Journal of Magnetism and Magnetic Materials 329 (2013) 71–76

Contents lists available at SciVerse ScienceDirect

Journal of Magnetism and Magnetic Materials

0304-88

http://d

n Corr

Canal U

fax: þ2

E-m

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

Structural and magnetic properties of cadmium substituted manganeseferrites prepared by hydrothermal route

Nasser Y. Mostafa a,b,n, Z.I. Zaki a,c, Z.K. Heiba a,d

a Faculty of Science, Taif University, P.O. Box: 888, Al-Haweiah, Taif, Saudi Arabiab Chemistry Department, Faculty of Science, Suez Canal University, Ismailia 41522, Egyptc Advanced Materials Division, Central Metallurgical R&D Institute (CMRDI), P.O. Box: 87 Helwan, Cairo, Egyptd Physics Department, Faculty of Science, Ain Shams University, Cairo, Egypt

a r t i c l e i n f o

Article history:

Received 5 June 2012

Received in revised form

6 August 2012Available online 16 October 2012

Keywords:

MnFe2O4

Cd-substitution

Hydrothermal

Magnetization

Ferrite

53/$ - see front matter Crown Copyright & 2

x.doi.org/10.1016/j.jmmm.2012.09.004

esponding author at: Chemistry Departmen

niversity, Ismailia 41522, Egypt. Tel.: þ20 64

0 64 322381.

ail address: nmost69@yahoo.com (N.Y. Mosta

a b s t r a c t

Cd-substituted manganese ferrite Mn1�xCdxFe2O4 powders with x having values 0.0, 0.1, 0.3 and

0.5 have been synthesized by hydrothermal route at 180 1C in presence of NaOH as mineralizer.

The obtained ferrite samples were characterized by X-ray diffraction (XRD), transmission electron

microscope (TEM) and vibrating sample magnetometer (VSM). The XRD analysis showed that pure single

phases of cubic ferrites were obtained with x upto 0.3. However, sample with xZ0.5 showed hexagonal

phase of cadmium hydroxide (Cd(OH)2) besides the ferrite phase. The increase in Cd-substitution upto

x¼0.3 leads to an increase in the lattice parameter as well as the average crystallite size of the prepared

ferrites. The average crystallite size increased by increasing the Cd-content and was in the range of

39–57 nm. According to VSM results, the saturation magnetization increased with Cd ion substitution.

Crown Copyright & 2012 Published by Elsevier B.V. All rights reserved.

1. Introduction

Nanocrystalline spinel ferrites have been investigated inten-sively in recent years due to their potential applications in non-resonant devices, radiofrequency circuits, high quality filters, rodantennas, transformer cores, read/write heads for high-speeddigital tapes and operating devices [1–6]. Manganese ferrite,MnFe2O4, is a well-studied ferrite that has low magnetic aniso-tropy at room temperature arising from the low magnetocrystal-line anisotropy energy common to cubic magnetic structures [7].Manganese ferrite is a partial inverse spinel ferrite, where about20% of the Mn2þ ions are on the B sites (octahedrally coordinatedsites) and the rest reside on the A sites (tetrahedrally coordinatedsites) [8]. The remaining of the A and B sites are occupied by theFe3þ cations. Magnetic properties of ferrites can be suitablytailored by varying composition of cations. The cation type,occupancy, and valence determine the magnetic and electronicproperties of this important class of materials doping of ferritewith small amount of nonmagnetic ions such as Zn2þ or Cd2þ

results in the increase of saturation magnetization [9].Various methods [10–16] such as mechanical milling, inert

gas condensation, hydrothermal reaction, oxidative precipitation,

012 Published by Elsevier B.V. All

t, Faculty of Science, Suez

382216;

fa).

sol–gel synthesis and reverse micelle technique are employed forthe preparation of ferrites nanoparticles. It was reported that atelevated temperatures (200–1000 1C), MnFe2O4 was unstable inair and Mn2þ ions on the surface oxidize to form Mn3þ ionsresulting in the dissociation of the ferrites [17]. Therefore pre-paration methods involving calcination steps are not suitable forthe preparation of manganese ferrite nanoparticles. On the otherhand, ferrites were prepared via the hydrothermal method attemperatures ranging from 150 to 200 1C without the need ofhigh processing temperature or calcination steps [13]. In recentyears, commercial interest in hydrothermal synthesis has beenrevived [18]. The main fascinated advantage of hydrothermalsynthesis is the significant improvement in the chemical activityof the reactant [19,20]. The shape, size distribution and crystal-linity of the final product can be precisely controlled throughadjusting the reaction parameters such as temperature, time,solvent type, surfactant type and precursor type [18–21].

