Structural and Magnetic Properties of Mo-Zn Substituted (BaFe12-4xMoxZn3xO19) M-type Hexaferrites

Molybdenum-zinc substituted hexaferrites were synthesized by high-energy ball milling and subsequent sintering at different temperatures (1100, 1200, and 1300° C). The samples sintered at 1100° C exhibited good hard magnetic properties, although a decrease in saturation magnetization from 70.2 emu/g for the unsubstituted sample down to 57 emu/g for the sample with x = 0.3 was observed. The drop in saturation magnetization results mainly from the presence of secondary nonmagnetic oxides. The samples sintered at temperatures >1200° C showed an improvement in saturation magnetization, and a sharp drop in coercivity. This behavior was associated with the development of the W-type hexaferrite, the particle growth, and possibly the spin reorientation transition from easy-axis to easy-plane.

of 160-255 kA m -1 .BaMhexaferrite is the most important hexaferrite in terms of production (more than 50% of the total globally manufactured magnetic materials [4]).
The unit cell of M-type barium hexaferritesis built by stacking R (BaFe 6 O 11 ) and S (Fe 6 O 8 ) blocks in the sequence RSR*S*, where the star denotes a block rotated by 180° about the c-axis of the hexagonal lattice [4].The unit cell therefore contains two (BaFe 12 O 19 ) molecules.The S block contains two hexagonal layers of four oxygen ions in each, while the R block consists of three hexagonal oxygen layers, with one oxygen ion in the middle layer replaced substitutionallyby a Ba ion.The metal ions occupy five different interstitial site: two sites in the S block (the octahedral 2a site and the tetrahedral 4f 1 site), two sites in the R block (the octahedral 4f 2 site and the by-pyramidal 2bsite) and one site at the R-S interfaces (the 12k octahedral site).These sublattices, their coordinations, the number of metal ions in each, and their spin directions are listed in Table 1.
The magnetic and electrical properties of barium hexaferrites were found to depend critically on the substitution of barium ions or iron ions by other cations and cations combinations.Trivalent metal ions, or combinations of divalent and tetravalent ions were used to substitute Fe 3+ ions in the hexaferrite lattice.These included Al 15,19,26 , Ga 19,27 , (Mo x Zn 0.4-x ) 14 , Mn 12 , Cr 19 , Ti-Ru 16 , Zn-Ti 13 , and Sn-Ru 18 .
In the present work, we synthesized barium hexaferrites with Fe ions partially substituted by Mo 6+ -Zn 2+ combinations.In order to maintain the cationic valence states the ratio of Mo:Znwas fixed at 1:3.The effects of the type of substitution and substitution level, and heat treatment on the structural, magnetic and physical properties of the prepared samples were investigated using x-ray diffraction (XRD), scanning electron microscopy (SEM), and vibrating sample magnetometry (VSM).
The structural refinement for the prepared samples was achieved using FULLPROF software based on Rietveld refinement techniques.

