Microstructural Evolution of Dy2O3-TiO2 Powder ... - Semantic Scholar

1 downloads 0 Views 4MB Size Report
Dec 28, 2016 - Jinhua Huang, Guang Ran *, Jianxin Lin, Qiang Shen, Penghui Lei, Xina ...... Kong, L.B.; Ma, J.; Zhu, W.; Tan, O.K. Preparation of PMN-PT ...
materials Article

Microstructural Evolution of Dy2O3-TiO2 Powder Mixtures during Ball Milling and Post-Milled Annealing Jinhua Huang, Guang Ran *, Jianxin Lin, Qiang Shen, Penghui Lei, Xina Wang and Ning Li College of Energy, Xiamen University, Xiamen 361102, China; [email protected] (J.H.); [email protected] (J.L.); [email protected] (Q.S.); [email protected] (P.L.); [email protected] (X.W.); [email protected] (N.L.) * Correspondence: [email protected]; Tel./Fax: +86-592-2185278 Academic Editor: Jan Ingo Flege Received: 30 October 2016; Accepted: 23 December 2016; Published: 28 December 2016

Abstract: The microstructural evolution of Dy2 O3 -TiO2 powder mixtures during ball milling and post-milled annealing was investigated using XRD, SEM, TEM, and DSC. At high ball-milling rotation speeds, the mixtures were fined, homogenized, nanocrystallized, and later completely amorphized, and the transformation of Dy2 O3 from the cubic to the monoclinic crystal structure was observed. The amorphous transformation resulted from monoclinic Dy2 O3 , not from cubic Dy2 O3 . However, at low ball-milling rotation speeds, the mixtures were only fined and homogenized. An intermediate phase with a similar crystal structure to that of cubic Dy2 TiO5 was detected in the amorphous mixtures annealed from 800 to 1000 ◦ C, which was a metastable phase that transformed to orthorhombic Dy2 TiO5 when the annealing temperature was above 1050 ◦ C. However, at the same annealing temperatures, pyrochlore Dy2 Ti2 O7 initially formed and subsequently reacted with the remaining Dy2 O3 to form orthorhombic Dy2 TiO5 in the homogenous mixtures. The evolutionary mechanism of powder mixtures during ball milling and subsequent annealing was analyzed. Keywords: microstructure; ball milling; dysprosium oxide; neutron absorber; phase evolution

1. Introduction High-energy ball milling has been widely used to prepare various types of materials, such as supersaturated solid solutions, metastable crystalline materials [1], quasicrystal phases [2], nanostructured materials [3], and amorphous alloys [4]. The technology was initially used in place of blending and sintering at elevated temperatures to prepare ceramic-strengthened alloys [5,6]. A large amount of mechanical energy is transformed into intrinsic energy in the target materials, which induces the formation of numerous defects in the crystal structure, such as vacancies, interstitials, cavities and dislocations, which are always in a non-equilibrium state [7–9]. The defects and structural disorders will increase the mobility of atomic diffusion and induce chemical reactions amongst components that are not present under equilibrium conditions [10]. Therefore, based on its excellent characteristics, ball milling was used to prepare bulk Dy2 TiO5 , which can be used as a neutron absorber in control rods in nuclear power plants. Control rods are very important in both operating and accident conditions because the nucleon reactivity must be controlled in order to safely operate a nuclear reactor [11]. In fact, bulk Dy2 TiO5 prepared by ball milling and sintering has been used in Russian power plant water reactors, such as MIR and VVER-1000 RCCAs [12,13], because of the excellent nucleon characteristics of the element dysprosium, as natural dysprosium consists of five stable isotopes with high thermal neutron absorption cross sections. The decay products are Ho and Er, which are also able to absorb neutrons. All of the radionuclides have low gamma activity and short half-life periods. The absorption

Materials 2017, 10, 19; doi:10.3390/ma10010019

www.mdpi.com/journal/materials

Materials 2017, 10, 19

2 of 12

cross sections of dysprosium isotopes range from 130 barn to 2600 barn. The region of resonance absorption is 1.6–25 eV, in which the absorption cross-section can reach approximately 1000 barn [14]. According to its equilibrium phase diagram, Dy2 TiO5 has three crystal structural types depending 1350 ◦ C