Substitution of magnetic and non-magnetic cations in ferritematerials changes their magnetic and electrical properties[22–24]. The changes in physical and chemical properties of ionsubstituted spinel ferrites arise from the ability of these com-pounds to distribute the cations between the available A-sites(tetrahedral) and B-sites (octahedral) [25–27]. Several cationshave been used by many researchers in order to improve theelectrical and magnetic properties of manganese ferrites [28–30].Reviewing the literature [31–35], one can assume that cadmiumsubstitution modify the electric and magnetic properties of

rights reserved.

N.Y. Mostafa et al. / Journal of Magnetism and Magnetic Materials 329 (2013) 71–7672

ferrites considerably. Due to its larger ionic radius, the occupancyof Cd2þ ions into the spinel lattice would possibly create a latticedistortion [35] and modification of all the material properties toan appreciable extent could be expected. The electronic config-uration of Cd2þ ion with free 5s5p orbitals permits it to formcovalent bonds with oxygen ions (sp3 hybridization) and predis-pose them to occupy tetrahedral sites only [33].

In the present investigation, we report a simple hydrothermalroute to synthesize pure MnFe2O4 ferrite and Cd-substitutedMnFe2O4 ferrite. Moreover, the effect of Cd-substitution on theformation and the magnetic properties of Mn1�xCdxFe2O4 parti-cles (with x varying from 0.1 to 0.3) was also investigated.

Fig. 1. XRD patterns of Mn1�xCdxFe2O4 synthesized by hydrothermal route

(cadmium hydroxide (Cd(OH)2); JCPDS no. 01-073-0969).

Table 1Relative intensities of Cd(OH)2 peaks in sample and in JCPDS cards.

Peak 2H I% (JCPDS) I% (sample)

(001) 18.858 93.8 100

(100) 29.479 58 2.69

(011) 35.228 100 2.5

(002) 38.252 6.4 10.23

(110) 52.294 20.7 0.63

2. Materials and methods

Ferric chloride (FeCl3), manganese chloride (MnCl2 �4H2O),cadmium chloride (CdCl2 �4H2O), and sodium hydroxide (NaOH)of analytical grade reagents were used. In a typical synthesisprocess, 80 mmol ferric chloride and 40 mmol manganese chlor-ide and cadmium chloride were mixed, ground uniformly, dis-solved in 350 ml distilled water with the assistance of magneticstirring. The pH was adjusted to 12 with NaOH solution. Thevolume was completed to 400 ml with distilled water andtransferred into a 600 ml teflon-lined autoclave then heated at180 1C for 20 h. Magnetic stirrer was used within the teflon-linedautoclave to increase the homogeneity of the producedCd-substituted ferrite powder. The compositions of the Cd-substituted samples were represented by Mn1�xCdxFe2O4 with x

having values 0.0, 0.1, 0.3 and 0.5. The final product was obtainedafter being washed and filtered several times with distilled waterand anhydrous ethanol, then dried at 50 1C in vacuum.

X-ray diffraction (XRD) analysis was performed using anautomated diffractometer (Philips type: PW1840), at a step sizeof 0.021, scanning rate of 21 in 2y/min, and a 2y range from 41 to801. The values of full width at half-maximum (FWHM) of thepeak of the (311) plane was used to calculate the crystallite sizesaccording to Scherrer’s formula.

D¼ Kl=½b1=2cos y� ð1Þ

where D is the crystallite size in angstrom, K is Scherrer’s constant(0.89), l is the wavelength of X-rays beam (1.5405 A), y is thediffraction angle for the reflection (311) and b1/2, is defined as thediffraction full width at half-maximum (FWHM), expressed inradians. The FWHM was extracted using X’Pert HighScore Plusprogram [36]. Determination of the lattice constants were madeby least squares refinement of the X-ray diffraction data. Indexingof the powder patterns and least squares fitting of the unitcell parameters was possible using the software X’Pert HighScorePlus [36].

