IEEE TRANSACTIONS ON MAGNETICS, VOL. 45, NO. 10, OCTOBER 2009 4219

Analysis of Grain Shape and Orientation in BaFe��O��-Ferrites UsingElectron Backscatter Diffraction (EBSD)

A. Koblischka-Veneva�, M. R. Koblischka�, Y. Chen�, and V. G. Harris�

Institute of Functional Materials, Saarland University, 151150, D-66041 Saarbrücken GermanyInstitute of Experimental Physics, Saarland University, 151150, D-66041 Saarbrücken Germany

Department of Electrical and Computer Engineering and Center for Microwave Magnetic Materials and Integrated Circuits,Northeastern University, Boston MA 02115 USA

The electron backscatter diffraction (EBSD) technique enables an advanced analysis of anisotropic materials like ferrites. Here, thespatially highly resolved EBSD mappings provide additional information as compared to the standard analysis techniques, which cancontribute to an optimization of the growth process. Furthermore, an analysis of the grain aspect ratio is possible which provides furtherinsight to the microstructural dependence of the magnetic properties of ferrites.

Index Terms—Crystal growth, electron microscopy, electron backscatter diffraction, ferrite, position measurement.

I. INTRODUCTION

T HE analysis of the achieved texture is of great impor-tance for ferrite materials, either bulk or thin films. The

magnetic properties were found to be strongly dependent onthe orientation, size and shape of the crystallites, and samplehomogeneity plays an important role as well [1]–[4]. Fur-thermore, in the case of hexaferrites, a strong texture of the-axis is important especially in applications of next generationmicrowave devices [5]–[7]. The recently developed electronbackscatter diffraction (EBSD) technique, which works withina scanning electron microscope, enables a spatially resolvedstudy (a resolution of about 20 nm is possible on perfectlyprepared surfaces of oxidic materials [8]–[11]) of the crys-tallographic orientations by means of recording of Kikuchipatterns. A good surface polishing/cleaning is essential forthis analysis, as the method requires an undisturbed surfacearea for a high image quality. This information is recordedto each measured Kikuchi pattern, together with a parameterdescribing the quality of indexation. To each pattern, the threeEulerian angles determine the crystallographic orientation[12]. From this information, orientation maps and local polefigures can be constructed, enabling further a detailed analysisof grain boundaries. In this contribution, we present an EBSDanalysis of two bulk M-type hexaferrite BaFe O (Ba-ferrite)samples, considering especially the grain aspect ratio analysis.

II. EXPERIMENTAL PROCEDURE

The EBSD system employed here consists of a FEI dual beamworkstation (Strata DB 235) equipped with a TSL OIM anal-ysis unit [13]. The Kikuchi patterns are generated at an accel-eration voltage of 20 kV, and are recorded by means of a Di-giView camera system, allowing a maximum recording speed ofthe order of 0.05 s/pattern. The time employed in the case of a

Manuscript received March 06, 2009. Current version published September18, 2009. Corresponding author: A. Koblischka-Veneva (e-mail: a.koblischka-veneva@mx.uni-saarland.de).

Color versions of one or more of the figures in this paper are available onlineat http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/TMAG.2009.2022020

Fig. 1. Kikuchi patterns and the corresponding indexation of (a) BaFe O .The EBSD-determined Eulerian angles are �� � ����� �� � ���� and�� � ���� .

multi-phase scan is much longer, of the order of 0.4 s/pattern, asa higher image quality/confidence index is required. To producea crystallographic orientation map, the electron beam is scannedover a selected surface area and the resulting Kikuchi patternsare indexed and analysed automatically (i.e., the Kikuchi bandsare detected by means of the software, see Fig. 1). An imagequality (IQ) parameter and a confidence index (CI) is recordedfor each Kikuchi pattern.

The dimensionless IQ parameter is the sum of the detectedpeaks in the Hough transform employed in the image recording;the CI value yields information about how exact the indexationwas carried out. The CI value ranges between 0 and 1 [13].Based on the analysis of the recorded CI value, a multi-phaseanalysis is realised. A detailed description of the measurementprocedure can be found in [14]–[16]. The results of the EBSDmeasurement are presented in form of maps and graphs, themost important thereof are the so-called inverse pole figure(IPF) maps, indicating the crystallographic orientation of eachindividual point. In the present contribution, we focus on amore specific analysis of the grain shape and orientation. Thesample N1 was prepared by conventional ceramic processing[17], whereas sample N2 was prepared by a one-step topotacticreaction technique that will be published elsewhere.

The surface preparation process of the bulk ferrites is a del-icate task, as for a high spatial resolution a mechanical pol-ishing down to 40 nm colloidal silica is required, which is then

0018-9464/$26.00 © 2009 IEEE

4220 IEEE TRANSACTIONS ON MAGNETICS, VOL. 45, NO. 10, OCTOBER 2009

Fig. 2. Grain shape aspect ratio plot and histogram for sample N1. The his-togram gives the color code for the mapping.

followed by a thermal etching or by Ar-ion polishing. The IQvalues achieved range about 300, which is very high for oxidicceramic samples; thus automated multi-phase scans are possibleto be performed.

