FE-SEM image of nanostructures synthesized by P1. (b) HR-TEM image and cFig. 4a.

(the inset image is the magnified center of the hysteresis diagram).The nonlinear and reversible diagram with zero coercivity and zeroremanence exhibits the superparamagnetic behavior of the parti-cles. It means that at room temperature, the thermal energy is suf-ficient to randomize the magnetization moment due to smallcrystal size of MNs [38]. However, the value of magnetic saturation(Ms) for synthesized nanoparticles is 70.22 emu/g which is lowerthan that for the bulk magnetite (84.5 emu/g [39]). Begin-Colinet al. reported that the difference between these two valuesdepending on the particle size is due to surface or volume spincanting as well as an intermediate composition of magnetite andmaghemite of the nanostructures conducting to the presence ofoxidation defects [40]. Furthermore, an estimate of the averagemagnetic particle size can be calculated by fitting the slope ofthe magnetization near zero field region from magnetization curvein which the major contribution comes from the largest particlesfor superparamagnetic nanoparticles. Hence, the maximum mag-netic diameter (Dm) can be estimated from following equation[41,42]:

Dm ¼18kBT

pxi

qM2s

!1=3

ð1Þ

Here, kB is the Boltzmann constant (1.3806 � 10�16 erg/K), T is thetemperature (300 K), q is the density of magnetite (5.18 g/cm3

[43]), Ms is the magnetic saturation (70.22 emu/g) and xi = (dM/dH)H ? 0 is the initial magnetic susceptibility which was obtainedby measuring the slope of linear portion of the hysteresis diagramnear the zero region (0.087 emu/g * Oe). Thus, according to the

orresponding SAED pattern of magnetic gold nanoshells. (c) EDS pattern related to

Fig. 5. FTIR spectra of (i) bare and (ii) amino functionalized magnetitenanoparticles.

S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76 71

Eq. (1), the maximum magnetic particle of the sample was calcu-lated to be 9.3 nm. This value is smaller than that observed forTEM image and the calculated one from XRD pattern. Accordingto previous studies, it is premised that the difference is associatedwith a thin magnetically disordered layer called dead layer on thesurface of MNs [44,45].

3.2. Characterization of silica-gold nanoshells

TEM image of silica nanoparticles synthesized according to Stö-ber method [46] is shown in Fig. 2a. From the image, the averagediameter of these spherical particles was measured to be about130 nm (standard deviation about 15%). Due to presence of silanolgroup (Si–OH) on the surface of silica particles, they are negativelycharged (zeta potential of the particles was ��34 mV atpH = 11.5). Since, pH of the solution at seeding step of preparingGNs in this work is about 9, so the affinity between the silica par-ticles and negatively charged gold nanocolloids is restricted.Because of presence of hydroxyl groups on silica surface, the cova-lent bond between silica spheres and organosilanes can be formed[47]. Hence, we modified the silica cores with APTES before seed-ing step in order to introduce amine group on the surface andchange the surface charge. Fig. S2 presents the FTIR spectra of bareand modified silica particles demonstrating formation of amine-terminated silica nanoparticles after functionalizing step. Fig. 2band 2c shows FE-SEM and TEM images of gold immobilized silicaspheres. It is obvious that gold nanoparticles deposited on the sil-ica surface partially due to repulsive Coulomb interactionsbetween gold colloids [48]. In the next step; gold shell growth step,reduction of gold hydroxide ions on the surface of the cores leadsto a 25 nm complete gold shell around silica spheres and formationof gold nanoshell with an average diameter about 155 nm and astandard deviation about 9% (Fig. 2d and 2e). XRD pattern of sil-ica-gold nanoshell is shown in Fig. S3 depicting the diffractionpeaks of face centered cubic structure (FCC) of the gold which illus-trates crystalline structure of gold shell. Furthermore, the 24� peakassigned to SiO2 is diminished indicating the uniform surfacecoverage of silica spheres with gold and formation of a completeshell around silica cores.

3.3. Characterization of multilayer magnetic gold nanoshells

In this work, the last step of producing magnetic gold nanoshellsis attachment of MNs onto the GNs surface. Hydrous magnetite

particle is an amphoteric species which means that the positive(Fe–OH2

+) or negative (Fe–O�) charges can develop on the surfaceof the particle depending on pH of the solution. The surface chargeof iron oxide particles is positive below the pH of the point of zerocharge (PZC), while it is negative above it [49,50]. The PZC of MNswas determined to be within pH range 6–7 in previous studies[50,51]. On the other hand, the zeta potential of synthesized GNswas measured to be ��12.2 mV at pH = 6.5. As the formation ofthe magnetite shell around the GNs carries out at pH = 8–8.5 whichthe surface charge of both particles is negative within this range, theimmobilizing of MNs on GNs surface through attraction betweenthe opposite charges cannot be obtained. Moreover, the covalentbonding cannot lead to formation of magnetic gold nanoshells. So,surface modification of the prepared nanostructures is essential sothat the attachment of MNs to GNs becomes possible.