The powders morphology was investigated using SEM (JOEL,Model: JSM-5600, Japan) equipped with secondary electrondetector and EDX. All samples were coated with gold. The shapeand particle size distribution were studied using transmissionelectron microscope operated at 120 kV accelerating voltage(JTEM-1230, Japan, JEOL). The samples were prepared by makinga suspension from the powder in distilled water using ultrasonicwater bath. Then a drop of the suspension was put into the carbongrid and left to dry. The magnetic properties of the specimenswere measured by a PAR vibrating sample magnetometer (VSM)at room temperature in a maximum field of 10 kOe.

3. Results and discussion

Fig. 1, shows the XRD patterns of Cd-substituted Mn-ferritepowders prepared by hydrothermal route. Phase analysis of the

obtained diffraction patterns ensured the single-phase cubicspinel structure for all Cd-substituted MnFe2O4 samples with x

values upto 0.3. Only, the diffraction peaks corresponding to(220), (311), (221), (400), (422), (511) and (440) planes ofMnFe2O4 are observed (JCPDS no. 10-0319). Samples with x¼0.5shows very strong peaks, marked as ‘*’, corresponding to the linesof hexagonal phase of cadmium hydroxide; Cd(OH)2 (JCPDS no.01-073-0969) with very weak peaks of MnFe2O4. This means thatthe hydrothermal process resulted in complete conversion ofreactants to yield Mn1�xCdxFe2O4 as a single phase for xr0.3without any calcinations processes. It was reported that manga-nese ferrite is unstable in air at temperatures 200–1000 1C andpartially dissociated into a–Fe2O3 and Mn2O3 [17]. Thus, thehydrothermal route is a promising process for the synthesis of asingle and pure phase of manganese ferrite. This is an advantageover the other wet chemical routes which need calcination steps.

Table 1, summarizes the relative intensity of Cd(OH)2 peaks ofsample Mn0.5Cd0.5Fe2O4 and those standard Cd(OH)2 (JCPDS no.01-073-0969). Comparing the obtained relative intensities withthose of JCPDS card, the 100% reflection becomes the (001)instead of the (011) reflection. The increase in intensity of (001)

N.Y. Mostafa et al. / Journal of Magnetism and Magnetic Materials 329 (2013) 71–76 73

peak of Ca(OH)2 (100%) compared with that (93.8%), is attributedto the preferred orientation along this crystal plane. This pre-ferred orientation along the [001] direction results from pressing

Table 2

Lattice parameter (a A) of Mn1�xCdxFe2O4 solid solutions.

Sample x Lattice constant (a A) Crystallite size (nm)

MnFe2O4 0.0 8.49 39.8

Mn0.9Cd0.1Fe2O4 0.1 8.51 46.7

Mn0.7Cd0.3Fe2O4 0.3 8.52 57.5

Table 3

Ionic radii (A) of different metal cations in A-sites (tetrahedral) and B-sites

(octahedral) [38].

Metal cation A-sites (tetrahedral) B-sites (octahedral)

Mn2þ 0.66 0.83

Fe3þ 0.49 0.55

Cd2þ 0.83 0.95

Fig. 2. (a) SEM micrograph and (b) EDX of manganese ferrite (MnFe2O4) powder

prepared by hydrothermal route.

the powders with microsheets morphology in XRD holder(see TEM Fig. 6). The high relative intensity of (00l) with respectto other reflections in the XRD patterns suggested that theobtained Cd(OH)2 sheets grow in both the a-axis and the b-axisdirections [37].

The lattice parameters for the cubic Mn1�xCdxFe2O4 spinelpowders with 0.0rxr0.3 were calculated. The variation oflattice parameter ‘a’ as a function of Cd concentration ‘x’ is shownin Table 2. The lattice constant increases with increasing cad-mium concentration, which can be explained based on therelative ionic radius. The ionic radius (0.78 A) of Cd2þ ions islarger than the ionic radius (0.66 A) of Mn2þ ions. Replacement oflarger Cd2þ ion for smaller Mn2þ ions in the manganese ferritecauses an increase in lattice constant.

The increase in the lattice parameter with low level of Cd-substitutions (xo0.3) can be understood on the basis of averagetetrahedral ionic radii of Mn2þ ions and Cd2þ ions, as summar-ized in Table 3 [38]. From Vegard’s law, if the radius of displacingion is larger than the displaced ion, the lattice constant increases.It is well known that the tetravalent Cd2þ ions have a strongtetrahedral (A) sites preference [39]. The octahedral ionic radius

Fig. 3. (a) SEM micrograph and (b) EDX of manganese ferrite (Mn0.9Cd0.1Fe2O4)

powder prepared by hydrothermal route.

