lable at ScienceDirect

Materials Chemistry and Physics 175 (2016) 46e53

Contents lists avai

Materials Chemistry and Physics

journal homepage: www.elsevier .com/locate/matchemphys

Chemical synthesis of low-coercivity, silica/Co composites for high-frequency magnetic components

Himani Sharma*, Markondeya Raj Pulugurtha, Rao TummalaPackaging Research Center, Georgia Institute of Technology, 813 Ferst Drive, Atlanta GA 30332, USA

h i g h l i g h t s

� Chemical synthesis of <10 nm cobalt nanoparticles with thin silica shell for electronic device application.� Correlation of structural and magnetic properties.� Low coercivity of 20 Oe showed exchange-coupled cobalt nanoparticles.� High-resolution TEM showed thin silica shell of ~4e5 nm.

a r t i c l e i n f o

Article history:Received 18 August 2015Received in revised form23 January 2016Accepted 30 January 2016Available online 9 March 2016

Keywords:Magnetic materialsNanostructuresChemical synthesisElectron microscopyMagnetometerMagnetic properties

* Corresponding author.E-mail address: himani.sharma@gatech.edu (H. Sh

http://dx.doi.org/10.1016/j.matchemphys.2016.01.0690254-0584/© 2016 Elsevier B.V. All rights reserved.

a b s t r a c t

The paper reports correlation between processing, structural and magnetic properties of cobalt nano-particles dispersed in an insulating oxide matrix. High-temperature, chemical decomposition approachwas utilized to achieve highly-dispersed, ultra-fine cobalt nanoparticles in silica matrix. Synthesis pa-rameters such as precursor to passivating agent ratio were designed to control the metal-to-oxide ratioso as to promote magnetic interactions between the metal particles while keeping the cobalt particle sizein sub-nanometer. High-resolution TEM and XRD studies were carried out to demonstrate face-centeredphase of single-crystal cobalt particles of size ~8e10 nmmonodispersed in thin silica matrix. The particlesize variation as a function of precursor to passivating agent ratio was also studied to yield processconditions for ultra-fine nanoparticles. Composites with sub-10 nm metal particles showed superior softmagnetic properties with a lowest coercivity of 10 Oe. The reduction in coercivity with such fine particlesis attributed to the transition to superparamagnetic regime. Larger particles showed higher saturationmagnetization and higher coercivity of 100 Oe. The coercivity decreased with further increase in particlesize from ~50 nm to 80 nm because of the transition from single domain to multi-domain structures.

© 2016 Elsevier B.V. All rights reserved.

1. Introduction

The technological applications of highly dispersed nanoparticleswith superior properties have potential to revolutionize many as-pects of modern electronics. The need for miniaturization inportable electronic devices with high efficiency, has led to growinginterest in such nanoparticle materials and processes especially forpassive electronic components such as capacitors, antennas andinductors [1,2]. In particular, nanomagnetic materials in inductorshave gained momentum due to their desirable properties such ashigh saturation magnetization (Ms), low coercivity (Hc), and anappropriate anisotropy field (Hk) [3]. This combination of magnetic

arma).

properties enhances the key parameters such as permeability,quality factor, and frequency stability, directly affecting theinductor performance for power and RF components. Traditionally,thick magnetic cores are composed of ferrites that are limited bylower Ms and poor frequency stability because of their low FMR(ferromagnetic resonance) frequency [4,5]. Metal-polymer com-posites with larger spherical particles show higher losses fromeddy-current, lower resistivity and domain wall resonance [6,7].Nanoparticle composites, on the other hand, show low perme-ability because of demagnetization. Improved permeability andfrequency-stability are achieved by reducing the metal particle sizeto nano-scale while maintaining close proximity within particles tofacilitate exchange coupling. However, such nano-scale inter-par-ticle separation between nano-grains, can only be achieved insputtered magnetic thin films [3,8,9] that are grown in highly