III. RESULTS AND DISCUSSION

Maps of the crystallographic orientation and of the detectedgrain boundaries were presented in a previous publication [18].Here, we focus on the results of the grain size aspect ratio anal-ysis, which provides an important step for the preparation of theanisotropic ferrite materials, whereas the maps of the crystallo-graphic orientation and of the detected grain boundaries will bedetailed in another publication [18]. Fig. 1 shows the EBSD-de-termined grain shape aspect ratio map and the correspondinghistogram for sample N1. Fig. 2 presents the same analysis forsample N2.

Within the EBSD software, the grain shape aspect ratio, , isdefined as the minor axis of an ellipse divided by its major axis,so a grain with a needle-like shape (i.e., commonly called a highaspect ratio) has . Only the grains which arefully included within the selected area take part in this analysis;the other grains are represented in white. Furthermore, as the

Fig. 3. Grain shape aspect ratio plot and histogram (color code) for sample N2.

crystallographic orientation of each individual grain is knownin the EBSD analysis, also the exact location of the major andminor axes of the ellipses which are employed to define a grainwithin the EBSD software [13], are always well defined andbeing treated accordingly in the analysis.

Therefore, this type of plot enables a correct analysis ascompared to e.g., optical methods. Sample N1 exhibits a largenumber fraction of grains with . Furthermore, there isa relatively high number of tiny grains with . Thesesmall grains are located along the edges of the bigger grains,and act quasi as a filler material between the bigger grains.Several such grains are even located inside the bigger grains.Grains with do practically not exist within theselected sample area.

In contrast to Fig. 2, the grain shape aspect ratio histogramof sample N2 looks much more equalized, with a maximumnumber fraction of 0.122. Furthermore, grains with all aspectratios are present in the selected area. As a consequence, a muchmore homogeneous grain matrix is obtained here.

Figs. 2 and 3 reveal clear differences between the two sampleswhere sample N1 shows a large maximum at .

KOBLISCHKA-VENEVA et al.: ANALYSIS OF GRAIN SHAPE AND ORIENTATION IN BaFe O -FERRITES 4221

Fig. 4. Major axis orientation map and histogram for sample N1. The histogramgives the color code for the mapping as shown above.

This grain shape aspect ratio analysis enables now to analysein detail what happens if a ferrite is prepared by different tech-niques.

Only a very small number of extremely large grains exists,and also the amount of grains with corresponds tothe number fractions found for the other orientations. Also, thenumber of grains with is considerably reduced. Theoverall number of small grains is observed to be reduced as well;producing a much more compact ceramic matrix.

Another piece of information can be collected from Figs. 4and 5, where the major axis orientation maps and the corre-sponding histograms for both samples are given. For a definitionof this procedure, see [19]. In the histogram of Fig. 4 (sampleN1), a large maximum is obtained for a major axis orientationat around 15 . This result is, however, again corrupted by thelarge amount of tiny grains being present. Sample N2 is showsagain a much more equalized histogram, see Fig. 5. As a conse-quence, the major axis orientation map reveals relatively largeareas with only a small axis misorientation, which is very usefulfor the resulting magnetic properties of the sample.

Fig. 5. Major axis orientation map and histogram (color code) for sample N2.The histogram gives the color code for the mapping as shown above.

The spatially highly resolved EBSD analysis contributes,therefore, to choose an optimum processing condition for agiven compound. The information on the individual grain ori-entation provides important input to understand the variation ofthe magnetic properties of ferrite samples, and helps to improvethe properties of ferrite materials even further. Additionally,the crystallographic orientation of grains even down to severaltens of nanometres can be directly analysed.

In summary, we have performed a grain shape aspect ratioanalysis by means of EBSD. Knowing the individual crystallo-graphic orientation of each grain enables an exact analysis ofthe grain shape aspect ratio also for tiny grains, so that exactresults can be obtained. Especially for materials with a largeanisotropy, this analysis may provide important information forthe processing of the respective samples.

IV. CONCLUSION

The EBSD technique successfully measures the crystallo-graphic orientation of bulk ferrite samples. Based on thesedata, a detailed analysis of bulk Ba-ferrite samples allows us

4222 IEEE TRANSACTIONS ON MAGNETICS, VOL. 45, NO. 10, OCTOBER 2009

to identify a correlation between the grain texture and theorientation technique applied during the sample preparation.Furthermore, the spatially resolved EBSD maps indicate achange of the major axis orientation within the samples due tothe processing route.

ACKNOWLEDGMENT

This work was supported by DFG project Mu959/19, whichis gratefully acknowledged.