1680 ◦ C

on the temperature, orthorhombic ↔ hexagonal ↔ cubic [15], which have different physical properties and radiation resistance abilities. In fact, bulk Dy2 TiO5 in the cubic crystal structure has the lowest neutron irradiation swelling and highest irradiation resistance. Therefore, it is necessary to synthesize bulk Dy2 TiO5 in the cubic crystal structure. Jung [16] synthesized bulk Dy2 TiO5 with high purity and density using a polymer carrier chemical synthesis process, in which ethylene glycol was used as an organic carrier for metal cations. An amorphous phase was detected below 800 ◦ C, and orthorhombic Dy2 TiO5 was observed after sintering for 1 h at 1300 ◦ C, while little else was observed while sintering in the range of 800 to 1300 ◦ C. Panneerselvam [17] used both solid-state synthesis and wet chemical synthesis to prepare Dy2 TiO5 . However, the effect of sintering temperature on the phase evolution needs further investigation. The sinterability of Dy2 O3 and TiO2 with different molar ratios was determined for various ball-milling and sintering conditions [18]. Amit Sinha [19] reported the synthesis of bulk Dy2 TiO5 from mixtures of equimolar Dy2 O3 and TiO2 powders in a two-step process: (I) pyrochlore Dy2 Ti2 O7 was initially formed and (II) Dy2 Ti2 O7 then reacted with the remaining Dy2 O3 to form orthorhombic Dy2 TiO5 . The powder mixtures used in sintering were simply mixed during ball milling. Garcia-Martinez [7] observed the experimental phenomena of the transformation of Dy2 O3 from cubic to monoclinic and the synthesis of a hexagonal high-temperature phase, reported as Dy2 TiO5 , in an equimolar Dy2 O3 -TiO2 mixture during ball milling. Therefore, further investigation is needed into the evolutionary behavior of the microstructure under different ball-milling conditions and the effect of the state of the ball-milled powder on the sintering behavior. In the present work, the microstructural evolutionary behavior and corresponding reaction mechanism of Dy2 O3 -TiO2 powder mixtures under two types of ball-milling parameters were investigated. The annealing behavior of the ball-milled mixtures was also examined. 2. Experiments Powders of Dy2 O3 (cubic crystal structure) and TiO2 (rutile crystal structure) with an average particle diameter of 5 µm and 50 nm, respectively, were used as raw materials. The raw powders of Dy2 O3 and TiO2 were purchased from Beijing HWRK Chem Co., Ltd. (Beijing, China). The purity of the raw powders of both Dy2 O3 and TiO2 was 99.9%. Ball milling of the molar fraction Dy2 O3 -50% TiO2 (Dy2 O3 :TiO2 = 1:1) powder mixtures was carried out on an SFM-1 high-energy planetary ball mill at room temperature. Stainless steel balls that were 5 mm in diameter were used as the milling media. The ball-to-powder mass ratio was 10:1, and the rotational speed was 200 rpm and 500 rpm. No more than one weight percent stearic acid was added to the powder mixtures as a process control agent to prevent excessive cold welding and aggregation amongst the powder particles. During ball milling, a 5-min stopping interval was used after milling for 55 min to prevent excess heat generation, which has an obvious effect on the ball-milling procedure. The powder mixtures used for microstructural analysis were extracted from the loose powders in the steel can, not from powders adhered to the surface of the stainless balls or the steel can wall, after ball milling for 4, 12, 24, 48, and 96 h. After various milling times, a small amount of ball-milled powders taken from the container were characterized and analyzed by X-ray diffraction (XRD) on a Rigaku Ultima IV X-ray diffractometer (Rigaku, Tokyo, Japan) with Cu Kα radiation (λ = 0.1540598 nm) and transmission electron microscopy (TEM) on a JEM-2100 instrument (JEOL, Tokyo, Japan). Analysis was also carried out for ball-milled powder mixtures annealed at different temperatures. The grain size was calculated using Suryanarayana and Grant Norton’s formula [20]. Br cos θ =

Kλ + η sin θ L

(1)

Materials 2017, 10, 19

3 of 12

where, K is a constant (with a value of 0.9); λ is the wavelength of the X-ray radiation; L and η are the grain size and internal strain, respectively; and θ is the Bragg angle. Br is the full width at half-maximum (FWHM) of the diffraction peak after instrumental correction and can be calculated from the following equation: B = Br + Bs (2) where, B and BS are the FWHM of the broadened Bragg peaks and the standard sample’s Bragg peaks, respectively. The ball-milled mixtures and annealed mixtures were first put in ethyl alcohol, and then adequately dispersed by ultrasonic vibration. A carbon-coated copper grid was used to collect the dispersed powders in the ethyl alcohol and then dried by ultraviolet lamp. After that, the prepared samples were observed by TEM. Differential scanning calorimetry (DSC) was used to analyze the thermal behavior of ball-milled powders at a 5 ◦ C/min heating rate in argon atmosphere using a SAT 449C instrument (NETZSCH, Bavarian State, Germany ). The powder mixtures milled for 96 h were annealed at temperatures ranging from 700 to 1150 ◦ C in a tube furnace under atmospheric conditions. The heating and cooling rates were both 5 ◦ C/min. 3. Results and Discussion The XRD patterns of Dy2 O3 -TiO2 powder mixtures milled at 500 rpm and 200 rpm for different times are shown in Figure 1. The XRD results show that the crystal structure of the original Dy2 O3 phase and TiO2 phase are cubic and rutile, respectively. At the condition of 500 rpm, the diffraction peaks of cubic Dy2 O3 and TiO2 broadened significantly and reduced in intensity with increased milling time. The broadening of the X-ray diffraction peaks is associated with the refinement in grain size and lattice distortions. Meanwhile, the diffraction peaks of monoclinic Dy2 O3 can be observed in the X-ray patterns as indicated by the black inverted triangles in Figure 1a. Ball milling induces a Dy2 O3 phase transformation from the cubic to the monoclinic crystal structure. A broad, singular diffraction peak is also present that indicates the formation of the amorphous phase during ball milling. Interestingly, the amorphous peak is present at the location of the diffraction peak for the monoclinic Dy2 O3 phase, not at the location of the diffraction peak for the cubic Dy2 O3 phase, which indicates that the formed amorphous phase is derived from the monoclinic Dy2 O3 phase, not from the cubic Dy2 O3 phase. The transformation from cubic to monoclinic increases with increased milling time. After milling for 96 h, only the amorphous phase can be observed, which indicates that the monoclinic Dy2 O3 phase was fully converted to the amorphous phase. Additionally, this behavior indicates that no new compounds are synthesized during ball milling. Even if new compounds were formed in the milled powders, the amount is very low and does not reach the sensitivity range of the X-ray measurement. Therefore, in the present work, the evolution of Dy2 O3 -TiO2 powder mixtures is as follows: ball milling first induces the transformation of Dy2 O3 from the cubic to the monoclinic crystal phase, then monoclinic Dy2 O3 undergoes amorphization, and finally the powder mixtures completely transform to the amorphous phase. However, the ball-milling behavior of powder mixtures at 200 rpm is distinctly different from that at 500 rpm. The change of the diffraction peaks with increased milling time at 200 rpm is shown in Figure 1b. Although the diffraction peaks of cubic Dy2 O3 and TiO2 are also broadened and reduced in intensity with increased milling time, the diffraction peaks of TiO2 can be observed in the XRD spectrums and are not disappeared. After ball milling for 96 h, the intensity of diffraction peaks of Dy2 O3 and TiO2 are also high. The powder mixtures are not changed completely to amorphization. According to the shape of XRD diffraction spectrums, the effect of ball milling on powder mixtures after milling for 96 h at 200 rpm is only similar to that after milling for 4 h at 500 rpm. Therefore, at low ball-milling rotation speeds, the powder mixtures are only fined and homogenized. Our experimental results are different from the results of G. Garcia-Martinez [7]. In their research, ball milling induced a phase transformation in Dy2 O3 from cubic to monoclinic. However, a Dy2 TiO5 compound with a hexagonal crystal structure was formed simultaneously. The ball-milled powders