Here, we used three different kinds of strategy to functionalizethe negatively charged particles and produce the magnetic goldnanoshells.

First, we used APTES to modify synthesized GNs (P1). By com-paring Fig. 2S and FTIR result of GNs shown in Fig. 3, curve i, itcan be inferred that although no new peak is rose up upon forma-tion of gold shell around silica spheres, decreasing the intensitiesof the peaks corresponding to the silica network as well as elimina-tion of NH peak demonstrate production of GNs. According toFig. 3, after modification step of GNs with APTES, the infrared peakassigning to NH bond was not appeared. Moreover, the intensitiesof existing peaks were increased. Based on obtained results, weinterpreted that APTES connected to the GNs surface through itsamine group. In fact, in amine group, nitrogen has a lone electronpair which can be used as electron donor. It means that aminegroup is able to bind dative covalent bond with other elementsor compounds via electron donation by nitrogen [52]. Therefore,it is supposed that during mixing step, available APTES aroundGNs surface tends to share its electron pair with gold and binds achemical bond with gold nanoshells particles. On the other hand,it is reported that amine group of APTES molecule protonates atpH = 7 and the obtained protonated functional group (–NH3

+) canbind to metallic particles via electrostatics interaction [48,53]. Inconclusion, APTES attaches to the GNs surface from amine endbecause of either the coordinate bond or electrostatics interactionsimilar to what reported by Zare et al. [54]. As a consequence,ethoxysilane group presents on the surface of gold nanoshellsconfirming FTIR results.

FE-SEM and HR-TEM images of obtained magnetic nanoparti-cles using P1 presented in Fig. 4a and 4b (another image is shownin Fig. S4a) demonstrate formation of the complete magnetic shellaround GNs and preparation of multifunctional magnetic nanoshellwith a mean diameter about 168 nm (standard deviation about16%). The observed rings in SAED pattern (the inset image atFig. 4b) are attributed to the (220), (311), (400), (511) and(440) planes of magnetite from inside to outside. It confirms thatthe magnetic particles cover the metallic shell completely andmagnetic gold nanoshells were obtained. Moreover, it indicatespolycrystalline structure of synthesized magnetite particles. EDSpattern of magnetic nanoshells corresponding to Fig. 4a presentedin Fig. 4c certifies microscopic images and SAED pattern. It must benoted that before taking FE-SEM images, the samples were coatedwith a thin layer of gold to increase the sample conductivity inorder to get better images. It may have small influence on theEDS pattern result and enhance the intensity of gold in the sample.

The second method involves modification of the MNs withAPTES (P2), and as the consequence, amine-functionalized magne-tite particles were obtained. The FTIR spectra of MNs before andafter modification are shown in Fig. 5. As mentioned in the graphi, two peaks observed at 445 and 569 cm�1 represent the stretchvibration of Fe–O bond [55] demonstrating preparation of MNs.

Fig. 6. (a) FE-SEM image of nanocomposites obtained by P2. (b) HR-TEM image and corresponding SAED pattern of magnetic gold nanoshells. (c) EDS pattern related toFig. 6a.

Fig. 7. FTIR spectra of (i) SiO2@Au nanoshells and (ii) carboxylic acid modified goldnanoshells.

72 S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76

After modification, development of several new peaks shown up at981, 1119, 1402, 1560 and 3442 cm�1 confirms production ofamine terminated magnetite particles through formation of cova-lent bond between hydroxyl groups of MNs and APTES. It must

be noted that the Fe–O–Si bond forms at around 580 cm�1 overlap-ping Fe–O stretching vibration of magnetite at 569 cm�1.

Amine group on the surface of MNs can bind to the GNs surfacethrough covalent and non-covalent interaction. Initially, the aminegroup can donate two electrons and bind the dative covalent bondwith gold. Hence, the Au–N coordinate bond between functional-ized MNs and GNs can be formed leading to attachment of MNson GNs surface. Second, it was reported that presence of aminegroup on the surface of MNs shifts PZC of the particles to�pH = 10 [47,56]. Thus, the surface charge of the MNs at attach-ment step would be positive which means the positively chargedMNs can attach to GNs surface via electrostatic attraction betweenopposite charges. Therefore, it can be interpreted that bondingbetween amine functionalized MNs and GNs is a combination ofcoordination bond and electrostatic interaction.