REFERENCES

[1] J. Smit and H. P. J. Wijn, Ferrite. Eindhoven, The Netherlands:Philips Technical Library, 1960.

[2] S. Krupicka, Physik der Ferrite und der verwandten magnetischenOxide. Braunschweig, Germany: Vieweg & Sohn, 1973.

[3] A. Goldman, Modern Ferrite Technology. New York: Van Nostrand,1990.

[4] S. Rösler, P. Wartewig, and H. Langbein, “Synthesis and characteri-zation of hexagonal ferrites by thermal decomposition of freeze-driedpowders,” Cryst. Res. Technol., vol. 38, pp. 927–934, Nov. 2003.

[5] Y. Chen, T. Sakai, T. Chen, S. D. Yoon, A. L. Geiler, C. Vittoria, andV. G. Harris, “Oriented barium hexaferrite thick films with narrow fer-romagnetic resonance linewidth,” Appl. Phys. Lett., vol. 88, p. 062516,2006.

[6] Y. Chen, A. L. Geiler, T. Chen, T. Sakai, C. Vittoria, and V. G. Harris,“Low-loss barium ferrite quasi-single-crystals for microwave applica-tion,” J. Appl. Phys., vol. 101, p. 09M501, 2007.

[7] V. G. Harris, A. Geiler, Y. Chen, S. D. Yoon, M. Z. Wu, A. Yang, Z.Chen, P. He, P. V. Parimi, X. Zuo, C. E. Patton, M. Abe, O. Acher,and C. Vittoria, “Recent advances in processing and applications ofmicrowave ferrites,” J. Magn. Magn. Mater., 2009, in press.

[8] A. Koblischka-Veneva, M. R. Koblischka, F. Mücklich, S. Murphy, Y.Zhou, and I. V. Shvets, “Crystallographic orientation analyses of mag-netite thin films using electron backscatter diffraction (EBSD),” IEEETrans. Magn., vol. 42, pp. 2873–2876, 2006.

[9] A. Koblischka-Veneva, M. R. Koblischka, J. D. Wei, Y. Zhou, S.Murphy, F. Mücklich, U. Hartmann, and I. V. Shvets, “Misorienta-tions in [001] magnetite thin films studied by electron backscatterdiffraction and magnetic force microscopy,” J. Appl. Phys., vol. 101,p. 09M507, 2007.

[10] A. Koblischka-Veneva, M. R. Koblischka, S. Murphy, S. K. Arora, F.Mücklich, U. Hartmann, and I. V. Shvets, “Microtexture of magnetitethin films of [001] and [111] orientations on MgO substrates studied byelectron-backscatter diffraction,” J. Appl. Phys., vol. 103, p. 07E505,May 2008.

[11] T. Tepper, C. A. Ross, and G. F. Dionne, “Microstructure and opticalproperties of pulsed-laser-deposited iron oxide films,” IEEE Trans.Magn., vol. 40, pp. 1682–1690, 2004.

[12] K. Z. Baba-Kishi, “Electron backscatter kikuchi diffraction in the scan-ning electron microscope for crystallographic analysis,” J. Mater. Sci.,vol. 37, pp. 1715–1746, 2002.

[13] Orientation Imaging Microscopy Software Version V4.0, user manual.TexSEM Laboratories (TSL), Draper, UT, 2004.

[14] M. R. Koblischka and A. Koblischka-Veneva, “Characterization of bulksuperconductors through EBSD methods,” Physica C, vol. 392–396,pp. 545–556, 2003.

[15] A. Koblischka-Veneva, F. Mücklich, M. R. Koblischka, N. H. Babu,and D. A. Cardwell, “Crystallographic orientation of Y Ba Cu�O(� � Nb, Zr, Ag) nanoparticles emebedded in bulk, melt-texturedYBCO using EBSD,” J. Am. Ceram. Soc., vol. 90, pp. 2582–2588, Aug.2007.

[16] A. Koblischka-Veneva, M. R. Koblischka, P. Simon, F. Mücklich, K.Ogasawara, and M. Murakami, “Application of electron backscatterdiffraction to bulk high-� superconductors,” Supercond. Sci. Technol.,vol. 15, pp. 796–802, 2002.

[17] Y. Chen, M. J. Nedoroscik, A. L. Geiler, C. Vittoria, and V. G. Harris,“Perpendicularly oriented polycrystalline BaFe Sc O hexafer-rite with narrow FMR linewidths,” J. Am. Ceram. Soc., vol. 91, pp.2952–2956, 2008.

[18] Y. Chen, A. L. Geiler, A. Koblischka-Veneva, M. R. Koblischka, C.Vittoria, and V. Harris, unpublished.

[19] S. I. Wright and D. P. Field, “Scalar measures of texture heterogenity,”Mater. Sci. Forum, vol. 495–497, pp. 207–212, 2005.