Materials 2017, 10, 19

4 of 12

Materials 2017, 10, 19 4 of 13 consisted of mixed phases of hexagonal Dy2 TiO5 and monoclinic Dy2 O3 . The Dy2 O3 -TiO2 powder mixtures did not completely transform to the amorphous phase, but instead produced the hexagonal powder mixtures did not completely transform to the amorphous phase, but instead produced the Dy2 TiO5 phase. This difference can be attributed to the different ball-milling conditions used in this hexagonal Dy2TiO5 phase. This difference can be attributed to the different ball-milling conditions research, special attention to the different During ball During milling,ball experimental used inwith this research, with special attention to ball-milling the differentfacilities. ball-milling facilities. milling, parameters such as rotation speed, ball-milling time, ball-milling media, and the ball-to-powder mass experimental parameters such as rotation speed, ball-milling time, ball-milling media, and the ballratio have an mass important influence on the ball-milled products even when using the when same using proportion to-powder ratio have an important influence on the ball-milled products even the and type of oxides. For example, Gajovi´ c reported that nanosized ZrTiO formed in ZrO2 -TiO 4 same proportion and type of oxides. For example, Gajović reported that nanosized ZrTiO4 formed in 2 powder mixtures, whereas only amorphous mixtures were obtained ball milling Stubiˇcin ar’s ZrO2-TiO 2 powder mixtures, whereas only amorphous mixtures wereduring obtained during ballinmilling work [21,22].work In addition, polymorphic transformation of Ln2 O3 was also Gd2 O3 -TiO Stubičar’s [21,22]. the In addition, the polymorphic transformation of Ln 2Oobserved 3 was alsoin observed in 2 and Y22OO3-TiO systems [23].systems [23]. Gd 2 and Y2O3-2TiO 2 powder 3 -2TiO 2 powder

Figure 1. X-ray diffraction patterns of the powder mixtures milled at (a) 500 rpm and (b) 200 rpm for Figure 1. X-ray diffraction patterns of the powder mixtures milled at (a) 500 rpm and (b) 200 rpm for various times, respectively. various times, respectively.

The variation of Dy2O3 grain size with ball-milling time at the rotational speeds of 500 rpm and The variation size with at the rotational speeds of 500 rpm and 2 O3 grain 200 rpm is shownofinDy Figure 2. Actually, theball-milling size of Dy2Otime 3 grain was calculated for Dy2O3 with cubic 200structure, rpm is shown in Figure 2. Actually, the size of Dy O grain wasball calculated Dy2 O3 awith cubic 2 3 because not for Dy2O3 with monoclinic structure, milling for induced phase structure, not for Dy becauseand ball simultaneously milling induced afrom phase transformation 3 with transformation in2 ODy 2O3 monoclinic from cubicstructure, to monoclinic monoclinic to in Dy cubic to monoclinic andthe simultaneously from to amorphous. is difficult amorphous. is difficult to calculate grain size of Dy 2O3 monoclinic with monoclinic structure. ItItcan be seen to 2 O3 from It calculate the grain results size of in Dya2 O monoclinic Itin can seen stage that ball milling in a that ball milling fast decrease of Dy2Ostructure. 3 grain size thebeinitial at both 500 results rpm and 3 with fast200 decrease of Dy sizeofincrystallite the initialsize stage at both 500 rpm andwith 200 rpm. The refinement rate rpm. The refinement rate is roughly logarithmic ball-milling time at 200 2 O3 grain rpm. After size 96 h is of roughly ball milling, the size ofwith Dy2O 3 grain is uptime to approximately 60 nm.96However, 500 of crystallite logarithmic ball-milling at 200 rpm. After h of ball at milling, diffraction peaks 2O3 with cubic60 structure are hardly observed in the XRD spectrum therpm, size the of Dy is upoftoDy approximately nm. However, at 500 rpm, the diffraction peaks of 2 O3 grain after milling for 24 h as shown in Figure 1a, especially, when the ball-milling time is over h. Dy2 O3 with cubic structure are hardly observed in the XRD spectrum after milling for 24 h as48 shown Therefore, the size of Dy 2O3 the withball-milling cubic structure before the 12 hsize of ball-milling time.cubic It in Figure 1a, especially, when timeisiscalculated over 48 h.only Therefore, of Dy2 O3 with can be seen that the grain size is quickly decreased. In addition, after same ball-milling time, the grain structure is calculated only before 12 h of ball-milling time. It can be seen that the grain size is quickly size of DyIn 2O3 phase at the 500 rpm is obviously smaller than that at the 200 rpm. The effect of ball decreased. addition, after same ball-milling time, the grain size of Dy2 O3 phase at the 500 rpm is milling on the grain refinement of powder mixtures at 500 rpm is significantly more intense than that obviously smaller than that at the 200 rpm. The effect of ball milling on the grain refinement of powder at 200 rpm. The size of Dy2O3 grain in the powder mixtures after milling for 4 h at 500 rpm is about mixtures at 500 rpm is significantly more intense than that at 200 rpm. The size of Dy2 O3 grain in 52 nm, which is smaller than that after milling for 96 h at 200 rpm (approximately 60 nm). the powder mixtures after milling for 4 h at 500 rpm is about 52 nm, which is smaller than that after milling for 96 h at 200 rpm (approximately 60 nm). The morphology evolution of Dy2 O3 -TiO2 powder mixtures with increasing ball-milling time at 500 rpm is shown in Figure 3. Both TiO2 and Dy2 O3 are brittle components, which are fragmented during ball milling and particle size reduces continuously as a consequence of the energy provided during ball milling. The morphology of large particles is changed significantly due to fracture, agglomeration, and deagglomeration processes. The morphology of the original powder mixtures consists of large-sized Dy2 O3 particles in micrometer size and small-sized TiO2 particles in nanometer. The shape of the powder particles is irregular. The line-scanning results of elemental Dy, Ti, and O in the characteristic position in Figure 3a are shown in Figure 3b and also inserted in Figure 3a. It can be seen that the small-sized particles are TiO2 component and the large-sized particles are Dy2 O3