H. Sharma et al. / Materials Chemistry and Physics 175 (2016) 46e53 47

controlled environment that not only makes them expensive butalso difficult to form multilayered thick structure that can handleenough power. Solution-derived, larger scale nanocrystals synthe-ses have more recently gained interest for their low-cost and tightsize control [10], however, most of these superstructures are self-assembled with application in biological, medical and pharmaco-logical fields as biosensors, drug-delivery capsules, catalysis ormagnetic data storage [11,12]. The primary reason for limitedapplication of chemically-derived nanoparticles in consumertechnologies is the lower product yields [13] that lends insufficientamounts for thickmagnetic cores in components such as inductors.In the recent past, a few chemical methods have been devised toproduce silica-coated metal nanoparticles at gram scale [14,15],however they showed hard magnetic properties with much largercoercivity rendering them unsuitable for power applications. A keyfactor to utilize solution-derived magnetic nanoparticles in powerpassives is to achieve nanocomposites with high metal densitiesthat can be correlated to energy storage densities. This paper in-vestigates the magnetoestructural correlation of solution-derivedcobalt nanoparticles with high metal densities for their applica-tion in power electronics. Synthesis process was tuned to achievecobalt nanoparticles (<10 nm) with high mono-dispersity in aninsulating oxide matrix to yield exchange-coupled metal nano-particles with superior soft magnetic properties.

2. Experimental

The silica-coated nanoparticles were synthesized using a high-temperature organoemetallic reaction that was pioneered byHyunjoon Song's group [15]. This synthesis approach targets toachieve silica shell on a fine cobalt nanoparticle core. A schematic ofthe methodology is shown in three steps in Fig. 1, illustratingdifferent coating on the nanoparticles at different stages of thesynthesis, including the organic and the oxide. The organic shell ofolelyamine (OA) serves dual purposes, acting as a solvent as well as

Fig. 1. Schematic illustration of evolution of core/shell nanoparticles; Step-1) Thermal decomhydrolysis of TMOS on CoO nanoparticles; Step 3) Gas-phase reduction of CoO to Co, remo

a passivating layer that prevents particle agglomeration and Ost-wald ripening at higher temperatures. The process begins withthermal decomposition of (cobalt acetylacetonate Co(acac)2) pre-cursor, to form cobalt oxide seeds that are passivated by the organicligand, oleylamine (Fig. 1, step-1). The second step involves sus-pending the OA/CoO nanoparticles in a non-polar organic solvent(cyclohexane) and hydrolyzing silica precursor, tetramethylortho-silicate (TMOS) on the cobalt oxide/OA nanoparticles through anonionic surfactant-stabilized water-in-oil microemulsion processat room temperature [16].

The hydrolysis of the alkoxide in this base-catalyzed (withNH4OH) reaction tends to form a three-dimensional network ofxerogel, entrapping the chelated cobalt oxide nanoparticles in it(Fig. 1, Step-2). The facile silica hydrolysis on oxide nanoparticleswas carried out at room temperature for nearly 60 min, followed bymultiple cycles of dispersion and precipitation in ethanol forwashing away unreacted TMOS and byproducts of the hydrolysis.The CoO/SiO2 nanoparticles were finally dried at room temperatureto yield stable brown/green powders. The final step (Step-3, Fig. 1)consisted of thermal treatment for reduction of metal oxide tometal, densification of silica, and elimination of organics. It wascarried out in a reducing atmosphere of 10% H2 at 600 �C for variousintervals of time to produce optimized fine cobalt nanoparticleswith silica shells.

2.1. Material characterization

The nanoparticles were characterized for structural, morpho-logical as well as magnetic properties at all stages. Morphology andthe particle size distribution were studied using TEM (JEOL 100CX-II transmission electron microscope, operated at 100 kV). The highresolution images were captured using FEI Tecnai F30 transmissionelectron microscope (HRTEM) with a TEM resolution limit of 1.7 A,operated at 200 kV. Sample preparation for all TEM imaging wascarried out by suspending nanoparticles in ethanol and dispensing

position of cobalt precursor in OA to form OA-coated CoO; Step 2) Room temperatureval of organics and silica compaction in at 600 �C for 4 h in 10% H2.

