The Mechanical Alloying Behavior of Fe2O3 into NiO in the High- Energy Ball-Milling Process

,eNiO and Fe2O3 powders were mixed by the high-energy ball-milling, followed by a sintering of the mixture at 1340°C for 0.5 h. XRD, SEM, DSC, and size measurements were preformed to study the microstructure evolution in the high-energy ball-milled mixture and the sintered ones, as well. It showed that the high-energy ball-milling processes resulted in a severe lattice distortion in the powder of Fe2O3, but only a slight lattice distortion in NiO. Meanwhile, a solid solution of iron atoms into the NiO lattice was also detected in the milling process. It was also found that the solubility of the iron atoms into the NiO lattice delayed the synthesizing reaction in the following sintering process.


Introduction
Most of the oxides are stable at the room temperature [1][2][3][4]. NiO and Fe 2 O 3 are the typical ones of these oxides. ey are often used as raw materials to synthesize ceramics, for instance, NiFe 2 O 4 . However, the transformation efficiency of both oxides into the ceramics is often limited by their stable microstructures in a usual sintering process [5]. us, it is necessary to understand the possible parameters in both the mixing and the sintering processes of the powders, which affects their synthesizing reaction in the sintering process.
High-energy ball-milling is a process that mixes multiple raw powders with greater efficiency before their sintering process [6,7]. It may alter the microstructure of the powder mixture. In the high-energy ball-milling process, powder particles are possibly broken and undergo significant plastic deformation [8][9][10][11][12].
us, a subsequent reaction may be changed in the milled powder mixture in the following sintering processes [13][14][15].
Researches are available on synthesizing of NiFe 2 O 4 ceramics by sintering of the NiO and Fe 2 O 3 powders [16,17]. But few studies have been reported about the effects of highenergy ball-milling on the microstructure of these two powders, especially its effect on the subsequent synthesizing reaction. e main aim of this study is to clarify the microstructure evolution of the NiO and Fe 2 O 3 powders and the possible interaction behavior between them in the highenergy ball-milling process, and further to evaluate its effect on the following sintering reaction.

Experimental Works
e purity of the as-received powders of NiO and Fe 2 O 3 was no less than 99.0%. e mean sizes of both powder particles were 1.4 μm and 10 μm, respectively.
Both powders were ball-milled in an ND7 planetary mill for 0.5, 2, 5, and, 10 h, respectively. e ball to powder weight ratio was 7 : 1. e molar proportion of Fe 2 O 3 /NiO powders was 11:10 to ensure a complete surface contact between the NiO and Fe 2 O 3 powders because the size of the Fe 2 O 3 particle was much larger than that of NiO. e whole high-energy ball-milling mixing process parameters are illustrated in Table 1.
All of the milled powders were dispersed in a solution with 99.99 wt.% alcohol using an ultrasonic system and then dried for further microstructure measurements and sintering experiment. In this sintering process, no cold densifying was performed, and only the effect of the sintering temperatures on the microstructure transformation was considered. us, part of the milled powders was sintered at 1340°C for 0.5 h. Both the microstructures of the milled powders and the sintered ones were measured in the study.
Size measurement of the milled powders was performed using the MASTERSIZER 2000 size analysis system. e micrograph of the powders was observed on a Navo 400 field emission SEM system at a voltage of 20 kV. X-Ray powder diffraction analysis was carried out using a SHIMADZU XRD-6000 X-ray powder diffractometer. Cu-Kα radiation and a scanning speed of 0.03342°/min were utilized. Phases were identified using the Bruker EVA (evaluation) software. Quantitative phase, grain size, and microresidual stress analyses were implemented using the software Topas3. ermogravimetric analysis was performed on a NETZSCH STA 449C simultaneous TG-DSC-STA system at a heating rate of 10°C/min in the air.