N.Y. Mostafa et al. / Journal of Magnetism and Magnetic Materials 329 (2013) 71–7674

of Cd2þ ion is 0.61 A, which is greater than that of Mn2þ (0.55 A),hence the unit cell expands and the lattice constant increaseswith the increase of Cd2þ ions content up to x¼0.3.

A clear change is observed in FWHM of XRD diffractograms,which is reflected in the calculations of the crystallite sizes.Table 2, shows the variation of the average crystalline sizes forthe cubic spinel phase. It was found that the average crystallitesize increased from 39.8 nm for the sample without Cd substitu-tion to 57.5 nm for the sample substituted with x¼0.3. Theseresults are consistent with the results obtained for the latticeparameters, where increasing the lattice parameter may results inincrease in the total crystallite sizes. Similar results were reportedfor other spinel ferrites [13] and PZT ferroelectric materials[21,40–41].

The microstructures of the as prepared manganese ferritesample without substitution and with Cd-substitution (uptox¼0.3) have been examined by SEM as shown in Figs. 2–4.It can be observed that the pure ferrite powders (Fig. 2a) arequasi-spherical agglomerated particles with a quite uniform sizedistribution. Fig. 3a, shows SEM of Cd-substituted sample(x¼0.1). The chemical compositions of the present ferrite sampleswere determined by energy dispersive X-ray spectroscopy (EDX)shown in Figs. 2–4. The EDX spectra of the ferrite samples show

Fig. 4. (a) SEM micrograph and (b) EDX of manganese ferrite (Mn0.7Cd0.3Fe2O4)

powder prepared by hydrothermal route.

the presence of elements (Mn, Cd, Fe and O) without impuritieswhich indicate the completeness of solid state reaction. Fig. 5shows SEM microstructure of the as prepared manganese ferritesample with x¼0.5. The figure clearly shows the large crystals ofhexagonal phase of cadmium hydroxide. The microstructures ofthe as prepared manganese ferrite sample (MnFe2O4) and withCd-substitution (x¼0.3) have been examined by TEM as shown inFig. 6. It can be observed that the pure ferrite powders (Fig. 6(a))were cubic particles with a quite uniform size distribution.Fig. 6(b), shows TEM of Cd-substituted sample (x¼0.3). Theparticles are also cubic with slight large crystallite sizes. Theseresults are consistent with the results of crystallite sizes obtainedfrom XRD data. Fig. 6(c) shows TEM of a selected area of sampleMn0.5Cd0.5Fe2O4. The formation of hexagonal microsheet ofCd(OH)2 is responsible for the preferred orientation that appearsin XRD.

The specific hysteresis loop and saturation magnetization (MS),of the as prepared Mn1�xCdxFe2O4 with x¼0.0 to 0.3 powders aremeasured at room temperature in a maximum field of 10 kOe. Thehysteresis loops are shown in Fig. 7. In general, the producedferrites are soft magnetic material due to the deviation fromrectangular form and their very low coercivity. The saturationmagnetization MS of the prepared ferrites is strongly dependent

Fig. 5. (a) SEM micrograph and (b) EDX of manganese ferrite (Mn0.5Cd0.5Fe2O4)

powder prepared by hydrothermal route.

Fig. 6. TEM micrograph of: (a) manganese ferrite, (b) Cd-substituted manganese

ferrite (x¼0.3) and (c) Cd(OH)2 produced in sample Mn0.5Cd0.5Fe2O4.

-10000 -5000 0 5000 10000

-40

-30

-20

-10

0

10

20

30

40 MnFe O Mn Cd Fe O Mn Cd Fe O

Mag

netiz

atio

n (e

m/g

)

Field (Oe)

Fig. 7. Effect of Cd ion concentration on the M–H hysteresis loop of Mn1�x

CdxFe2O4 powders synthesized by hydrothermal route.

