IEEE TRANSACTIONS ON MAGNETICS, VOL. 42, NO. 10, OCTOBER 2006 3611

Magnetoelectric Properties of CoFe2O4-BaTiO3

Core-Shell Structure CompositeGiap V. Duong1;2, Roland Groessinger2, and Reiko Sato Turtelli2

Faculty of Chemical Engineering, Hanoi University of Technology, Hai Ba Trung, Hanoi, VietnamInstitute of Solid State Physics, Vienna University of Technology, A-1040 Vienna, Austria

The CoFe2O4-BaTiO3 core-shell structure composite has been successfully synthesized by wet chemical method. X-ray characteri-zation showed that the composite consisted of two single phases: CoFe2O4 and BaTiO3. The saturation magnetization of the CoFe2O4

component in composite was found to be similar to those of the bulk sample. It was observed that the longitudinal and transverse mag-netoelectric coefficients of the core-shell structure are 18 times higher than those of the mixed structure.

Index Terms—Barium titanate, cobalt ferrite, magnetoelectric effect, multiferroics.

I. INTRODUCTION

MAGNETOELECTRIC (ME) materials become magne-tized when placed in an electric field and electrically

polarized when placed in a magnetic field. Thus an effectiveconversion between electric and magnetic energy becomespossible. They can be realized as single phase and compositematerials. The number of single phase ME materials is limiteddue to the primary requirement that magnetic and electricdipoles have to coexist in an asymmetric structure. The MEcoefficient, , where is the induced elec-trical field in an applied magnetic field , in single phasematerials is small, the working temperature is low and theyinvole expensive materials and processing techniques. Theselimitations can be overcome when shifting to composites, whichusually consist of magnetostrictive and piezoelectric phases.Multilayer or laminate composites of ferrite and piezoelectricthick layers show a large ME effect [1]–[3]. The ME effectin such structure originates from “product properties” of theconstituent phases and can reach a maximum ME coefficient of5900 mV cm Oe [1]. However, in the CoFe O -BaTiOsystems, the ME coefficients vary in a wide range, from 0.19 in[4] to 130 mV cm Oe in [5]. The highest value is for the(BaTiO ) -[(CoFe O ) -(CoTiO ) ] composite(eutectic composition with alternate layers prepared by unidi-rectional solidification method). For other composition, the MEcoefficients vary from 1–4 mV cm Oe . Differences in thesample characteristics (constituent, composition, microstruc-ture, size, impedance, etc.), measuring techniques (static,quasistatic, pulse, resonant, etc.) and measuring conditions(input impedance of instruments, field direction, frequencyof the AC field, etc.) make a conclusion what is the highestpossible ME effect for a certain material up to now impossible.

The motivation of this work is to synthesize aCoFe O -BaTiO core-shell structure composite and tostudy its ME properties. The core-shell structure was chosenbelieving that this can cause a better coupling between the twophases. CoFe O and BaTiO were chosen not only due to its

Digital Object Identifier 10.1109/TMAG.2006.879748

good magnetic and piezoelectric properties but also its highelectrical resistance that can help to prevent the dischargingprocess during the measurements. Additionally, its chemicaland mechanical stability and nontoxic properties are importantfor environment and health in applications.

II. EXPERIMENT

The sample was synthesized using wet chemical method.All chemicals are in analytic pure grade. CoFe O powder ob-tained from appropriate solution containing Co(NO ) .6H Oand Fe(NO ) .9H O and co-precipitating at 75 C using NaOHsolution. To obtain CoFe O -BaTiO core-shell structure com-posite, the co-precipitated CoFe O powder, calculated to geta composite containing 50% in mass of each constituent, wasintroduced into a solution of acetic acid, barium hydroxideand titanium (IV) n-Butoxide, which is then gelated on thesurface of the CoFe O grains or particles during heatingand stirring. The obtained material is dried, pre-sintered at700 C for 2 h, ground into fine powder, then pressed under apressure of 6 tones/cm to obtain a sample in shape of a disc of10 mm in diameter and 1.5 mm thick. This sample is sintered at1250 C for 12 h to get a core-shell structure composite. Afterheat treatment, due to shrinkage, the diameter and thicknessof the sample are 8.7 and 1.3 mm, respectively. The samplewas painted by silver paste for electric contacts, then poledelectrically under an electric field of 7500 V/cm (field directionis perpendicular to the surface of the sample) in silicon oilfrom 150 C down to room temperature. For comparison, asample with similar composition was produced by mixing theCoFe O and BaTiO powders following by the same heattreatment and poling procedure.