Materials 2017, 10, 19

5 of 12

components from the variation of the elemental diffraction intensity. The brittle Dy2 O3 particles are fragmented by ball-powder-ball collisions, leading to a considerable reduction in the powder particle size and subsequent amorphization as milling time increases. After ball milling for 4 h, the size of particles decreases significantly. A large number of small size of Dy2 O3 particles in nanometer can be observed in the milled powders as shown in Figure 3c. The morphology of the powder mixtures is transformed to uniform, as shown in Figure 3c–g, where the ball-milling time ranges from 4 to 96 h. The morphologies demonstrate that the refining effects of the powder particles are proportional to the ball-milling time for the same rotational speed. After 96 h of ball milling, a large number of nanoparticles agglomerate to form a large-sized particle, as shown in Figure 3g. In addition, TiO2 particles disappear after ball milling for 96 h, as shown in Figure 1a. The surfaces of the Dy2 O3 particles in Figure 3g are clean compared with those in Figure 3a. It can be concluded that the particle size in the powder mixtures is refined to the nanoscale after ball milling for 96 h. The line-scanning results of elemental Dy, Ti, and O in the characteristic position in Figure 3g is shown in Figure 3h and also inserted in Figure 3g, which indicates these elements are uniformly distributed in the ball-milled particles according the variation of the elemental diffraction intensity. In addition, the morphology evolution of powder mixtures at an 200 rpm dose are not provided in the present work because the powder mixtures are only fined and homogenized according to the XRD results as shown in Figure 1b. The morphology of mixtures after ball milling for 96 h at 200 rpm is similar with that after ball milling forMaterials 4 h at 500 2017,rpm. 10, 19 5 of 13

Materials 2017, 10, 19

Figure 2. Curves of Dy2O3 grain size vs. ball-milling time. Figure 2. Curves of Dy2 O3 grain size vs. ball-milling time.

6 of 13

The morphology evolution of Dy2O3-TiO2 powder mixtures with increasing ball-milling time at 500 rpm is shown in Figure 3. Both TiO2 and Dy2O3 are brittle components, which are fragmented during ball milling and particle size reduces continuously as a consequence of the energy provided during ball milling. The morphology of large particles is changed significantly due to fracture, agglomeration, and deagglomeration processes. The morphology of the original powder mixtures consists of large-sized Dy2O3 particles in micrometer size and small-sized TiO2 particles in nanometer. The shape of the powder particles is irregular. The line-scanning results of elemental Dy, Ti, and O in the characteristic position in Figure 3a are shown in Figure 3b and also inserted in Figure 3a. It can be seen that the small-sized particles are TiO2 component and the large-sized particles are Dy2O3 components from the variation of the elemental diffraction intensity. The brittle Dy2O3 particles are fragmented by ball-powder-ball collisions, leading to a considerable reduction in the powder particle size and subsequent amorphization as milling time increases. After ball milling for 4 h, the size of particles decreases significantly. A large number of small size of Dy2O3 particles in nanometer can be observed in the milled powders as shown in Figure 3c. The morphology of the powder mixtures is transformed uniform, asresults shownofinthe Figure 3c–g, wheremilled the ball-milling time rangesball-milling from 4 to 96 h. Figure 3.to SEM analysis powder mixtures at 500 rpm for different Figure 3. SEM analysis results of the powder mixtures milled at 500 rpm for different ball-milling times. The images showing thethat morphology of ball-milled for (a)particles 0 h; (c) 4 h; h; (e) 24 to The morphologies demonstrate the refining effects ofmixtures the powder are(d)12 proportional times. The images showing the morphology of ball-milled mixtures for (a) 0 h; (c) 4 h; (d)12 h; (e) 24 h; h; (f) 48 h; and (g) 96 h; (b,h) are EDS analysis results of ball-milled particles in (a,g), respectively. the ball-milling time for the same rotational speed. After 96 h of ball milling, a large number of (f) 48 h; and (g) 96 h; (b,h) are EDS analysis results of ball-milled particles in (a,g), respectively. nanoparticles agglomerate to form a large-sized particle, as shown in Figure 3g. In addition, TiO2 Figure 4 showsafter TEM images area diffraction particles disappear ball millingand for corresponding 96 h, as shownselected in Figure 1a.electron The surfaces of the(SAED) Dy2O3 patterns in of Figure Dy2O3-TiO 2 powder mixtureswith milled forin4 Figure h and 96 After for that 4 h, the nano-sized particles 3g are clean compared those 3a. h. It can bemilling concluded particle ball-milled powdermixtures particles isaggregate form large-sized as shown inThe Figure 4a, due to size in the powder refined totothe nanoscale afterparticles, ball milling for 96 h. line-scanning the high active surface energy created upon ball milling. The size of the original TiO 2 particles is results of elemental Dy, Ti, and O in the characteristic position in Figure 3g is shown in Figure 3h and approximately 50 nm. From the XRD results, it can be observed that ball milling leads to TiO 2 particle also inserted in Figure 3g, which indicates these elements are uniformly distributed in the ball-milled refinement, dissolution, and finally disappearance after 96 h. Therefore, the main particles presented

Materials 2017, 10, 19

6 of 12

Figure 4 shows TEM images and corresponding selected area electron diffraction (SAED) patterns of Dy2 O3 -TiO2 powder mixtures milled for 4 h and 96 h. After milling for 4 h, nano-sized ball-milled powder particles aggregate to form large-sized particles, as shown in Figure 4a, due to the high active surface energy created upon ball milling. The size of the original TiO2 particles is approximately 50 nm. From the XRD results, it can be observed that ball milling leads to TiO2 particle refinement, dissolution, and finally disappearance after 96 h. Therefore, the main particles presented in the TEM image are Dy2 O3 . The bright zones near the edge of the powder particles are the thin areas where the electron beam penetrates. The dark areas in the images of the powder particles are the thick areas where the electron beam rarely penetrates. Indexing and analyzing the ring-shaped SAED pattern taken from the area denoted by the letter “A” indicates that the Dy2 O3 grains are already nanocrystalline. The diffraction spots coming from cubic Dy2 O3 , monoclinic Dy2 O3 , and TiO2 grains are present in the SAED pattern in Figure 4b. The diffraction halo in the SAED pattern also indicates the formation of an amorphous phase during high-energy ball milling. Materials 2017, 10, 19 7 of 13