H. Sharma et al. / Materials Chemistry and Physics 175 (2016) 46e5348

two drops on carbon-coated grids (Ted Pella) and drying them inair. Structural information was obtained by X-ray diffraction (XRD-Philips 1813 diffractometer, Westborough, MA) while the magneticcharacterization was performed using BeH loops obtained with avibration sample magnetometer (VSM, Lakeshore 736 series,Westerville, OH) at room temperature (25 �C). Formagnetic studies,the samples were prepared as metal-polymer nanocomposite,where the silica-coated cobalt nanoparticles were dispersed inepoxy polymer with volume fraction of above 80%. The compositepowder was mechanically pressed as pellets, followed by polymercuring at 125 �C for 30 min. The silica shell on the cobalt nano-particles protected the metal core from oxidation at elevatedtemperatures during polymer curing.

3. Results and discussion

3.1. Role of Co(acac)2:OA ratio in CoO synthesis

The first part of this study was to optimize the cobalt precursor(cobalt acetylacetonate Co(acac)2) to organic ligand (Oleylamine)ratio, to yield finest metal core in a silica shell. Thermal decom-position of cobalt acetylacetonate (shown in step-1 in Fig. 1) atelevated temperatures of 220 �C for 30e40 min was carried out toform cobalt oxide seeds that were passivated with OA. The particlesize of these nucleated cobalt oxides were determined by theCo(acac)2:OA ratio. The ratio was thus changed from very lowpassivating agent content (1:10) to high content with ratio of 1:100.The morphological effect on the particle size of CoO is shown inFig. 2 (aed). As expected, the lower oleylamine concentration at1:10 yielded much larger CoO particle size of ~80 nm. The particlegrowth is attributed to Ostwald ripening at a higher reactiontemperature of 220 �C when the finer particles fuse with largerones due to low OA concentration. As the oleylamine concentrationis increased, tighter control on the particle size is achieved withincreased metal-ligand chelation that prevents particle fusion.Increasing the Co(acac)2:OA to 1:25 and then to 1:50 had showncontinuous drop in CoO particle size to 30 nm for 1:25 and 15 nmfor 1:50. However, further increase in the OA to 1:75 ratio, tendedto increase the particle size to 35 nm and thereby stabilizing thesize beyond OA concentration of 1:100 (Fig. 2(e)). It is believed thatvery high concentration of oleylamine leads to high steric hin-drance between the ligands which lead to ineffective chelation of

Fig. 2. TEM micrographs of CoO nanoparticles with different precursor to ligand ratio (a) 1:OA ratio.

oleylamine with the metal oxide particles leading to particlegrowth, as seen in case of 1:75 and 1:100. Literature describes thecorrelation between particle size and capping agent by statisticalmechanical model which quantifies the free-energies of nanoscalebonding (or particle aggregation) at different capping agent con-centrations [17]. Beyond a critical capping agent concentration, theaverage particle size increases with capping agent concentrationdue to increasingly stronger cross-linking interaction between thecobalt (nanoparticle) domains. These interactions influence boththe energy and entropy of the cluster formation and lower the freeenergy of the system, thereby favoring larger cluster (particle)formation.

The particle size variation as a function of precursor to passiv-ating agent ratio is summarized in Fig. 2(e), indicating lowest par-ticle size of CoO is obtained at critical ratio of 1:50.