Results
It is shown in Figure 1 the measured particle size of the powders milled for different hours. Two strong peaks can be observed on the size distribution plot of the powders without subsequent high-energy ball-milling (Figure 1(a)). e first peak occurs at the particle diameter sizing in 1.5 μm, which corresponds to the mean particle size of the as-received Fe 2 O 3 powder.
e second peak appears at the particle diameter sizing in approximate 10 μm, which corresponds to the mean particle size of the as-received NiO powder. Around each of these two peaks, the particle sizes of the powders show the Gaussian distribution ( Figure 1(a)). However, if the ball-milling time increases from 0 h to 0.5 h, the mean particle size of Fe 2 O 3 shifts from ∼1.5 μm to ∼0.6 μm, as shown in Figure 1(b), which implies that the Fe 2 O 3 particles were brittle and had been broken in the ballmilling process. NiO particles with an original mean size of ∼10 μm remained unchanged with increasing the milling time, which indicates that the NiO particles were hard, ductile, and not easily broken in the ball-milling process. However, the volume fraction of the ∼10 μm NiO powder particles (obtained by integration of the whole peak) increased with increasing the ball-milling time (Figures 1(a)  and 1(b)). e possible reason for this increase is because the fine hard NiO particles were wrapped with the fine broken Fe 2 O 3 powders in the frequently ball-milling processes, which increased the size of the NiO particle originally finer than 10 μm. Furthermore, a weak peak can be observed at about 3um in the plot (Figure 1(b)). is weak peak (also see Figure 1(c)) is possibly owing to the allegation of the Fe 2 O 3 particle or the presence of the minor broken NiO particle.
If the ball-milling time increases to 2 h, the mean size of the small particles shifts from ∼1.5 μm of the as-received powder down to ∼0.3 μm of the milled powder, as shown in Figure 1(c). By contrast, no apparent size shift can be observed in the large particles with the mean size of ∼10 μm (Figure 1(c)). When the milling time exceeds 2 h, e.g., at 5 and 10 h, the mean size of the large particles remains the same but their relative volume fraction increases with increasing the milling time (comparing Figures 1(d) and 1(e) with Figure 1(a)). e mean size of the small particles increases slightly with increasing the milling time at this time. SEM images of the mixed powders with different milling time are shown in Figure 2. ese images demonstrate the size changes of the milled powders. e X-ray diffraction patterns of the milled powders are shown in Figure 3. It is illustrated that both the intensities and the peak widths of Fe 2 O 3 varied as a function of the milling time. In particular, the intensities decreased and the peak widths were broadened with increasing the milling time. Increases in intensities indicate the accumulation in the volume fraction of Fe 2 O 3 in the mixture. e broadening of the peak widths implies an increase in the microstress and a decrease in the grain size in Fe 2 O 3 . No apparent intensity changes are observed in the diffraction pattern of NiO with increasing the milling time, and very slight broadening may be seen here, as well. Meanwhile, a trace of the Fe 3 O 4 phase was detected in the powders milled for over 2 h, which may be owing to the decreasing in the oxygen partial pressure in the ball mill with increasing the milling time.
Grain sizes determined from the X-ray diffraction patterns are shown in Figure 4. It reveals there is virtually no change in the grain size of the NiO powders with increasing the milling time (see Figure 4(a)). By contrast, the Fe 2 O 3 grain sizes drastically declined from 1 μm before the ballmilling down to 0.2 μm after a ball-milling of 2 hours. e grain size of Fe 2 O 3 further decreased down to 0.01 μm when the ball-milling lasted for 5 hours, and then this grain size became almost constant when the ball-milling time increased from 5 h to 10 h.
For further understanding the dynamical changes that occurred in the high-energy ball-milling process, quantitative calculation of the phases in the powders was performed. e volume fraction of Fe 3 O 4 can be neglected in this calculation because of the low content of this compound in the milled powders detected from the X-ray diffraction.
us, only NiO and Fe 2 O 3 are considered in the calculation. Furthermore, the effect of the powder texture that resulted from the ball-milling can also be neglected because the milled powders were fully dispersed by ultrasonication.