Figure 4. 4. TEM TEM analysis analysis results results of of the the mixtures mixtures ball ball milled milled at at 500 500 rpm: rpm: (a) (a) the the bright bright field field TEM TEM image image Figure and (b) (c)(c) Bright field TEM image andand (d) and (b) corresponding corresponding SAED SAEDpattern patternofofmixtures mixturesmilled milledfor for4 h; 4 h; Bright field TEM image corresponding SAED pattern of mixtures milled for 96 h. (d) corresponding SAED pattern of mixtures milled for 96 h.

Monoclinic Ln2O3 is initially formed in the mechanical alloying of lanthanum titanate or After ball milling for 96 h, the ball-milled powders agglomerate to form large particles with large dititanate. In Moreno’s research, the formation of Gd2(Ti(1−y)Zry)2O7 pyrochlores occurred in the final thicknesses such that the electron beam rarely penetrates, showing as a dark color in the TEM image step of ball milling starting from an amorphous matrix of Gd2O3, TiO2, and ZrO2 [23]. However, in in Figure 4c. It is difficult to observe the microstructure of the agglomerated particles. The SAED the present work, after 96 h of ball milling, the monoclinic Dy2O3 phase could not be detected and pattern from the area marked with the letter “B” indicates that the powder mixtures are almost was completely transformed to the amorphous phase. Moreover, dysprosium titanate also could not completely converted to the amorphous phase, although sporadic diffraction spots are also present be detected. To investigate the sintering behavior of the ball-milled powder mixtures, DSC was in this pattern. The atom arrangement in the amorphous phase is disordered over long distances, carried out on the powder mixtures milled for 96 h at test temperatures ranging from 200 to 1200 °C. but ordered over short distances. As grain size decreases, the number of atoms at the grain boundaries The DSC curve in Figure 5 shows one exothermic peak close to 880 °C and one endothermic peak increases. The proportion of atoms in the crystal volume relative to the crystal boundary decreases. close to 1145 °C. An exothermic peak in a DSC curve can generally be attributed to a transition from Schwarz and Koch noted that the formation of an amorphous phase in as-milled powders was similar disordered to ordered, the recrystallization of original components from the amorphous phase or the to the amorphization that occurs during the isothermal annealing of crystalline metallic thin films [24]. formation of a new compound from the ball-milled amorphous powders. Therefore, subsequent XIn high-energy ball milling, the intense deformation accelerates interdiffusion, and the large defect ray analysis of the ball-milled powders annealed at different temperatures is used to further analyze density increases the free energy of the components in the mixture to form an amorphous product. occurrences in the heating process in more detail. The powder mixtures transform completely to the Monoclinic Ln2 O3 is initially formed in the mechanical alloying of lanthanum titanate or dititanate. amorphous state after ball milling for 96 h. Therefore, it seems feasible that the transition from In Moreno’s research, the formation of Gd2 (Ti(1−y) Zry )2 O7 pyrochlores occurred in the final step of ball disordered to ordered produced the exothermic peak in the DSC curve. In fact, only the amorphous peak is observed in the XRD pattern of the 96 h ball-milled powder after annealing for 24 h at 700 °C. No diffraction peaks for Dy2O3 or TiO2 are detected. Even with prolonged annealing time, the XRD results are the same. Therefore, the exothermic peak in the DSC curve should be related to the new phase generated from the amorphous mixtures.

Materials 2017, 10, 19

7 of 12

milling starting from an amorphous matrix of Gd2 O3 , TiO2 , and ZrO2 [23]. However, in the present work, after 96 h of ball milling, the monoclinic Dy2 O3 phase could not be detected and was completely transformed to the amorphous phase. Moreover, dysprosium titanate also could not be detected. To investigate the sintering behavior of the ball-milled powder mixtures, DSC was carried out on the powder mixtures milled for 96 h at test temperatures ranging from 200 to 1200 ◦ C. The DSC curve in Figure 5 shows one exothermic peak close to 880 ◦ C and one endothermic peak close to 1145 ◦ C. An exothermic peak in a DSC curve can generally be attributed to a transition from disordered to ordered, the recrystallization of original components from the amorphous phase or the formation of a new compound from the ball-milled amorphous powders. Therefore, subsequent X-ray analysis of the ball-milled powders annealed at different temperatures is used to further analyze occurrences in the heating process in more detail. The powder mixtures transform completely to the amorphous state after ball milling for 96 h. Therefore, it seems feasible that the transition from disordered to ordered produced the exothermic peak in the DSC curve. In fact, only the amorphous peak is observed in the XRD pattern of the 96 h ball-milled powder after annealing for 24 h at 700 ◦ C. No diffraction peaks for Dy2 O3 or TiO2 are detected. Even with prolonged annealing time, the XRD results are the same. Therefore, the exothermic peak in the DSC curve should be related to the new phase generated from the amorphous mixtures. Materials 2017, 10, 19 8 of 13

Figure -TiO22 powder powder mixtures mixtures milled milled for for 96 96 h. h. Figure5.5.DSC DSCcurve curveof ofDy Dy22O O33-TiO