EXPERIMENTAL
High purity (~99%) powders of BaCO 3 , Fe 2 O 3 , ZnO and MoO 2 were used as starting materials to prepare the powder precursors of Mo-Zn substituted(BaFe 12-4x Mo x Zn 3x O 19 )M-type hexaferrites.The barium to metal molar ratio was 1:11, and the molybdenum to zinc ratio was 1:3.Since zinc is divalent, the substitution of this combination for Fe 3+ ions in these hexaferritesensures that molybdenum has the Mo 6+ valence state, which ensures the chemical neutrality of the hexaferrite.
The required amounts of the starting powders were weighed accurately and then mixed and transferred to the zirconia milling vessels.The mixture was then milled in an acetone bath (8 ml for each 5 grams of the powder) for 16 hours.The milling time was sectioned into 10 min.grinding periods interrupted by 5 min.pause periods to allow for cooling the sample and avoid overheating.The ball-to-powder mass ratio was 14:1 and the rotation speed was 250 rpm.The resulting wet powder mixture was left in the container overnight to dry at room temperature, and then the dry powder was collected in clean glass vials.An adhesive agent of aqueous solution of 2% wt. of polyvinyl alcohol (PVA) was prepared and transferred to the vial containing the milled powder and thoroughly mixed with the powder.After drying at room temperature, parts of the powder (about 1 gram each) were pressed into a discs (1.5 cm in diameter) in a stainless steel die under a 4 ton force.The discswere subsequently sintered at temperatures in the range 1100 -1300°C.
The density for each disc was calculated by dividing the mass of the disk by its volume, and found to be independent on the level of substitution (2.7 ± 0.1 g/cm 3 for all samples sintered at 1100° C).However, we noticed a tendency toward increasing the density with increasing sintering temperature, where the average density of the discs sintered at 1200° C increased to about 3.0 ± 0.1 g/cm 3 and that for discs sintered at 1300° C increased up to about 3.3 ± 0.05g/cm 3 .This behavior could be attributed to the growth in particle size with increasing the sintering temperature, which in turn reduces the porosity of the samples.
Scanning electron microscope (SEM) system (FEI-Inspect F50/FEG)was used to investigate the microstructural characteristics of the samples.The chemical compositions of the samples were determined using the energy dispersive x-ray spectroscopy (EDX) facility available in the SEM system.
The structural characteristics of the samples were investigated using x-ray diffraction (XRD).XRD patterns of the samples were obtained in the angular range 20° -70° using XRD 7000-Shimadzu machine with Cu-Ka radiation.The scanning step was 0.02° and the scanning rate was 1deg/min.The patterns were then analyzed using dedicated software routines (FULLPROF and Expert High Score) to determine the structural parameters of the samples.
The coercivity (H c ),remnant magnetization (M r ) and saturation magnetization (M s ) of the samples were determined from the hysteresis loops measured using a standard vibrating sample magnetometer (VSMMicro Mag 3900, Princeton Measurements Corporation).Al lmagnetic measurements were performed at room temperature in an applied field up to 10 kOe.

Scanning electron microscopy (SEM)
Fig. 1 shows SEM images for the samples at T = 1100° C. The images indicate that the particle size and morphology does not seem to be influenced by the substitution, and show hexagonal platelets with diameters ranging from 200 nm to about 500nm for all samples under investigation.However, lighter particles with generally smaller size and different morphology was also observed, which gave the impression that we may have different phases in the samples.This was examined by EDX analysis of the spectra collected at darker (D) and lighter (L) particles of the sample with x = 0.2 (BaFe 11.2 Mo 0.2 Zn 0.6 O 19 ) as an example (Fig. 2).The atomic Ba:Fe ratio in the darker particles was found to be 1:12.2which is close to the stoichiometric ratio of BaFe 12 O 19 .On the other hand, the Ba:Mo ratio in thelighter particles was 1:1.07, which is consistent with the stoichiometric ratio of BaMoO 4 .These results indicate that the Mo-Zn substituted samples with BaM stoichiometry may contain a secondary barium molybdenum oxide phase coexisting with thehexaferrite major phase under these experimental conditions.This method, although does not provide an accurate quantitative analysis of the compositions of the samples due to the low concentrations of some elements in comparison with the experimental uncertainty of the technique, it is useful in checking for the possibility of having secondary phases.Therefore, a detailed structural analysis is required to identify the existing phases and their structural properties.A large number of parameters can be obtained directly from the refinement routine such as the lattice constants (a and c), cell volume V, Miller indices of the diffraction peaks (hkl), and the goodness of fit parameters.Some refined parameters resulting from fitting the experimental diffraction data are shown in Table 2.The refinement results indicates slight fluctuations of the lattice constants around a = 5.89 Å and c = 23.21Å.The slight increase in cell volume upon substitution of iron by zinc and molybdenum could be due to partial substitution of the largerZn 2+ ions (radius = 0.74 Å) for the smaller Fe 3+ ions (radius = 0.49 Å)at the 4f 1 tetrahedral sites 28 .