In addition to the increase in particle size, the OA concentrationalso played a role in determining the crystallographic phase of CoOnanoparticles. At the precursor-to-ligand ratio of 1:75, CoO parti-cles were hexagonal in shape (shown in Fig. 2(c)) as compared tospherical shaped at all other Co(acac)2:OA ratio. A high resolutionTEM image, shown as an inset in Fig. 2(c), clearly shows six edges tothe hexagonal pyramidal CoO particle of size ~35 nm. These resultsare in accordance to literature [18] where it was shown that hex-agonal pyramid-shaped nanocrystals are predominant over nano-rods in bigger nanocrystals, which are favored by abruptdecomposition in higher metal precursor concentration. This phaseselectivity between hexagonal and cubic phase of CoO was alsoseen earlier but was attributed to the addition of water during thereaction that changes the reaction mechanism to yield differentcrystallographic oxide phases [19]. Other studies have shown thatthe metal oxide phases can be tuned by pressure-driven trans-formation or adding external reagent [20].

The crystallographic phases of the oxides were also confirmedby XRD and are shown in Fig. 3 (aed). CoO with starting ratio of1:10, 1:50 and 1:100 exhibited face-centered cubic (fcc) oxide forwhich Co2þ ions were octahedrally coordinated by the oxygenlattice (JCPDS No. 48-1719; space group, Fm3m) with peaksassigned to (111), (200), and (220) lattice plane. The ratio of 1:75 onthe other hand, clearly showed diffraction peaks corresponding tonanocrystalline hexagonal hcp phase of CoO characterized by theirsignature three peaks (100), (002), (101) at 2Ɵ of 30�e40�. Thecrystallite size was also calculated with XRD using Scherrer's

10; (b) 1:50; (c) 1:75 and (d) 1:100; (e) Effect of particle size as a function of Co(acac)2:

Fig. 3. XRD spectra of CoO from varied Co(acac)2:OA ratio (a) 1:10; (b) 1:50; (c) 1:75 and (d) 1:100.

H. Sharma et al. / Materials Chemistry and Physics 175 (2016) 46e53 49

formula and was found to be in good agreement with the TEM data(Table 1). A lowest particle size of ~15 nm (TEM) was obtained for aratio of 1:50 with spherical particles while the crystallite size forthe same ratio was calculated to be 7 nm from the broad peaks seenin XRD spectrum (Fig. 3(b)). The higher noise level observed inspectra corresponding to ratio 1:100 can be attributed to thickamorphous passivation layer of oleylamine that scatters the x-rayscausing the increase in noise level as seen in Fig. 3(d). The resultsobtained from XRD and TEM imaging for particle size and crystal-lographic phases are tabulated in Table 1, suggesting the optimizedcobalt precursor to oleylamine ratio of 1:50 to yield finest cubic CoOnanoparticles.

3.2. Optimizing the silica thickness for softer magnetic properties

The coreeshell cobalt/silica (Co/SiO2) structure was obtained byreducing the cobalt oxide in hydrogen atmosphere at 600 �C for3e4 h. A thin silica shell not only prevents particle sintering and

Table 1Consolidated morphology and particle sizes of CoO nanoparticles from different ratio, ob

Co:OA ratio XRD phase Crystallite size (XRD)

1:10 fcc 60 nm1:50 fcc 7 nm1:75 hcp 30 nm1:100 fcc 20 nm

growth during thermal reduction but also provides the muchneeded electrical insulation between the particles that therebyreduces the eddy currents in the final device. Thus, the thickness ofsilica shell depends on the CoO to silica precursor (TMOS) ratio andplays a critical role in determining the magnetic properties of thefinal product. Several ratios were studied to understand the metaloxide dispersion in the silica matrix and yield the thinnest andstable silica coat on the metal nanoparticles. The results of tworatios are presented here, Ratio 1: 0.05 M CoO in cyclohexane with1 ml TMOS; Ratio 2: 0.15 M CoO in cyclohexane with 1 ml TMOS.

On using a low ratio of CoO:TMOS (Ratio 1), a fluffy network ofsilica that embeds extra-fine cobalt metal particles (4e5 nm) isobtained, as shown in Fig. 4 (a, b). Literature reports that as high as42% reduction in metal volume can be achieved during thermalreduction of CoO to fcc Co [15,21] due to oxygen elimination athigher temperature. Although these individual fine cobalt nano-particles possess superior magnetic properties due to their shapeand size, their applications are limited since the interparticle

tained through XRD and TEM.