N.Y. Mostafa et al. / Journal of Magnetism and Magnetic Materials 329 (2013) 71–76 75

on the Cd2þ ion concentration. The MS of the Cd-substitutedMn-ferrite powders, as shown in Fig. 7, increased with increasingCd2þ ion concentration upto x¼0.3. The MS increases from31.08 emu/g for the unsubstituted MnFe2O4 to 33.88 emu/g forMn0.9Cd0.1Fe2O4; and slightly increases to 34.46 emu/g forMn0.7Cdi0.3Fe2O4 ferrite. The magnetization of Cd-substitutedMnFe2O4 can be explained on the basis of Neel’s molecular fieldmodel and cations distribution between A-sites and B-sites.According to this model, A–B interaction is stronger and moreeffective than A–A and B–B interactions. The net magneticmoment of the lattice is given by the vector sum of magneticmoments of A and B sublattices M(x)¼MB(x)�MA(x), where MB

and MA are the magnetic moment at the octahedral and tetra-hedral sites respectively. Generally, MnFe2O4 prepared byconventional solid state method is known to be 20% inversespinel [17]. Thus, its structure can be represented by Mn0.8Fe0.2

[Mn0.2Fe1.8]O4; where, ions in square bracts occupy B-sites.According to Neel’s model discussed above, increasing the occu-pancy of B-sites with magnetic ions (Fe3þ or Mn2þ) increases MS.However, increasing the occupancy of A-sites with magnetic ionsdecreases MS.

The increase in saturation magnetization with low level Cd-substitution (x¼0.1) is due to the substitution of diamagneticCd2þ ions in the A-sites. Thus Cd-substitution at these low levelsdoes not affect the occupancy of Fe3þ in B-sublattice. However, itdecreases the occupancy of Mn2þ in A-sublattice, because enter-ing one Thus, the charge compensation mechanism forces Mn2þ

to expel from the A-sublattice leading to slight overall increase inMS values.

4. Conclusion

Cd2þ ions substitution manganese ferrite particles with dia-meters ranging from 39 to 47 nm have been synthesized by thehydrothermal method at 180 1C. At low substitution level(x¼0.1), the Cd2þ ions enter the spinel lattice on the tetrahedralA-sites. This leads to an increase in both the lattice parameter andthe saturation magnetization. As the substitution level increasedto x¼0.3, the Cd2þ ions start to substitute Mn2þ and Fe3þ ions inthe A-sublattice and B-sublattice, respectively, which results inlattice parameter increase. However, the saturation magnetiza-tion slightly increased due to the dilution of Fe3þ in B-sublattice.

Acknowledgments

This work was financially supported by a grant from TaifUniversity (Grant no. 1105-432-1).

References

[1] P. Ravindernathan, K.C. Patil, Journal of Materials Science 22 (1987) 3261.[2] H. Igarash, K. Ohazaki, Journal of the American Ceramic Society 60 (1997) 51.[3] A. Goldman, American Ceramic Society Bulletin 63 (1984) 582.[4] D. Stoppels, Journal of Magnetism and Magnetic Materials 160 (1996) 323.[5] K.H. Lee, D.H. Cho, S.S. Jeung, Journal of Materials Science Letters

16 (1997) 83.[6] C.S. Kim, Y.S. Yi, K.T. Park, H. Namgung, J.G. Lee, Journal of Applied Physics 85

(1999) 5223.

N.Y. Mostafa et al. / Journal of Magnetism and Magnetic Materials 329 (2013) 71–7676

[7] Xu Zuo, Aria Yang, Soack-Dae Yoon, Joseph A. Christodoulides,G.Harris Vincent, Carmine Vittoria, Applied Physics Letters 87 (2005) 152505.

[8] Xu Zuo, Aria Yang, Soack-Dae Yoon, Joseph A. Christodoulides, G.Harris Vincent,Carmine Vittoria, Journal of Applied Physics 97 (2005) 10G103.

[9] B. Viswanathan, V.R.K. Murthy, Ferrite Materials, Narosa Publications, 1990,p. 11.