Fig. 6 presents FE-SEM and HR-TEM images of P2 prepared mag-netic gold nanoshells (another image is shown in Fig. S5). Accord-ing to the results, the complete magnetic shell around the GNs wasobtained. However, several clusters of aggregated MNs are obser-vable in the Figs. 6a and S5. Although using APTES for functionali-zation of MNs can prevent agglomeration of the magnetiteparticles, as what Pramanik et al. reported it may decrease theionic species on the particles surface in some concentration as wellas weakening the repulsive interaction between the MNs leadingto aggregation of the particles [17]. We assumed that a long mixingtime may have the same effect as high concentration of APTES andas a result, the functionalized MNs stuck together during mixingstep. Therefore, for P2, the mixing time is a two-edged sword

Fig. 8. (a) FE-SEM image and (b) TEM image and corresponding SAED pattern of magnetic gold nanoshells prepared by P3. (c) EDS pattern related to Fig. 8a.

S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76 73

which not only it increases the surface coverage of GNs with mag-netite, but also it causes agglomeration of unattached MNs that isobvious in the microscopic results. Furthermore, the rings of SAEDpattern of magnetic nanoshells shown in Fig. 6b are not as bright asthose for P1 produced nanostructures (Fig. 4b) that may be due tofunctionalization of MNs.

For the last strategy, we modified both the magnetite particlesand gold nanoshells with APTES and MPA, respectively (P3).According to FTIR spectrum of MPA-modified GNs (Fig. 7), theintensities of peaks indexed to silica network are decreased andthree new peaks are shown up at 1385, 1457 and 1714 cm�1

assigning to stretch vibration of C–O bond, bending vibration ofC–O–H bond and stretch vibration of C@O bond [57], respectively;confirming preparation of carboxylic acid terminated goldnanoshells.

Previous searches revealed that among different functionalgroups, thiol has the strongest affinity toward gold nanoparticlesand it links to the gold particles surface through binding covalentbond and forming Au–S bond which is much stronger than Au–None [58,59]. Thus, we suppose that more negatively charged siteswould be available on the surface of MPA-modified gold nanoshellsrelative to APTES-modified ones. On the other hand, at a pH valuenear 9, both the amine and carboxylic acid group are completelyionized [60]. Hence, we supposed that the attractive electrostaticinteraction between negatively charged carboxylic acid and posi-tively charged amine group reaches to its maximum value leadingto much more attachment of amine-functionalized MNs and MPA-functionalized gold nanoshells. Moreover, while both the amine

and carboxylic acid groups are ionized; at pH values above the acidpKA and below the amine pKA (pKA(COOH) < pH < pKA(amine)), thehydrogen bonding between the functional groups can be obtainedwhich is classified among the strong hydrogen bond categories[61,62]. Here, the hydrogen bond can increase the affinity betweenmodified particles. Therefore, in the case of P3, the immobilizationof MNs on the surface of GNs would be due to non-covalentinteractions; attractive electrostatic interaction between positivelycharged particles as well as hydrogen binding between thecarboxylic acid and the protonated amine groups.

As shown in Figs. 8a and S6, although the magnetic gold nano-shells were produced, the surface of the nanostructures is not assmooth as that for P1 and P2 produced nanoparticles. In fact, thesurface roughness of P3 synthesized nanocomposites is more thanprevious ones. It seems that the magnetic shell did not form homo-geneously around the GNs so that the particles even lose theirspherical shapes. Moreover, the obtained magnetic nanoshellsaggregate and the particles stick together as distinguishing oneparticle is somewhat hard in some areas. Besides, more aggregatedcolonies of unimmobilized MNs can be seen in the images. TheSAED rings are not as clear as previous structures and the ringschanged from solid lines to semi-dotted type that could be dueto irregular attachment of MNs onto the surface of gold nanoshells[63].

Crystallographic structure of the nanostructures was studiedand the XRD diffraction patterns are presented in Fig. 9a. As char-acterized in the diagram, the observed peaks are corresponded togold and magnetite cubic planes indicating formation of magnetic

Fig. 9. (a,b) XRD diffraction patterns and magnetic hysteresis loops of magnetic gold nanoshells prepared using (i) P1, (ii) P2 and (iii) P3 method. (c) Normalized UV–visiblespectra of (i) GNs and magnetic nanostructures obtained from (ii) P1, (iii) P2 and (iv) P3 method.