Morphology (TEM) Particle size (nm) (TEM)

Spherical and agglomerated 60e80Spherical and dispersed 15e20Hexagonal and dispersed 30e40Spherical and dispersed 25e35

Fig. 4. TEM micrographs showing extra-fine cobalt nanoparticles embedded in thick silica matrix.

H. Sharma et al. / Materials Chemistry and Physics 175 (2016) 46e5350

distance prevents any magnetic coupling between them and alsolowers the effective cobalt volume fraction. The high level ofmonodispersity of the cobalt nanoparticle is attributed to thehigher silica concentration that formed a porous but thicker(~30 nm) silica matrix during hydrolysis.

XRD of the reduced silica-coated cobalt nanoparticles was alsocarried out, revealing the cubic phasedmetallic cobalt nanoparticle,with assigned peaks to (111) and (200) plane of fcc-cobalt (JCPDSNo. 15-0806) [15]. The calculated crystallite size of ~8 nm was alsoin agreement with the TEM images (Fig. 4). The high noise level inthe XRD spectrum (Fig. 5) can be attributed to the large amount ofamorphous silica matrix that dispersed the x-ray during dataacquisition.

An optimized ratio of TMOS:CoO (Ratio 2, 0.15 M CoO in cyclo-hexane with 1 ml TMOS) was later employed to coat CoO nano-particles with thinner yet stable silica shell, shown in Fig. 6. HRTEMmicrographs showed cobalt cores mostly present as single crystalsof diameter ~8e10 nm. The particles were monodispersed withnarrow size distribution in amorphous silica matrix with inter-particle distance of 3e5 nm. Fig. 7(a) shows cluster of mono-dispersed cobalt nanoparticles that were mapped for elementaldistribution. Based on elemental mapping of cobalt (Fig. 7(b)), thediameter of the cobalt spheres was estimated to be 10 nm, whilethe thickness of the silica shell, whose signatures are obtained from

Fig. 5. XRD spectrum of reduced cobalt nanoparticle with a metal core of ~8 nm,showing fcc metal phase.

Si and O mapping (Fig. 7 (c&d), respectively) was much larger.Assuming a uniform silica coating on cobalt core, the thickness ofsilica around the core, is calculated to be ~ 5 nm.

3.3. Magnetic properties of Co/SiO2 nanoparticles

The static magnetic properties were studied withmagnetizationcurves measured using vibration sample magnetometry (VSM) atroom temperature and correlated to nanostructures. The Co/SiO2nanoparticles were mixed with polymer to form composites suchthat the metal-silica to polymer volume ratio was maintained at0.80:0.20 for all samples. In addition, for better comparison, allsamples were pressed so as to maintain same density of ~3 g/cc.Fig. 8(A) shows a hysteresis loop of cobalt nanoparticles with variedparticle sizes. As expected, with the increase in cobalt nanoparticlesize, saturation magnetization (Ms) was seen to increase from15emu/g for < 10 nm cobalt to 100emu/g for the largest particle inthis study (100 nm). The increase in Ms is attributed to the increasein the effectivemetal core volume fraction, compared to oxide shell,with larger particle size. The coercivity, in addition to Ms, alsoshowed strong dependence on particle size (Fig. 9). The magneticdata of cobalt nanoparticles with varied sizes is summarized inTable 2.

For the smallest particle size (<10 nm), a very low coercivity of20 Oe (Fig. 8(A), curve a) is achieved, indicative of exchangecoupling between the particles. To the best of the author's knowl-edge, such low coercivity has not been previously observed in solid-state and solution-derivedmetal nanoparticles. Furthermore, as theparticle size is increased (~15e30 nm), the coercivity increased asseen in curve b and c of Fig. 8(A). However, further increase inparticle size to 100 nm (curve d, Fig. 8(A)) leads to decrease incoercivity. This grain-size dependence of coercivity is explainedthrough Fig. 9 which follows the classical bell-shaped curve whichis well-understood from the literature [22] [24], [29].