X-ray diffraction (XRD) measurements
In order to investigate the effect of compaction on the development of phases and crystallinity, samples in powder form were sintered at 1100° C and investigated by XRD.The refinement results indicated that these patterns are almost identical to those for disk-sintered samples, which indicates that the compaction of the powder under a 4-ton pressure did not influence the phase evolution in the samples.

The average crystallite size for each sample was calculated from the Scherrer equation [29]:
where k is a constant equals 0.9, λ = 1.542Å, β is the peak-width at half maximum, and è is the angular position of the peak.The average crystallite sizes for the investigated samples with x ranging from 0.0 to 0.2 are listed in Table 3.The data indicates that the crystallite size tends to decrease from about 15 nm down to about 9 nm with increasing x.This leads to the conclusion that the Mo-Zn substitution for Fe results in poorer crystallinity of the samples, probably arising from crystal defects.To examine the effect of sinter ing temperature on the structure of the ferrites, all samples were resintered at 1200° C for 2 h, and their structure investigated by XRD.Fig. 4 shows X-ray patterns for resintered samples.The patterns clearly indicate that ZnFe 2 O 4 phasedisappeared in the samples, were as the BaMoO 4 phase persisted throughout the whole concentration range and a newhigh-temperature phase evolved, which was identified as W-type BaZn 2 Fe 16 O 27 (Zn 2 W) phase.Baring in mind that the unit cell of the W-type phase is a combination of the unit cell of BaM and a spinel (Zn 2 Fe 4 O 8 ) block, this phase apparently evolved according to the reaction: The structural parameters for the M-type phase in these samples were derived from the FULLPROF analysis and listed in Table 4.The data indicate slight decrease in cell volume toward that of the un-substituted sample with increasing the sintering temperature.This is associated with the incorporation of Zn 2+ ions in the evolving Zn 2 W hexaferrite rather than in the BaMhexaferrite lattice.To fur ther investigate the effect of sintering t e m p e r a t u r e , t h e s a m p l e w i t h x = 0 . 1 5 (BaFe 11.4 Mo 0.15 Zn 0.45 O 19 )was re-sintered at a temperature 1300° C and the patterns for this sample at different temperature are shown in Fig. 5.It is evident that the BaMoO 4 phase existed at all temperatures, while ZnFe 2 O 4 disappeared completely at temperatures higher than 1200° C. At such high sintering temperatures, the reaction between the (intermediate) M-type and spinel phases is completed to form the Zn 2 W-type phase.Thus, we conclude that the reaction the M-typephase and the zinc spinel Zn 2 Fe 4 O 8 (S) phase is favored at high temperatures (~1300° C).Our results are consistent with previously reported results 30 , which indicated the presence of M-phase at low sintering temperatures, and the development of the W-phase at temperatures higher than 1200° C.