74 S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76

gold nanocomplexes. It can be understood easily that gold diffrac-tion peaks dominate the XRD patterns and their intensity is higherthan those for magnetite because of the much heavier atomic massof gold atoms [64]. Besides, the peaks of the curve ii related to P2are sharper may due to a few nanocomposites with incompletemagnetic shell around GNs that can be detected in Fig. 6a resultingin increasing the gold mass and subsequently higher mass ratio ofgold relative to others elements.

Magnetic properties of these nanostructures were characterizedat 300 K and the results demonstrate their superparamagneticbehavior at room temperature due to showing no remanence andno coercivity (Fig. 9b). The magnetic saturation of nanocomplexesobtained from P1, P2 and P3 were measured to be 22.9, 7.4 and24.9 emu/g at 10 kOe, respectively that are considerably higherthan saturation magnetization of previously reported magneticgold nanostructures [9,22,23,26,27,29,65]. However, they arelower than magnetic saturation of synthesized MNs (70.22 emu/g) and bulk magnetite (84.5 emu/g), especially in the case of P2.This might be due to diamagnetic nature of gold shell contributingthe main portion of total mass of magnetic gold nanocomposites[64]. Besides, configuration of noncollinear (canted) spin structuresrequiring high magnetic field for magnetic saturation was reportedfor the diamagnetically substituted ferrimagnetic materials (suchas magnetite) which could be assisted in decreasing the saturationof synthesized nanocomposites [30,66]. Thus, according to these

mentioned reasons, we inferred that lower magnetic saturationof nanocomplexes prepared from P2 relative to that for P1 andP3 could be attributed to higher gold mass ratio in this structurethat was noticed above while discussing their XRD patterns.

UV–visible absorption spectra of GNs and magnetic gold nano-shells normalized at their maximum absorption are shown inFig. 9c. The surface plasmon resonance of GNs was measured tobe 550 nm. Upon formation the magnetic layer around GNs, theSPR band of nanocomposites was shifted to longer wavelengths.From the previous studies, it is evolved that the position of SPRband of metallic core–shell nanostructure definitely depends onrefractive index of the core, shell and the medium, and by increas-ing refractive index of the surrounding medium, the maximumabsorption peak shows a red shift to longer wavelengths [67,68].Here, since the refractive index of water and magnetite aren = 1.33 and n = 3, respectively, the red-shift of SPR band afterchanging the surrounding matter of gold shell from water to mag-netite particles is due to increasing of the refractive index of thesurrounding medium. Moreover, for the nanocomposites preparedby P2 and P3, the SPR band is shown up at longer wavelengths incomparison to that for P1 produced magnetic nanoshells. Thismay be due to less inter-particle gap between magnetic gold nano-shells arising from sticking of the particles (discussed before whilestudying microscopic results) for P1 synthesized particles leadingto less red-shift of the SPR [69].

S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76 75

According to the above discussions, between these three meth-ods, by choosing the most appropriate mixture of the reactantsand experiment condition, the particles obtained by P1 are moreuniform as well as the particles surface is smoother. While, it seemsthat upon modification of the magnetite particles, especially in thecase of P3, the nanocomplexes stick together and the consequence issemi-aggregated particles. Additionally, the surface roughness ofthe particles increased and non-attached magnetite particles aggre-gated and colonized in the samples. Therefore, as all the preparedmagnetic nanocomplexes almost show the same magnetic and opti-cal properties, it can be concluded that among these three testedprocedures to produce magnetic gold nanoshells, the most properone is P1 involving three subsequent steps: synthesizing of magne-tite nanoparticles, preparation of APTES-modified gold nanoshellsand attachment of MNs onto the surface of gold nanoshells.

4. Conclusions

In summary, three new, simple and reproducible procedureswere introduced to produce magnetic gold nanoshells consistingof silica-gold nanoshells surrounded with a magnetic shell. For P1,magnetite particles were attached onto the surface of APTES-modi-fied GNs via forming covalent bind between amine functional groupof modified GNs and magnetite particles or electrostatic interactionbetween oppositely charged particles. The P2 synthesized nano-composites were produced by mixing amine-functionalized MNsand GNs through forming coordinate bond or electrostatic interac-tion, too. However, for P3, the magnetic shell was formed aroundGNs via electrostatic interaction and forming hydrogen bondbetween the carboxylic acid-modified gold nanoshells and theamine-terminated magnetite particles. The characterization datareveals that not only the obtained multilayer magnetic nanocom-posites are plasmonic nanostructures showing optical response tothe electromagnetic wavelengths, but also they show superpara-magnetic behavior enable their manipulation by an external mag-netic field. Therefore, these nanocomposites are likely to utilize ina variety of fields including catalysis and biomedical especiallysimultaneous diagnostic and therapeutic applications due to theircombined superparamagnetic and surface plasmon resonanceproperties.