The powder forfor 96 96 h were annealed for 3for h at3800, and 1150and °C. The powdermixtures mixturesmilled milled h were annealed h at900, 800,1000, 900,1050, 1000,1100, 1050, 1100, ◦ The XRD results of the annealed powder mixtures are shown in Figure 6a,b. Several diffraction peaks 1150 C. The XRD results of the annealed powder mixtures are shown in Figure 6a,b. Several diffraction different from the diffraction peakspeaks of Dy2of O3Dy (cubic and monoclinic crystalcrystal structure) and TiO 2 (rutile peaks different from the diffraction and monoclinic structure) and TiO2 2 O3 (cubic structure) are observed, whichwhich indicates new components withwith crystal structures generated from the (rutile structure) are observed, indicates new components crystal structures generated from amorphous mixtures. This experimental phenomenon to the amorphous mixtures. This experimental phenomenonofofsynthesizing synthesizingnew newcompounds compounds is is similar to Stubičar’s researchin in which the orthorhombic phase was generated from the highStubiˇ car’s research which the orthorhombic ZrTiO4ZrTiO phase4 was generated from the high-temperature temperature annealing ofmixtures amorphous mixtures theof ball milling of2 apowder ZrO2-TiO 2 powder annealing of amorphous formed fromformed the ballfrom milling a ZrO system [21] 2 -TiO system [21] andwith consistent Khor’s that zirconia was from produced from annealing amorphous and consistent Khor’swith results thatresults zirconia was produced annealing amorphous mixtures mixtures formed from the ball of an equimolar ZrSiO 43 and Al2O3system powder[25]. system [25]. formed from the ball milling of milling an equimolar ZrSiO4 and Al2 O powder ◦ , hexagonal According to the XRD standard database, cubic Dy 2 TiO 5 {111} presents at 2θ = 30.043°, hexagonal According to the XRD standard database, cubic Dy2 TiO5 {111} presents at 2θ = 30.043 ◦ ◦, Dy22TiO55 {102} DyDy 2TiO 5 {201} presents at 2θ and Dy {102} presents presentsatat2θ 2θ= =32.411°, 32.411 orthorhombic , orthorhombic presents at =2θ29.554°, = 29.554 2 TiO 5 {201} ◦ pyrochlore Dy2Ti 7 {111} presents at 2θ the; the difference in the above diffraction angles is and pyrochlore Dy2O presents at =2θ30.698°; = 30.698 difference in the above diffraction angles 2 Ti 2 O7 {111} not very powder mixtures mixtures is not verylarge. large.Three Threediffraction diffractionpeaks peaksare areobserved observedin inthe the XRD XRD patterns patterns of the powder ◦ C. The main diffraction peak representing the crystalline annealed for for 33 hh at 800, 900, and 1000 °C. crystalline phase phase annealed is at 2θ = 30.0°, as shown in Figure 6b. Therefore, the newly formed product in the powder system annealed between 800 °C and 1000 °C is not pyrochlore Dy2Ti2O7. This result is different than that found in Amit Sinha’s research in which pyrochlore Dy2Ti2O7 was initially created in the formation of Dy2TiO5 [19]. In their research using an equimolar Dy2O3-TiO2 system, the chemical reaction of Dy2O3 and TiO2 initially formed pyrochlore Dy2Ti2O7, and then Dy2Ti2O7 reacted with the remaining

After annealing for 3 h at 1100 °C, orthorhombic Dy2TiO5 and a small amount of pyrochlore Dy2Ti2O7 and cubic Dy2O3 are detected; notably, cubic Dy2TiO5 is not observed in the ball-milled powder mixtures. The change of the grain size of the main characteristic phase in the annealed powder mixtures with annealing Materials 2017, 10, 19temperature is shown in Figure 6c. According to the above results, the grain size 8 of of 12 the intermediate phase and orthorhombic Dy2TiO5 are calculated using Suryanarayana and Grant Norton’s formula. It can be seen that the grain size increases with increasing annealing temperature. ◦ is at grain 2θ = 30.0 shown in Figurephase 6b. Therefore, thenm newly formed product in the powder The size ,ofasthe intermediate is about 27 in the powder mixtures annealed forsystem 3 h at ◦ C and 1000 ◦ C is not pyrochlore Dy Ti O . This result is different than that annealed between 800 2 2 7 for 3 h at 1050 °C. Because the 800 °C and is about 275 nm in the powder mixtures annealed found in Amitphase Sinha’s research in pyrochlore created in the formation 2 O7 was intermediate transforms to which orthorhombic Dy2Dy TiO25Tiwhen theinitially annealing temperature is above of Dy TiO [19]. In their research using an equimolar Dy O -TiO system, the chemical reaction of 2 5 grain size of orthorhombic Dy2TiO5 is calculated 2 3 at the 2 annealing temperature ranging 1050 °C, the Dy O and TiO initially formed pyrochlore Dy Ti O , and then Dy Ti O reacted with the remaining 2 31050 °C to 2 1150 °C. The grain size of orthorhombic 2 2 7 2 5 2is 7about 43 and 380 nm after from Dy2TiO Dy Dy21150 TiO5°C, . respectively. 2 O3 to form annealing for 3orthorhombic h at 1050 °C and

Figure 6. 6. XRD XRD patterns patterns of ofthe theball-milled ball-milledpowder powdermixtures mixturesannealed annealedforfor 3 at h at various temperatures Figure 3h various temperatures at at diffraction angles 2θ ranging from (a) 20° to 65° and (b) 28° to 31°; (c) Curves of the grain size of ◦ ◦ ◦ ◦ diffraction angles 2θ ranging from (a) 20 to 65 and (b) 28 to 31 ; (c) Curves of the grain size of main main characteristic phase vs. annealing temperature. The powder mixtures were previously milled characteristic phase vs. annealing temperature. The powder mixtures were previously milled for 96 h for500 96 rpm. h at 500 rpm. at