With large particles of above 100 nm, multiple domains formwithin the particle. The number of domains within the particle isdependent on the magnitudes of magnetocrystalline energy, ex-change energy and saturation magnetization. The coercivity is lowin this case because the magnetization within each domain can bealigned to the other domain by easily rotating the domainwall withlower energy. With reduction in particle size to below 50 nm, thenumber of domains within the particle reduces to less than 5 [23].The effective field anisotropy is enhanced in nanoparticles becauseof dominating magnetocrystalline anisotropy with finer and fewerdomains. The dominating surface area further increases the coer-civity because of the surface anisotropy and ferromagnetic/anti-ferromagnetic coupling at the Co/CoO interface [24], [25].

Fig. 6. HRTEM of silica coated cobalt nanoparticles with ~10 nm metal core and ~5e8 nm silica shell.

Fig. 7. Nanoscale elemental mapping of (a) silica coated Co nanoparticle in STEM mode; (b) mapped cobalt (10 nm); (c, d) mapped Si and O (20 nm) overlaying the Co metal core.

H. Sharma et al. / Materials Chemistry and Physics 175 (2016) 46e53 51

The surface anisotropy (K) can be estimated by assuming mag-netic anisotropy (Hk) to be of the form:

Hk ¼ Hbulk þHsurface

D(1)

K ¼ Kbulk þKsurface

D(2)

Where, D is the metal particle size.Based on expressions (1) and (2) with finer particles the effec-

tive anisotropy increases because of enhanced surface area. Thehighest coercivity is obtained when the particles transition frommultiple domains to a single domain, also known as Stoner-wohlfarth domains [26]. In these dimensions, the particle cannotaccommodate even a single domain wall. Coercivity of magnetic

particles is dependent on several factors (1) magnetic anisotropywhich can originate from magnetocrystalline or shape anisotropy;and (2) demagnetization factor, which depend on the shape andaspect ratio of the particles, (3) magnetic anisotropy averaging frommultiple domains within the particle, (4)Thermal effects.

The coercivity rapidly reduces with size below a critical valuebecause of thermal effects and also from increased exchangecoupling between the particles which reduces the effective fieldanisotropy based on Herzer's theory [28]. The thermal effects cancause easy reversal of magnetization in this so-called super-paramagnetic limit. The magnetic properties of finer nanoparticlesdepend strongly on the interplay of energies between local mag-netic anisotropy and ferromagnetic exchange which are in turngoverned by ferromagnetic exchange length, Lex [27]. Due to ex-change coupling, the field effective anisotropy constant Keff reducesas 1/√N due to averaging, where N is the number of exchange

Fig. 8. Hysteresis loop for (A) Co/SiO2 nanoparticles with various particle sizes; a) sub-10 nm, b) 10e15 nm, c) 52e30 nm, d) > 100 nm, e) 140 nm; (B) Magnified BeH loop ofcomposite with <10 nm cobalt nanoparticles.

0 20 40 60 80 100 120 1400

30

60

90

120

150 Ms Coercivity

Particle size (nm)

Ms

(em

u/g)

0

100

200

300

400

500C

oerc

ivity

(Oe)

Fig. 9. Saturation magnetization and coercivity as a function of particle size in Co/SiO2

nanoparticles.

Table 2Magnetic properties of silica-coated cobalt nanoparticles with varying particle size.

Particle size (nm) Ms (emu/g) Hc (Oe) Hk (Oe)

<10 15 20 50010e15 25 140 170025e30 50 500 2200~100 80 150 4000140 120 20 5000

H. Sharma et al. / Materials Chemistry and Physics 175 (2016) 46e5352

coupled grains within the volume of (Lex)3 with grain size. How-ever, exchange coupling is presumably absent at these inter-particle dimensions, which inhibits further suppression incoercivity.