Acknowledgements

This work was supported by Iran National Science Foundation(INSF) (Grant No. 90004760). The authors gratefully thank the staffmembers of Research Center for New Technologies in Life ScienceEngineering of University of Tehran and Department of Miningand Metallurgical Engineering of Amirkabir University of Technol-ogy (Tehran Polytechnic) for supporting this research. Further-more, S. N. Abdollahi greatly acknowledges Mr. Arash Ghazitabarfor his valuable help.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.cej.2014.11.059.

References

[1] R. Qiao, C. Yang, M. Gao, Superparamagnetic iron oxide nanoparticles: frompreparations to in vivo MRI applications, J. Mater. Chem. 19 (2009) 6274–6293.

[2] W. Cai, T. Gao, H. Hong, J. Sun, Applications of gold nanoparticles in cancernanotechnology, Nanotechnol. Sci. Appl. 1 (2008) 17–32.

[3] Y. Piao, A. Burns, J. Kim, U. Wiesner, T. Hyeon, Designed fabrication of silica-based nanostructured particle systems for nanomedicine applications, Adv.Funct. Mater. 18 (2008) 3745–3758.

[4] B. Chen, M. Zuberi, R.B. Borgens, Y. Cho, Affinity for, and localization of, PEG-functionalized silica nanoparticles to sites of damage in an ex vivo spinal cordinjury model, J. Biol. Eng. 6 (2012) 18.

[5] X. Yang, S.E. Skrabalak, Z.Y. Li, Y. Xia, L.V. Wang, Photoacoustic tomography of arat cerebral cortex in vivo with Au nanocages as an optical contrast agent,Nano Lett. 7 (2007) 3798–3802.

[6] T.T. Wang, F. Chai, C.G. Wang, L. Li, H.Y. Liu, L.Y. Zhang, Z.M. Su, Y. Liao,Fluorescent hollow/rattle-type mesoporous Au@SiO2 nanocapsules for drugdelivery and fluorescence imaging of cancer cells, J. Colloid Interface Sci. 358(2011) 109–115.

[7] S. Link, M.A. El-Sayed, Shape and size dependence of radiative, non-radiativeand photothermal properties of gold nanocrystals, Int. Rev. Phys. Chem. 19(2000) 409–453.

[8] S.K. Ghosh, T. Pal, Interparticle coupling effect on the surface plasmonresonance of gold nanoparticles: from theory to applications, Chem. Rev. 107(2007) 4797–4862.

[9] X. Ji, R. Shao, A.M. Elliott, R.J. Stafford, E. Esparza-Coss, J.A. Bankson, G. Liang,Z.P. Luo, K. Park, J.T. Markert, C. Li, Bifunctional gold nanoshells with asuperparamagnetic iron oxide-silica core suitable for both MR imaging andphotothermal therapy, J. Phys. Chem. C 111 (2007) 6245–6251.

[10] J.Z. Zhang, Biomedical applications of shape-controlled plasmonicnanostructures: a case study of hollow gold nanospheres for photothermalablation therapy of cancer, J. Phys. Chem. Lett. 1 (2010) 686–695.

[11] Y. Song, J. Ding, Y. Wang, Shell-dependent evolution of optical and magneticproperties of Co@Au core�shell nanoparticles, J. Phys. Chem. C 116 (2012)11343–11350.

[12] S.J. Oldenburg, R.D. Averitt, S.L. Westcott, N.J. Halas, Nanoengineering ofoptical resonances, Chem. Phys. Lett. 288 (1998) 243–247.

[13] D. Wu, X. Xu, X. Liu, Influence of dielectric core, embedding medium and sizeon the optical properties of gold nanoshells, Solid State Commun. 146 (2008)7–11.

[14] A.M. Gobin, M.H. Lee, N.J. Halas, W.D. James, R.A. Drezek, J.L. West, Near-infrared resonant nanoshells for combined optical imaging and photothermalcancer therapy, Nano Lett. 7 (2007) 1929–1934.

[15] L.B. Carpin, L.R. Bickford, G. Agollah, T.K. Yu, R. Schiff, Y. Li, R.A. Drezek,Immunoconjugated gold nanoshell-mediated photothermal ablation oftrastuzumab-resistant breast cancer cells, Breast Cancer Res. Treat. 125(2011) 27–34.

[16] I. Robinson, L.D. Tung, S. Maenosono, C. Wälti, N.T.K. Thanh, Synthesis of core–shell gold coated magnetic nanoparticles and their interaction with thiolatedDNA, Nanoscale 2 (2010) 2624–2630.