In addition, the pressure created during ball milling is not high enough to transform the crystal According to the equilibrium phase diagram, the Dy2 TiO5 phase has three crystal structure types structure of Dy2TiO5 from orthorhombic to 1350 hexagonal or from hexagonal to cubic. The average ◦C 1680 ◦ C ↔ cubic [15]. depending thecontact temperature: Transformation from pressure onon the surfaceorthorhombic of two colliding↔ millhexagonal balls is approximately 8.5 GPa [26], which is far the high-temperature phase to the low-temperature phase is an exothermic process. For example, below the 100 GPa needed to induce a pressure wave to cause the Gd2TiO5 phase transformation from the transformation of Gd hexagonal to orthorhombic produced an exothermic 2 TiO5tofrom the polymorphic low-temperature orthorhombic phase the high-temperature hexagonal phase [27]. Therefore, peak in the DSC curve [7]. However, in the present work, there is only one exothermic peakinto in the the intermediate phase is not produced by collision pressure. Further investigation is needed the DSC curve. Although the diffraction peaks in the XRD patterns match well with the diffraction peaks cause of the formation of the intermediate phase. in theTo standard of cubic Dy2 TiO5 , the generated phase should powder not be cubic Dy2 TiO further pattern investigate the annealing behavior of the ball-milled mixtures, the5 because powder low-temperature phases hexagonal TiO5900, were not detected after annealing mixtures milled for 96 h of at orthorhombic 200 rpm were and sintered for 3 hDy at2800, 1000, 1050, 1100, and 1150 °C. ◦ C for annealing times ranging from several minutes to over temperatures from 700powder to 1000 mixtures The phase evolutionranging of the annealed identified by XRD analysis is shown in Figure 7. ◦ C. It is 3Under h; additionally, as mentioned above, the temperature of cubic Dy2 TiO 5 is over 1680and these ball-milling conditions, theformation Dy2O3-TiO 2 powder mixtures are homogenized, the not possible to achieve such a high temperature during ball milling. Therefore, the high-temperature polymorphic transformation of Dy2O3 from cubic to monoclinic is not observed in Figure 7a. In the phase of cubicof Dythe not be produced. the1000 generated phase is an intermediate 2 TiO 5 should XRD pattern powder mixtures annealed Instead, for 3 h at °C, diffraction peaks for Dy2Ophase 3 and that has a similar crystal structure to cubic Dy TiO and is a metastable state that phase 2 5 Dy2Ti2O7 phase are observed. However, the main diffraction peak for orthorhombic Dy2transforms TiO5 is not ◦ C. After annealing for 3 h to orthorhombic annealing is above 2 TiO 5 when the detected. Under Dy these conditions, the annealedtemperature powder mixtures are1050 composed of cubic Dy2O3 and ◦ at 1100 C, orthorhombic and a small amount pyrochlore Dy cubic Dy2°C O3and are 2 TiO5 mixtures 2 Ti 2 O75 and pyrochlore Dy2Ti2O7. The Dy powder generate theoforthorhombic Dy 2TiO phase at 1050 detected; notably, cubicDy Dy the ball-milled powder mixtures. 2 TiO 5 is not observed are composed of cubic 2O 3, pyrochlore Dy2Ti2Oin 7, and orthorhombic Dy2TiO5. The powder mixtures The change of the grain size of the main characteristic in the annealed powder mixtures are almost completely transformed to orthorhombic Dy2TiOphase 5 after annealing at 1150 °C for 3 h. The with annealing temperature is shown in Figure 6c. According to the above results, the grain size of the diffraction peak intensity of orthorhombic Dy2TiO5 gradually increases with increasing annealing intermediate and orthorhombic calculated using Suryanarayana and Grant Norton’s 2 TiO5 arepeak temperature,phase and simultaneously, theDy diffraction intensity of Dy 2O3 and Dy2Ti2O7 decreases with formula. It can be seen that the grain size increases with increasing annealing temperature. increasing annealing temperature. For powder mixtures ball milled for 96 h at 200 The rpm,grain the size of the intermediate is about 27 nm in the powder mixtures annealed for 3presented h at 800 ◦in C and evolutionary behavior atphase various annealing temperatures is consistent with the data Ref. ◦ C. Because the intermediate phase is about 275 nmAmit in theSinha powder mixtures for 3 hreaction at 1050 of [19], in which reported thatannealed the chemical Dy2O3 and TiO2 initially formed ◦ C, the grain size transforms orthorhombic Dy2 TiO the annealing temperature above 1050 pyrochlore to Dy 2Ti2O7, and then Dy25Tiwhen 2O7 reacted with the remainingisDy 2O3 to form orthorhombic of orthorhombic Dy2 TiO5 is calculated at the annealing temperature ranging from 1050 ◦ C to 1150 ◦ C. The grain size of orthorhombic Dy2 TiO5 is about 43 and 380 nm after annealing for 3 h at 1050 ◦ C and 1150 ◦ C, respectively. In addition, the pressure created during ball milling is not high enough to transform the crystal structure of Dy2 TiO5 from orthorhombic to hexagonal or from hexagonal to cubic. The average