The crystalline structures of single phase materials as well ascomposites were investigated by X-ray diffractometer (XRD)using Co-K radiation. The hysteresis loops were measured atroom temperature using a pulse field magnetometer which gen-erates field up to 5 T with a typical pulse duration of 50 ms. Themagnetostriction was measured using strain gauge method witha 50 kHz bridge (HBM Type KWS 85.A1).

The ME coefficient was measured using lock-in techniqueemploying an AC field with frequencies from 1 Hz to 4 kHz

0018-9464/$20.00 © 2006 IEEE

3612 IEEE TRANSACTIONS ON MAGNETICS, VOL. 42, NO. 10, OCTOBER 2006

Fig. 1. Schema of the longitudinal and transverse ME coefficientmeasurements.

and amplitudes from 0.1 Oe to 20 Oe superimposed onto a DCmagnetic field (generated by an electromagnet) up to 6.5 kOe,

[6]. The input resistance and capacitance of the lock-in(EG&G model 5210) is 100 M and 25 pF, respectively. Themeasurements of the ME voltage were performed for twodifferent field orientations with respect to the sample plane toobtain the transverse and longitudinal ME coefficients. Theabsolute values of the voltage may depend on the ratio be-tween the resistances of the samples and the lock-in amplifier.The ME coefficient, , was determined using the equation

H) , where is the voltagemeasured with the lock-in amplifier, is the effective thicknessof the piezoelectric phase (in this work it was considered 47%of the thickness of the sample since the density of CoFe Oand BaTiO are 5.3 and 6.08 g/cm , respectively) and isthe amplitude of the applied AC field [6]. The schema showingthe ME transverse ( perpendicular to ) and longitudinal( parallel to ) measurements are represented in Fig. 1.All measurements were carried out at room temperature andambient pressure.

III. RESULTS AND DISCUSSIONS

The XRD patterns in Fig. 2 suggest that the compositesconsist of two single phases only: CoFe O and BaTiO .The particle size of the as-coprecipitated CoFe O evaluatedusing Scherrer equation [7] for the full-width at half maximumis about 10 nm. The saturation magnetization and coercivefield of the as-coprecipitated CoFe O is of 53 emu/g and310 Oe, respectively. However, the saturation magnetization ofthe magnetostrictive component (CoFe O ) in the core-shellstructure composite increases to 72 emu/g and the coercive fieldof the composite is similar to that of co-precipitated ferrite.This increase of the saturation magnetization may be attributedto the less spin canted surface layers due to heat treatment.To have a comparison, the bulk CoFe O prepared by sol-gelmethod presents a saturation magnetization of 78 emu/g, but acoercivity of 825 Oe.

Figs. 3 and 4 show the longitudinal and transverse MEcoefficients as function of the bias field, , measuredon the core-shell structure and mixed structure composites,respectively, under an AC field of 10 Oe with a frequency of

Fig. 2. XRD patterns of CoFe O (spinel structure), BaTiO (perovskitestructure) and CoFe O -BaTiO composite.

Fig. 3. Bias field dependent of � measured on CoFe O -BaTiO core-shellstructure composite.

Fig. 4. Bias field dependent of � measured on CoFe O -BaTiO mixedstructure composite.

270 Hz. The curves show hysteresis, remanenceand maximum, where the values and positions of the maximadepend on the samples and field orientations. As the bias fieldis increased from zero, in core-shell composite, for the longitu-dinal vs. curve, the maximum of occurs at 2.2 kOe

DUONG et al.: MAGNETOELECTRIC PROPERTIES OF CoFe O -BaTiO CORE-SHELL STRUCTURE COMPOSITE 3613

Fig. 5. Hystersis loops of CoFe O component measured parallel andperpendicular to the sample plane.