[17] S. Mohapatra, N. Pramanik, S. Mukherjee, S.K. Ghosh, P. Pramanik, A simplesynthesis of amine-derivatised superparamagnetic iron oxide nanoparticlesfor bioapplications, J. Mater. Sci. 42 (2007) 7566–7574.

[18] A.G. Roca, J.F. Marco, M.D. Puerto Morales, C.J. Serna, Effect of nature andparticle size on properties of uniform magnetite and maghemite nanoparticles,J. Phys. Chem. C 111 (2007) 18577–18584.

[19] Y. Song, R. Wang, R. Rong, J. Ding, J. Liu, R. Li, Z. Liu, H. Li, X. Wang, J. Zhang, J.Fang, Synthesis of well-dispersed aqueous-phase magnetite nanoparticles andtheir metabolism as an MRI contrast agent for the reticuloendothelial system,Eur. J. Inorg. Chem. (2011) 3303–3313.

[20] M.B. Cortie, A.M. McDonagh, Synthesis and optical properties of hybrid andalloy plasmonic nanoparticles, Chem. Rev. 111 (2011) 3713–3735.

[21] S.I. Stoeva, F. Huo, J.S. Lee, C.A. Mirkin, Three-layer composite magneticnanoparticle probes for DNA, J. Am. Chem. Soc. 127 (2005) 15362–15363.

[22] J. Kim, S. Park, J.E. Lee, S.M. Jin, J.H. Lee, I.S. Lee, I. Yang, J.S. Kim, S.K. Kim, M.H.Cho, T. Hyeon, Designed fabrication of multifunctional magnetic goldnanoshells and their application to magnetic resonance imaging andphotothermal therapy, Angew. Chem. 118 (2006) 7918–7922.

[23] V. Salgueiriño-Maceira, M.A. Correa-Duarte, M. Farle, A. López-Quintela, K.Sieradzki, R. Diaz, Bifunctional gold-coated magnetic silica spheres, Chem.Mater. 18 (2006) 2701–2706.

[24] Y. Song, W. Yin, Y.H. Wang, J.P. Zhang, Y. Wang, R. Wang, J. Han, W. Wang, S.V.Nair, H.E. Ruda, Magneto-plasmons in periodic nanoporous structures, Sci.Rep. 4 (2014) 1–9.

[25] M. Melancon, W. Lu, C. Li, Gold-based magneto/optical nanostructures:challenges for in vivo applications in cancer diagnostics and therapy, Mater.Res. Bull. 34 (2009) 415–421.

[26] S.E. Park, J.W. Lee, S.J. Haam, S.W. Lee, Fabrication of double-doped magneticsilica nanospheres and deposition of thin gold layer, Bull. Korean Chem. Soc. 30(2009) 869–872.

[27] J. Ge, Q. Zhang, T. Zhang, Y. Yin, Core–satellite nanocomposite catalystsprotected by a porous silica shell: controllable reactivity, high stability, andmagnetic recyclability, Angew. Chem. Int. Ed. 47 (2008) 8924–8928.

[28] L.N. Kim, E.G. Kim, J. Kim, S.E. Choi, W. Park, S. Kwon, Fabrication andmanipulation of gold 1D chain assemblies using magnetically controllable goldnanoparticles, Bull. Korean Chem. Soc. 33 (2012) 3735–3739.

[29] J. Kim, J.E. Lee, J. Lee, Y. Jang, S.W. Kim, K. An, J.H. Yu, T. Hyeon, Generalizedfabrication of multifunctional nanoparticle assemblies on silica spheres,Angew. Chem. 118 (2006) 4907–4911.

[30] E.V. Shevchenko, M.I. Bodnarchuk, M.V. Kovalenko, D.V. Talapin, R.K. Smith, S.Aloni, W. Heiss, A.P. Alivisatos, Gold/iron oxide core/hollow–shellnanoparticles, Adv. Mater. 20 (2008) 4323–4329.

[31] Y.S. Kang, S. Risbud, J.F. Rabolt, P. Stroeve, Synthesis and characterization ofnanometer size Fe3O4 and c-Fe2O3 Particles, Chem. Mater. 8 (1996) 2209–2211.

76 S.N.A. Keivani et al. / Chemical Engineering Journal 264 (2015) 66–76

[32] S.N. Abdollahi, M. Naderi, G. Amoabediny, Synthesis and physicochemicalcharacterization of tunable silica-gold nanoshells via seed growth method,Colloids Surf. A 414 (2012) 345–351.