[10] I.H. Gul, A. Maqsood, Journal of Alloys and Compounds 465 (2008) 227.[11] P.P. Sarangi, S.R. Vadera, M.K. Patra, N.N. Ghosh, Powder Technology 203

(2010) 348.[12] U. Ghazanfar, S.A. Siddiqi, G. Abbas, Materials Science and Engineering B 118

(2005) 132.[13] N.Y. Mostafa, M.M. Hessien, A.A. Shaltout, Journal of Alloys and Compounds

529 (15) (2012) 29.[14] S.D. Sartale, C.D. Lokhande, Indian Journal of Engineering and Materials

Science 7 (2000) 404.[15] G.A. El-Shobaky, A.M. Turky, N.Y. Mostafa, S.K. Mohamed, Journal of Alloys

and Compounds 493 (2010) 415.[16] N.I. Aljuraide, M.A.A. Mousa, N.Y. Mostafa, G.A. El-Shobaky, H.H. Hamdeh,

M.A. Ahmed, International Journal of Nanoparticles 5 (2012) 56.[17] E.D. Macklen, Journal of Inorganic and Nuclear Chemistry 30 (1968) 2689.[18] W.L. Suchanek, R.E. Riman, Advances in Science and Technology 45 (2006)

184.[19] N.Y. Mostafa, E.A. Kishar, S.A. Abo-El-Enein, Journal of Alloys and Compounds

473 (2009) 538.[20] N.Y. Mostafa, A.A. Shaltout, H. Omar, S.A. Abo-El-Enein, Journal of Alloys and

Compounds 467 (2009) 332.[21] S. Kumar, R. Kumar, S.K. Sharma, V.R. Reddy, A. Banerjee, Alimuddin, Solid

State Communications 142 (2007) 706.[22] B. Ramesh, S. Ramesh, R. Vijaya Kumar, M. Lakshmipathi Rao, Journal of

Alloys and Compounds 513 (2012) 289.[23] M.H. Alimuddin, S. Kumar, S. Ali, B.H. Koo, H. Chung, R. Kumar, Journal of

Alloys and Compounds 511 (2012) 107.[24] J. Du, Y. Liu, G. Yao, X. Long, G. Zu, J. Ma, Journal of Alloys and Compounds 510

(2012) 87.

[25] M.A. Khan, M.U. Islam, M. Ishaque, I.Z. Rahman, Journal of Alloys andCompounds 519 (2012) 156.

[26] T.J. Shinde, A.B. Gadkari, P.N. Vasambekar, Journal of Alloys and Compounds

513 (2012) 80.[27] J. Li, Z. Yu, K. Sun, X. Jiang, Z. Xu, Z. Lan, Journal of Alloys and Compounds 513

(2012) 606.[28] D.S. Birajdar, D.R. Mane, S.S. More, V.B. Kawade, K.M. Jadhav, Materials

Letters 59 (2005) 2981.[29] L. Nalbandian, A. Delimitis, V.T. Zaspalis, E.A. Deliyanni, D.N. Bakoyannakis,

E.N. Peleka, Journal of Microporous and Mesoporous Materials 114 (2008)465.

[30] M.K. Shobana, S. Sankar, V. Rajendran, Materials Chemistry and Physics 113(2009) 10.

[31] E. Rezlescu, L. Sachelarie, P.D. Popa, N. Rezlescu, IEEE Transactions onMagnetics 36 (2000) 3962.

[32] R.G. Kharabe, R.S. Devan, B.K. Chougale, Journal of Alloys and Compounds 463(2008) 67.

[33] E. Wolska, W. Wolski, P. Piszora, M. Pietrusik, J. Subrt, T. Grygar,M. Nejezchleba, International Journal of Inorganic Materials 1 (1999) 187.

[34] A.K. Nikumbh, A.V. Nagawade, S. Gugale, M.G. Chaskar, P.P. Bakare, Journal of

Materials Science 37 (2002) 637.[35] S.S. Bellad, S.C. Watawe, A.M. Shaik, B.K. Chougule, Bulletin of Materials

Science 23 (2000) 83.[36] HighScore Plus, Full Powder Pattern Analysis Software, V2.2, PANALYTICAL,

Almelo, Holland, 2004.[37] Zhao-Qian Li, Xin-Shan Lin, Lei Zhang, Xue-Tai Chen, Zi-Ling Xue, Crystal

Engineering Communications (2012), http://dx.doi.org/10.1039/C2CE06622J.[38] CRC Handbook of Chemistry and Physics, 90th Edition, CD-ROM Version,

2010.[39] M.A. Ahmed, S.T. Bishay, Journal of Physics D: Applied Physics 34 (2001) 1.[40] T.K. Kundu, D. Chakravorty, Journal of Materials Research 14 (1999) 3957.[41] Z. Xu, X. Dai, J.F. Li, D. Viehland, Applied Physics Letters 68 (1996) 1628.