Magnetic measurements
Fig. 6 shows the hysteresis curves for BaFe 12-4x Mo x Zn 3x O 19 samples sintered at 1100° C. The curves indicate that the magnetizations do not saturate in the magnetic field range of the study.According to the law of approach to saturation, the magnetization in the high field region is dominated by magnetic domain rotation 1 .Therefore, this law was used to determine the saturation magnetization for each sample from the high field region (H > 0.8 kOe), while the coercivity and remnant magnetization were determined directly from the hysteresis loops, and the results are listed in Table 6.
As shown in Table 6 the saturation magnetization for the un-substituted sample is 70.2 emu/g,the remnant magnetization is 40.6 emu/g.The squareness ratio of about 0.58 for this sample is close to the value (0.5) for a system of randomly oriented single domain particles 3 .The saturation magnetization (Fig. 7) decreased gradually with increasing Mo concentration, recording a 19% drop in the sample with x = 0.3.However, the magnetic properties of the products are still good for permanent magnet or data storage applications.
The initial drop in coercivity (Fig. 8)of about 25% for the sample with x = 0.1 is consistent with previously reported results 14 .This behavior was associated with the substitution of the Fe 3+ ions by Zn 2+ ions at 4f 1 sites and Mo 6+ ions at 2b sites (which have the highest contribution to the magnetic anisotropy).The progressive increase of Zn 2+ ions at 4f 1 sites and Mo 6+ ions at spin-up 2b sites is expected to result in an increase in the saturation magnetization and a further drop in coercivity, contrary to observed results.Therefore, we associate the noticeable decrease in saturation magnetization and the almost constant value of the coercivity with increasing x to the limited solubility of Zn -Mo ions in the BaM lattice, and the development of the nonmagnetic (ZnFe 2 O 4 and BaMoO 4 ) oxide phases, which influence the saturation magnetization, but not the coercivity.In view of the reduction of the saturation magnetization for the sample with x = 0.3, the nonmagnetic oxides apparently account for 19% of the sample weight.
It is well known that the sintering temperature has an important effect on the magnetic properties, especiallyon the coercivity value, which is sensitive to the grain morphology 3,31,32 .The hysteresis loops for samples with x = 0.2 sintered at different temperatures are shown in Fig. 9.The magnetic parameters derived from these loops are listed in Table 7.The increase in saturation magnetization with increasing sintering temperature is associated with the disappearance of the nonmagnetic ZnFe 2 O 4 phase and the development of the W-phase.Such structural developments were not observed in Mo-Zn substituted hexaferrites prepared by wet chemical mixing with different Mo to Zn ratio 14,33 .The sharp drop of the coercivity is in agreement with the results of Pasko et al., 30 who attributed this drop to the reaction of BaM and spinel phases to form the W-phase at temperatures higher than 1200° C. Further, the transition from a hard magnet (H C ~ 2800 Oe at 1100° C) to a soft magnet (H C ~ 300 Oe at 1300° C) could be associated with the reported spin reorientation transition from easy axis to easy plane in W-type hexaferrite.However, such transition in our sample should be confirmed by other techniques in a future work.In addition, the squarenes ratio of 0.53 for the sample sintered at 1100° C is consistent with single-domain particles randomly oriented.However, the decrease in squareness ratio at higher sintering temperatures is indicative of multi-domain particles due to particle growth at such high temperatures 30 .This multidomain structure typically results in a reduction of the coercivity 3 .

CONCLUSIONS
Barium hexaferrite phases were synthesized by high energy ball milling and subsequent sintering at temperatures 100, 1200, and 1300° C. Mo-Zn substitution for F 3+ ions was found to result in a decrease in saturation magnetization due to the presence of secondary impurity oxide phases in samples sintered at 1100° C. Samples sintered at higher temperatures showed an increase in saturation magnetization and a drastic drop in coercivity.The proposed spin reorientation transition at room temperature in the sample with x = 0.2 sintered at 1300° C suggests that this sample has a potential application in magnetic refrigeration 30

Figure 3
shows the XRD patterns for all samples sintered at 1100° C. The figure indicates the development of new phases with increasing x values as evidencedby the peaks around 26.6°, 30.0°, and 35.03°.The diffraction patternswere analyzed using FULLPROF software.The pattern for the un-doped sample shows a major phase with reflections consistent with BaFe 12 O 19 M-type hexaferrite(JCPDS: 00-043-0002) without other impurity phases.On the other hand, XRD patterns indicate that all doped samples consist of a major BaMhexaferrite phase and small amounts of other intermediate phases (ZnFe 2 O 4 and BaMoO 4 ).The formation of BaMoO 4 phase which is consistent with the results of EDX analysis, causes deficiency in the amount of Ba (or excessive amounts of Fe and Zn) required for BaM phase, resulting in the formation of ZnFe 2 O 4 phase.