[33] S.N. Abdollahi, M. Naderi, G. Amoabediny, Synthesis and characterization ofhollow gold nanoparticles using silica spheres as templates, Colloids Surf. A436 (2013) 1069–1075.

[34] K. Liu, L. Zhao, P. Klavins, F.E. Osterloh, H. Hiramatsu, Extrinsicmagnetoresistance in magnetite nanoparticles, J. Appl. Phys. 93 (2003)7951–7953.

[35] W. Cai, J. Wan, Facile synthesis of superparamagnetic magnetite nanoparticlesin liquid polyols, J. Colloid Interface Sci. 305 (2007) 366–370.

[36] B.Y. Yu, S.Y. Kwak, Assembly of magnetite nanocrystals into sphericalmesoporous aggregates with a 3-D wormhole-like pore structure, J. Mater.Chem. 20 (2010) 8320–8328.

[37] A. Guinier, X-ray Diffraction of Crystals, Imperfect Crystals and AmorphousBodies, first ed., Dover, New York, USA, 1994.

[38] D. Sellmyer, R. Skomski, Advanced Magnetic Nanostructures, first ed., Springer,New York, USA, 2006.

[39] Y. Hou, H. Kondoh, M. Shimojo, E.O. Sako, N. Ozaki, T. Kogure, T. Ohta,Inorganic nanocrystal self-assembly via the inclusion interaction of b-cyclodextrins: toward 3D spherical magnetite, J. Phys. Chem. B. 109 (2005)4845–4852.

[40] W. Baaziz, B.P. Pichon, S. Fleutot, Y. Liu, C. Lefevre, J.M. Greneche, M. Toumi, T.Mhiri, S. Begin-Colin, Magnetic iron oxide nanoparticles: reproducible tuningof the size and nanosized-dependent composition, defects, and spin canting, J.Phys. Chem. C 118 (2014) 3795–3810.

[41] Y.J. Cha, M.J. Kim, Y.H. Choa, J. Kim, B. Nam, J. Lee, D.H. Kim, K.H. Kim, Synthesisand characterizations of surface-coated superparamagnetic magnetitenanoparticles, IEEE Trans. Magn. 46 (2010) 443–446.

[42] M. Chastellain, A. Petri, A. Gupta, K.V. Rao, H. Hofmann, Superparamagneticsilica-iron oxide nanocomposites for application in hyperthermia, Adv. Eng.Mater. 6 (2004) 235–241.

[43] G. Goloverda, B. Jackson, C. Kidd, V. Kolesnichenko, Synthesis of ultrasmallmagnetic iron oxide nanoparticles and study of their colloid and surfacechemistry, J. Magn. Magn. Mater. 321 (2009) 1372–1376.

[44] D.K. Kim, Y. Zhang, W. Voit, K.V. Rao, M. Muhammed, Synthesis andcharacterization of surfactant-coated superparamagnetic monodispersed ironoxide nanoparticles, J. Magn. Magn. Mater. 225 (2001) 30–36.

[45] E.E. Carpenter, Iron nanoparticles as potential magnetic carriers, J. Magn.Magn. Mater. 225 (2001) 17–20.

[46] W. Stöber, A. Fink, E. Bohn, Controlled growth of monodisperse silica spheresin the micron size range, J. Colloid Interface Sci. 26 (1968) 62–69.

[47] B.J. Jankiewicz, D. Jamiola, J. Choma, M. Jaroniec, Silica–metal core–shellnanostructures, Adv. Colloid Interface Sci. 170 (2012) 28–47.

[48] S.L. Westcott, S.J. Oldenburg, T.R. Lee, N.J. Halas, Formation and adsorption ofclusters of gold nanoparticles onto functionalized silica nanoparticle surfaces,Langmuir 14 (1998) 5396–5401.

[49] Z.X. Sun, F.W. Su, W. Forsling, P.O. Samskog, Surface characteristics ofmagnetite in aqueous suspension, J. Colloid Interface Sci. 197 (1998) 151–159.

[50] E. Illés, E. Tombácz, The effect of humic acid adsorption on pH-dependentsurface charging and aggregation of magnetite nanoparticles, J. ColloidInterface Sci. 295 (2006) 115–123.

[51] D.K. Kim, W. Voit, W. Zapka, B. Bjelke, M. Muhammed, K.V. Rao, Biomedicalapplication of ferrofluids containing magnetite nanoparticles, Mat. Res. Soc.Symp. Proc. 676 (2001). Y8.32.1–6.

[52] K.C. Grabar, K.J. Allison, B.E. Baker, R.M. Bright, K.R. Brown, R.G. Freeman, A.P.Fox, C.D. Keating, M.D. Musick, M.J. Natan, Two-dimensional arrays of colloidalgold particles: a flexible approach to macroscopic metal surfaces, Langmuir 12(1996) 2353–2361.