us, the diffraction intensities of NiO and Fe 2 O 3 were proportional to their volume fraction in the milled powder mixture. en, the volume fractions of the milled NiO and Fe 2 O 3 were calculated using the software Topas3, and the relevant results are shown in Figure 5. It is illustrated that the shows an opposite trend. According to the X-ray diffraction pattern as shown in Figure 3, the calculated lattice constant of NiO and Fe 2 O 3 in the milled powders are given in Table 2. It can be seen that the lattice parameter of NiO increases from a � 4.1723Å in the asreceived powder to a � 4.1730Å in the powder mixture milled for 5 h. is result reveals that some Fe 2 O 3 was dissolved into the lattice of NiO, which led to a decrease in the content of Fe 2 O 3 and a relative increase in that of NiO in the X-ray diffraction pattern. e lattice constants "a" and "c" of Fe 2 O 3 change frequently with the milling time, indicating the presence of the lattice distortion in the powder.
To investigate the interaction of the different powders in the milling process, SEM and EDX analysis of the typical milled powder mixture were performed. EDX analysis was done at least 10 points on each of the samples. e typical results are given in Figure 6. Here, independent Fe 2 O 3 particles were detected in the mixtures milled for no longer than 2h (seen in Figures 6(a) and 6(b)). When the milling time increased to more than 2 h, no independent Fe 2 O 3 particles can be observed in the mixture powder (seen in Figures 6(c) and 6(d)). Moreover, Fe and Ni coexisted in the same powder particles but the content of Fe gradually decreased with increasing the milling time when the milling time was no less than 5 h, which implied that Fe 2 O 3 was dissolved into NiO in the further milling process. Figure 7 shows the microstrain ε in the powders plotted against the ball-milling time. Since the elastic moduli E of Fe 2 O 3 and NiO were 201 and 126 GPa [18,19], respectively, the third residual stress σ can be calculated according to Hooke's law by utilizing the equation σ � Eε. e third residual stresses obtained in the ball-milling process are shown in Figure 8. e strain energy of the milled powders could be calculated according to equation W � (1/2) Eε 2 [20], and the relevant results can be calculated using the data in Figure 7,  as shown in Figure 9. e strain energy of the milled powders initially increased with increasing the milling time, which peaked at a certain point and then decreased thereafter. e microstructure evolution in the mixing process would possibly affect the synthesis reaction of the mixed Fe 2 O 3 and NiO in the following sintering process. Before the sintering experiment, the sintering temperature was determined by DSC (see Figure 10). A maximum temperature of 1450°C was used in the DSC measurement. Figure 10 demonstrates that the shapes of the DSC curves of the powders milled for 0, 2, and 10 h at temperatures below 1200°C are similar. However, an exothermic peak could be observed in the powders without milling at 1340°C (seen in Figure 10(a)), which may correspond to the transformation point of the mixture powder into NiFe 2 O 4 . In DSC curve of the milled powders, an obvious endothermic peak could be  Advances in Materials Science and Engineering observed after milling for 2 h (Figure 10(b)) and 10 h (Figure 10(c)) at about 1340°C, as well. us, 1340°C was used as the temperature for sintering. Figure 11 shows the XRD patterns of the sintered powders which were ball-milled for 0, 2, 5, and 10 h, respectively. Good agreement in terms of the diffraction angles

Discussions
e high-energy ball-milled powder particle sizes illustrated in Figure 1 show that these sizes were changed during the ball-milling. Primarily, the small particles tended to attach to the large particles, which enhanced the volume fraction of the large ones. A further ball-milling, for instance, the ballmilling for 2 h, caused decreases in the size of both the coarse and the fine powder particles, which indicated that the particles were broken. After a ball-milling for over 2 h, e.g., 5 h and 10 h, the volume fraction of the coarse particles increased while that of the fine powders decreased. e activity of the powder particles at this stage can be an agglomeration of the fine powder particles and partially attached to the coarse particles, resulting in increasing the volume fraction of the coarse powder particles. ose illustrated in Figure 2 are the SEM images of the mixed powders obtained at different milling times, which shows that the sizes of the powder particles vary with the milling time. It seems that the size changes of the powder particles with the milling time may be divided into three stages. At the primary milling stages, the powder particle size increased with increasing the milling time (seen in Figures 