Fig. 6. Magnetostriction of CoFe O and CoFe O -BaTiO core-shellstructure composite at room temperature.

with a value of 3.4 mV cm Oe , and in the transverse case,at 1.4 kOe with a value of 2.0 mV cm Oe . These valuesare smaller compared to those in [5] (130 mV cm Oe forthe (BaTiO ) -[(CoFe O ) -(CoTiO ) ] compo-sition) but bigger than those in [4] (0.19 mV cm Oe forCoFe O -BaTiO sintered composite). The reason may be dueto the difference in the sample compositions and the actualmicrostructures.

The bias field where the maximum occurs in case of the lon-gitudinal is lager than the transverse case. The reason is: the de-magnetizing field in the longitudinal case (magnetic field par-allel to the sample plane) is bigger than that of the transverse(magnetic field perpendicular to the sample plane), which canbe seen clearly in Fig. 5.

As reported in [3], the tracks roughly the strengthof piezomagnetic coupling where is the magne-tostriction of the ferrite. As can be seen in Fig. 6 which shows thefield dependence of is the longitudinal magnetostrictionand is transverse magnetostriction), the maximum

occurs at around 1–3 kOe, where the maximum in themagnetization curve occurs, too (see magnetic hysteresis loopsin Fig. 5). The longitudinal and transverse are almost 18times higher as compared to those found in the mixed structurecomposite (see Fig. 4). Our results obtained in the mixed struc-ture composite are in agreement with the results reported in [4].One can deduce that the coupling between the two phases in thecore-shell structure is much better than in the mixture. In thiswork, in both cases, the longitudinal ME coefficient is biggerthan the transverse one. This is different from results reportedin literature [3]: the transverse ME coefficient is bigger than thetransverse one. However, our results actually suggest that the

curve follows the strength of the dependent of ,that is, , see Fig. 6. The decrease in ME coefficientat higher bias field may be attributed to the saturated state of themagnetization and total magnetostriction that weaken the piezo-magnetic coupling.

IV. CONCLUSSION

Composite with core-shell structure of 50% CoFe O and50% BaTiO in mass has been successfully synthesized. Themagnetoelectric properties of the composite have been studiedat room temperature and ambient pressure. The coupling be-tween the magnetostricve and piezoelectric phases in the core-shell structure composite is much better than that in the mixedstructure, resulting in a ME coefficient which is about 18 timesbigger in the former as compared to those in the later. In ourwork, vs. curve tracks roughly the strength of piezo-magnetic coupling .

ACKNOWLEDGMENT

This work was supported by the ÖAD (Technology GrantSoutheast Asia) and FWF Project P16500-N02.

REFERENCES

[1] J. Ryu, S. Priya, K. Uchino, and H.-E. Kim, “Magnetoelectric effect incomposites of magnetostrictive and piezoelectric materials,” J. Electro-ceramics, vol. 8, pp. 107–119, 2002.

[2] M. Fiebig, “Revival of the magnetoelectric effect,” J. Phys. D, Appl.Phys., vol. 38, pp. R123–R152, 2005.

[3] G. Srinivasan, E. T. Rasmussen, and R. Hayes, “Magnetoelectric effectsin ferrite-lead zirconate titanate layered composites: The influence ofzinc substitution in ferrites,” Phys. Rev. B, vol. 67, p. 014 418, 2003.

[4] R. P. Mahajan, K. K. Patankar, M. B. Kothale, S. C. Chaudhari, V. L.Mathe, and S. A. Patil, “Magnetoelectric effect in cobalt ferrite-bariumtitanate composites and their electrical properties,” Pramana-J. Phys.,vol. 58, p. 1115, 2002.

[5] J. Van Suchetelene, Philips Res. Rep., vol. 27, p. 28, 1972.[6] G. V. Duong, R. Groessinger, M. Schoenhart, and D. Bueno-Basques,

“The lock-in technique for studying magnetoelectric effect,” J. Magn.Magn. Mater., to be published.

[7] H. G. Jiang, M. Rühle, and E. J. Lavernia, “On the applicability of theX-ray diffraction line profile analysis in extracting grain size and mi-crostrain in nanocrystalline materials,” J. Mater. Res., vol. 14, no. 2, pp.544–549, 1999.

Manuscript received March 13, 2006 (e-mail: giap@ifp.tuwien.ac.at).