[53] M.A. Aziz, S. Patra, H. Yang, A facile method of achieving low surface coverageof Au nanoparticles on an indium tin oxide electrode and its application toprotein detection, Chem. Commun. 38 (2008) 4607–4609.

[54] D. Zare, A. Akbarzadeh, N. Bararpour, Synthesis and functionalization of goldnanoparticles by using of poly functional amino acids, Int. J. Nanosci.Nanotechnol. 6 (2010) 223–230.

[55] M. Ma, Y. Zhang, W. Yu, H.Y. Shen, H.Q. Zhang, N. Gu, Preparation andcharacterization of magnetite nanoparticles coated by amino silane, ColloidsSurf. A 212 (2003) 219–226.

[56] R.A. Bini, R.F.C. Marques, F.J. Santos, J.A. Chaker, M. Jafelicci Jr., Synthesis andfunctionalization of magnetite nanoparticles with different amino-functionalalkoxysilanes, J. Magn. Magn. Mater. 324 (2012) 534–539.

[57] T. Sainsbury, T. Ikuno, D. Okawa, D. Pacilé, J.M.J. Fréchet, A. Zettl, Self-assemblyof gold nanoparticles at the surface of amine- and thiol-functionalized boronnitride nanotubes, J. Phys. Chem. C 111 (2007) 12992–12999.

[58] E.C. Dreaden, A.M. Alkilany, X. Huang, C.J. Murphy, M.A. El-Sayed, The goldenage: gold nanoparticles for biomedicine, Chem. Soc. Rev. 41 (2012) 2740–2779.

[59] P.A. Buining, B.M. Humbel, A.P. Philipse, A.J. Verkleij, Preparation of functionalsilane-stabilized gold colloids in the (sub)nanometer size range, Langmuir 13(1997) 3921–3926.

[60] A. Gole, S.R. Sainkar, M. Sastry, Electrostatically controlled organization ofcarboxylic acid derivatized colloidal silver particles on amine-terminated self-assembled monolayers, Chem. Mater. 12 (2000) 1234–1239.

[61] M. Sastry, K.S. Mayya, V. Patil, D.V. Paranjape, S.G. Hegde, Langmuir–Blodgettfilms of carboxylic acid derivatized silver colloidal particles: role of subphasepH on degree of cluster incorporation, J. Phys. Chem. B. 101 (1997) 4954–4958.

[62] M. Valtiner, S.H. Donaldson Jr., M.A. Gebbie, J.N. Israelachvili, Hydrophobicforces, electrostatic steering, and acid�base bridging between atomicallysmooth self-assembled monolayers and end-functionalized PEGolated lipidbilayers, J. Am. Chem. Soc. 134 (2012) 1746–1753.

[63] S. Chah, J.H. Fendler, J. Yi, Nanostructured gold hollow microspheres preparedon dissolvable ceramic hollow sphere templates, J. Colloid Interface Sci. 250(2002) 142–148.

[64] R. Bardhan, W. Chen, C. Perez-Torres, M. Bartels, R.M. Huschka, L.L. Zhao, E.Morosan, R.G. Pautler, A. Joshi, N.J. Halas, Nanoshells with targetedsimultaneous enhancement of magnetic and optical imaging andphotothermal therapeutic response, Adv. Funct. Mater. 19 (2009) 3901–3909.

[65] A. Gole, J.W. Stone, W.R. Gemmill, H.C.Z. Loye, C.J. Murphy, Iron oxide coatedgold nanorods: synthesis, characterization, and magnetic manipulation,Langmuir 24 (2008) 6232–6237.

[66] S. Mørup, Spin-canting and transverse relaxation at surfaces and in the interiorof ferrimagnetic particles, J. Magn. Magn. Mater. 266 (2003) 110–118.

[67] Y. Sun, Y. Xia, Increased sensitivity of surface plasmon resonance of goldnanoshells compared to that of gold solid colloids in response toenvironmental changes, Anal. Chem. 74 (2002) 5297–5305.

[68] J. Butet, I. Russier-Antoine, C. Jonin, N. Lascoux, E. Benichou, P.F. Brevet, Effectof the dielectric core and embedding medium on the second harmonicgeneration from plasmonic nanoshells: tunability and sensing, J. Phys. Chem. C117 (2013) 1172–1177.

[69] P.K. Jain, M.A. El-Sayed, Plasmonic coupling in noble metal nanostructures,Chem. Phys. Lett. 487 (2010) 153–164.