4. Conclusions

Silica-coated cobalt nanoparticles were synthesized using ahigh-temperature organometallic process to yield highly-monodispersed nanoparticles with narrow particle size distribu-tion. The effect of metal precursor to capping agent ratio on particlesize and morphology of cobalt nanoparticles was studied usinghigh-resolution TEM and XRD. Hexagonal and face-centered cubicphases of CoO were observed with varying Co(acac)2:OA ratio. An

optimized ratio of 1:50 yielded smallest cubic, particle size of<10 nm. High-density, monodispersed and spherical cobalt nano-particles (10 nm) were embedded in thin silica matrix to facilitateexchange coupling between the magnetic particles.

Magnetic properties were correlated to the structure, size anddistribution of the nanoparticles. The composite with sub-10 nmnanoparticles in superparamagnetic regime showed superior softmagnetic properties with as low coercivity as 20 Oe. With increasein the single-domain particle size, a more coercive nanocompositewas formed due to dominating ferromagnetic anisotropy,increasing the field for reversal magnetization. Beyond the criticalparticle size (40e50 nm) multi-domain structures are formedwithin the particle, leading to lowering of coercivity.

References

[1] P.M. Raj, P. Muthana, T.D. Xiao, L. Wan, D. Balaraman, I.R. Abothu,S. Bhattacharya, M. Swaminathan, R. Tummala, Magnetic nanocomposites fororganic compatible miniaturized antennas and inductors, in: AdvancedPackaging Materials: Processes, Properties and Interfaces, 2005. Proceedings.International Symposium on, 2005, pp. 272e275.

[2] G. Oldfield, T. Ung, P. Mulvaney, Au@ SnO2 coreeshell nanocapacitors, Adv.Mater. 12 (2000) 1519e1522.

[3] K. Ikeda, K. Kobayashi, K. Ohta, R. Kondo, T. Suzuki, M. Fujimoto, Thin-filminductor for gigahertz band with CoFeSiO-SiO 2 multilayer granular films andits application for power amplifier module, Magn. IEEE Trans. 39 (2003)3057e3061.

[4] M.K. Kazimierczuk, G. Sancineto, G. Grandi, U. Reggiani, A. Massarini, High-frequency small-signal model of ferrite core inductors, Magn. IEEE Trans. 35(1999) 4185e4191.

[5] T. Nakamura, Snoek's limit in high-frequency permeability of polycrystallineNieZn, MgeZn, and NieZneCu spinel ferrites, J. Appl. Phys. 88 (2000)348e353.

[6] H. Sharma, S. Jain, P.M. Raj, K. Murali, R. Tummala, Magnetic and dielectricproperty studies in Fe-and NiFe-based polymer nanocomposites, J. Electron.Mater. (2015) 1e8.

[7] M. Wu, Y. Zhang, S. Hui, T. Xiao, S. Ge, W. Hines, J. Budnick, G. Taylor, Mi-crowave magnetic properties of Co50/(SiO2) 50 nanoparticles, Appl. Phys. Lett.80 (2002) 4404e4406.

[8] S. Li, Q. Xue, H. Du, J. Xu, Q. Li, Z. Shi, X. Gao, M. Liu, T. Nan, Z. Hu, Large E-fieldtunability of magnetic anisotropy and ferromagnetic resonance frequency ofco-sputtered Fe50Co50-B film, J. Appl. Phys. 117 (2015) 17D702.

[9] H. Geng, J. Wei, Z. Wang, S. Nie, H. Guo, L. Wang, Y. Chen, G. Yue, D. Peng, Softmagnetic property and high-frequency permeability of [Fe 80 Ni 20eO/TiO 2]n multilayer thin films, J. Alloys Compd. 576 (2013) 13e17.

[10] Q. Dai, M. Lam, S. Swanson, R.-H.R. Yu, D.J. Milliron, T. Topuria, P.-O. Jubert,A. Nelson, Monodisperse cobalt ferrite nanomagnets with uniform silicacoatings, Langmuir 26 (2010) 17546e17551.