2(a) and 2(b)). And then the size decreased with an increase in the milling time (seen in Figures 2(b) and 2(c)) at a further milling stage. In the third stage, the powder particle size increased with increasing the milling time again (seen in Figures 2(c) and 2(d)). ese three stages implied different physical behavior of the powder particles in the milling process. At the primary milling stage, the small Fe 2 O 3 powder particles were easier to be broken than the large NiO powder particles. e smaller broken pieces of Fe 2 O 3 tended to attach to the larger NiO particles, resulting in increases in both the volume fraction and the average size of the milled NiO particles. is finding also indicates that NiO is more difficult to break than Fe 2 O 3 (Figure 1). In the second stage of ball-milling, the broken Fe 2 O 3 powder particles became too small to deform further and to separate the adjacent NiO powder particles from each other. us, more opportunities for the collisions among NiO particles occurred, which resulted in the breakage of the NiO particles. It is noted that the average size of the NiO powder particles decreased at this stage. Finally, the average particle sizes of both NiO and Fe 2 O 3 increased once more because the temperature rising in the mixture powders were brought about by the collisions between different adjacent powder particles, and between powder particles and the wall of the jar mill. ese temperature increments enhanced the adhesive property of all  Advances in Materials Science and Engineering the powders, thereby leading to the aggregation of these powder particles, and further increased the average size of the milled particles. e strain energy of Fe 2 O 3 , as shown in Figure 9, changes more dramatically than that of NiO, which implies that the plastic deformation is more susceptible to occurring in Fe 2 O 3 than that in NiO in the ball-milling process. e deformation process can be clarified as follows: in the primary milling stage, the main physical processes occurring in the powders were the breakage of particles and the accumulation of strain in the powders. High-speed collisions between the adjacent powder particles, between the powder particles and the milling balls, and between the powder particles and the container wall generated a large amount of heat, which enhanced the temperature of the powders [21]. In this milling stage, the temperature in the powders was not high enough to lead to the recovery of them. If the milling time was adequately long, the temperature would increase to that higher than the recovery temperature of the powders.
us, the strain energy in the powders tended to decrease. When these decreases in the strain energy were balanced by the increases in the strain energy caused by the powder deformation, the total strain energy reached a peak. After this peak, the temperature of the powders, for instance, in Fe 2 O 3 milled for 5 h, was high enough to "soften" the powders. erefore, strain energy decreased with increasing milling time.
As shown in Figure 10, the shift from an exothermic process to an endothermic one caused by the ball-milling may be due to the changes in the reaction products, the changes in the reaction pathways, or both.
According to the XRD results shown in Figures 3 and 11, the ball-milling leads to a solid solution of Fe 2 O 3 into the lattice of NiO, which expands the lattice constant of the latter (see Table 2). It is also reported that the strong internal strains introduced during mechanical milling treatment causes change in lattice constant and results in the unit cell expansion [22,23]. e solid solution of the ferrite ion into the FCC NiO interstice and the severe distortion in the NiO lattice caused by the high-energy milling would probably result in a rearrangement of ferrite ion in the NiO lattice in the following sintering process. us, an obvious increasing in the intensity of the (311) face in the formed NiFe2O4 X-ray diffraction pattern can be seen with increasing the milling time (see Figure 11).
is rearrangement of the ferrite ion in the sintering process would absorb additional energy which finally affected the DSC curve of the milled powders (see Figure 10).

Conclusions
(1) Powder particle size decreases with increasing the milling time up to 2 h, while powder particle size increases with increasing the milling time when the time is longer than 2 h. (2) Fe 2 O 3 powder particles are extensively broken and deformed during the ball-milling, whereas the size of NiO powder particles changes very slightly. (3) e ball-milling leads to a solid solution of iron atoms into the lattice of NiO, which delayers the reaction of the powders during the following sintering process.
Data Availability e data in the article are available if necessary.  Advances in Materials Science and Engineering 9