[11] Z. Nie, A. Petukhova, E. Kumacheva, Properties and emerging applications ofself-assembled structures made from inorganic nanoparticles, Nat. Nano-technol. 5 (2010) 15e25.

[12] S. Sun, Recent advances in chemical synthesis, self-assembly, and applicationsof FePt nanoparticles, Adv. Mater. 18 (2006) 393e404.

[13] E. Rabani, D.R. Reichman, P.L. Geissler, L.E. Brus, Drying-mediated self-assembly of nanoparticles, Nature 426 (2003) 271e274.

H. Sharma et al. / Materials Chemistry and Physics 175 (2016) 46e53 53

[14] S. Liu, Z. Zhang, M. Han, Gram-scale synthesis and biofunctionalization ofsilica-coated silver nanoparticles for fast colorimetric DNA detection, Anal.Chem. 77 (2005) 2595e2600.

[15] J.C. Park, H.J. Lee, H.S. Jung, M. Kim, H.J. Kim, K.H. Park, H. Song, Gram-scalesynthesis of magnetically separable and recyclable Co@ SiO2 yolk-shellnanocatalysts for phenoxycarbonylation reactions, ChemCatChem 3 (2011)755e760.

[16] Y. Han, J. Jiang, S.S. Lee, J.Y. Ying, Reverse microemulsion-mediated synthesisof silica-coated gold and silver nanoparticles, Langmuir 24 (2008) 5842e5848.

[17] Y. Wei, K.J. Bishop, J. Kim, S. Soh, B.A. Grzybowski, Making use of bondstrength and steric hindrance in nanoscale “synthesis”, Angew. Chem. Int. Ed.48 (2009) 9477e9480.

[18] W.S. Seo, J.H. Shim, S.J. Oh, E.K. Lee, N.H. Hur, J.T. Park, Phase-and size-controlled synthesis of hexagonal and cubic CoO nanocrystals, J. Am. Chem.Soc. 127 (2005) 6188e6189.

[19] J.H. Shim, K.M. Nam, W.S. Seo, H. Song, J.T. Park, The role of water for thephase-selective preparation of hexagonal and cubic cobalt oxide nano-particles, Chem.-An Asian J. 6 (2011) 1575.

[20] B.J. Morgan, P.A. Madden, Pressure-driven sphalerite to rock salt transition inionic nanocrystals: a simulation study, Nano Lett. 4 (2004) 1581e1585.

[21] J.C. Park, J.U. Bang, J. Lee, C.H. Ko, H. Song, Ni@ SiO 2 yolk-shell nanoreactorcatalysts: high temperature stability and recyclability, J. Mater. Chem. 20(2010) 1239e1246.

[22] B.D. Cullity, C.D. Graham, Introduction to Magnetic Materials, John Wiley &Sons, 2011.

[23] K.H.J. Buschow, F.R. Boer, Physics of Magnetism and Magnetic Materials,Springer, 2003.

[24] W.H. Meiklejohn, C.P. Bean, New magnetic anisotropy, Phys. Rev. 105 (1957)904.

[25] S.P. Gubin, Magnetic Nanoparticles, John Wiley & Sons, 2009.[26] D.L. Atherton, J. Beattie, A mean field Stoner-Wohlfarth hysteresis model,

Magn. IEEE Trans. 26 (1990) 3059e3063.[27] G. Herzer, Grain size dependence of coercivity and permeability in nano-

crystalline ferromagnets, Magn. IEEE Trans. 26 (1990) 1397e1402.[28] G. Herzer, Soft magnetic nanocrystalline materials, Scr. Metall. Mater. 33

(1995) 1741e1756.[29] A.G. Kolhatkar, A.C. Jamison, D. Litinov, R.C. Willson, T.R. Lee, Tuning the

magnetic properties of nanoparticles, Int. J. Mol. Sci. 14 (8) (2013)15977e16009.