Materials 2017, 10, 19

9 of 12

pressure on the contact surface of two colliding mill balls is approximately 8.5 GPa [26], which is far below the 100 GPa needed to induce a pressure wave to cause the Gd2 TiO5 phase transformation from the low-temperature orthorhombic phase to the high-temperature hexagonal phase [27]. Therefore, the intermediate phase is not produced by collision pressure. Further investigation is needed into the cause of the formation of the intermediate phase. To further investigate the annealing behavior of the ball-milled powder mixtures, the powder mixtures milled for 96 h at 200 rpm were sintered for 3 h at 800, 900, 1000, 1050, 1100, and 1150 ◦ C. The phase evolution of the annealed powder mixtures identified by XRD analysis is shown in Figure 7. Under these ball-milling conditions, the Dy2 O3 -TiO2 powder mixtures are homogenized, and the polymorphic transformation of Dy2 O3 from cubic to monoclinic is not observed in Figure 7a. In the XRD pattern of the powder mixtures annealed for 3 h at 1000 ◦ C, diffraction peaks for Dy2 O3 and Dy2 Ti2 O7 phase are observed. However, the main diffraction peak for orthorhombic Dy2 TiO5 is not detected. Under 3 and Materials 2017, 10, 19 these conditions, the annealed powder mixtures are composed of cubic Dy2 O10 of 13 pyrochlore Dy2 Ti2 O7 . The powder mixtures generate the orthorhombic Dy2 TiO5 phase at 1050 ◦ C and are5composed cubic Dy2 O3 , pyrochlore Dy2 Tiis2 O Dy2 TiOin Theannealed powder Dy2TiO . However,ofthis experimental phenomenon different from that observed 7 , and orthorhombic 5 . the ◦C mixtures are almost completely transformed orthorhombic annealing at 1150 powder mixtures milled for 96 h at 500 rpm, astoshown in Figure Dy 6, due the initial conditions of the 2 TiOto 5 after for 3 h. The diffraction ball-milling mixtures. peak intensity of orthorhombic Dy2 TiO5 gradually increases with increasing annealing temperature, and simultaneously, the diffraction intensity ofpowder Dy2 O3 and Dy2 Tiwith The change of the grain size of main characteristic phasepeak in the annealed mixtures 2 O7 decreases increasing temperature. powder mixtures ball for 96 h at Dy 2002Ti rpm, annealingwith temperature is annealing shown in Figure 7c. TheFor grain size of the cubic Dymilled 2O3, pyrochlore 2O7 the behavior various annealing temperatures is consistent with the dataformula. presented in andevolutionary orthorhombic Dy2TiOat 5 are calculated using Suryanarayana and Grant Norton’s The Ref. reported that reaction Dy2 O3 and TiO2 initially formed grain[19], sizeinofwhich cubic Amit Dy2OSinha 3 and pyrochlore Dythe 2Ti2chemical O7 increases withof increasing annealing temperature. pyrochlore Dy52is Ti2detected O7 reacted withpowder the remaining orthorhombic Because theDy orthorhombic Dy2TiO in the mixturesDy annealed at 1050 °C for 3 h, 2 Ti2 O7 , and then 2 O3 to form Dy TiO5 . size However, this experimental is different from that observed in the2grain of orthorhombic Dy2TiO5 phenomenon is calculated when the annealing temperature is the overannealed 1050 °C. powder mixtures milled for 96 hDy at2TiO 500 5rpm, as shown Figure 6, due the initial3conditions ofand the The grain size of orthorhombic is about 38 nm in and 395 nm aftertoannealing h at 1050 °C ball-milling mixtures. 1150 °C, respectively.

Figure 7. 7. XRD XRD patterns patterns of ofthe theball-milled ball-milledpowder powdermixtures mixturesannealed annealedfor for 3h various temperatures Figure 3h atat various temperatures at at diffraction angles 2θ ranging from (a) 20° to 65° and (b) 28° to 31.0°; (c) Curves of the grain size size of of ◦ ◦ ◦ ◦ diffraction angles 2θ ranging from (a) 20 to 65 and (b) 28 to 31.0 ; (c) Curves of the grain characteristic phase phase vs. vs. annealing annealing temperature. temperature. The The powder were previously previously milled milled for for 96 96 h h characteristic powder mixtures mixtures were at 200 200 rpm. rpm. at

Figure 8a,b shows the bright field TEM image and corresponding SAED pattern of the ballThe change of the grain size of main characteristic phase in the annealed powder mixtures with milled Dy2O3-TiO2 powder mixtures annealed at 1000 °C for 3 h, respectively. The powder mixtures annealing temperature is shown in Figure 7c. The grain size of the cubic Dy2 O3 , pyrochlore Dy2 Ti2 O7 are previously milled for 96 h at 500 rpm. After annealing for 3 h, the grain size of powder mixtures and orthorhombic Dy2 TiO5 are calculated using Suryanarayana and Grant Norton’s formula. The grain is kept in nanometer scale, which can be supported by the corresponding SAED pattern taken from size of cubic Dy2 O3 and pyrochlore Dy2 Ti2 O7 increases with increasing annealing temperature. Because the region marked letter “A” in Figure 8a. The diffraction ring is a typical SAED pattern of the orthorhombic Dy2 TiO5 is detected in the powder mixtures annealed at 1050 ◦ C for 3 h, the nanocrystal materials. After analyzing and indexing the ring-shaped SAED pattern, it is indicated grain size of orthorhombic Dy2 TiO5 is calculated when the annealing temperature is over 1050 ◦ C. that this SAED pattern belongs to the intermediate Dy2TiO5 phase that has a similar crystal structure The grain size of orthorhombic Dy2 TiO5 is about 38 nm and 395 nm after annealing 3 h at 1050 ◦ C and to cubic Dy2TiO5, which is in accord with the XRD results as shown in Figure 6. The small-sized 1150 ◦ C, respectively. powders agglomerate to form large particles with large thicknesses as shown in Figure 8a. The bright Figure 8a,b shows the bright field TEM image and corresponding SAED pattern of the ball-milled zones near the edge of the powder particles is the◦ thin area where the electron beam penetrates. The Dy2 O3 -TiO2 powder mixtures annealed at 1000 C for 3 h, respectively. The powder mixtures are dark area in the image of the powder particles is the thick area where the electron beam rarely penetrates. After annealing, the amorphous ball-milled powder mixtures are changed to the intermediate Dy2TiO5 phase with crystal structure. Figure 8c,d is the bright field TEM image and corresponding SAED pattern of the annealed Dy2O3-TiO2 powder mixtures that were previously milled for 96 h at 200 rpm. Indexing and analyzing the ring-shaped SAED pattern taken from the area

Materials 2017, 10, 19

10 of 12

previously milled for 96 h at 500 rpm. After annealing for 3 h, the grain size of powder mixtures is kept in nanometer scale, which can be supported by the corresponding SAED pattern taken from the region marked letter “A” in Figure 8a. The diffraction ring is a typical SAED pattern of nanocrystal materials. After analyzing and indexing the ring-shaped SAED pattern, it is indicated that this SAED pattern belongs to the intermediate Dy2 TiO5 phase that has a similar crystal structure to cubic Dy2 TiO5 , which is in accord with the XRD results as shown in Figure 6. The small-sized powders agglomerate to form large particles with large thicknesses as shown in Figure 8a. The bright zones near the edge of the powder particles is the thin area where the electron beam penetrates. The dark area in the image of the powder particles is the thick area where the electron beam rarely penetrates. After annealing, the amorphous ball-milled powder mixtures are changed to the intermediate Dy2 TiO5 phase with crystal structure. Figure 8c,d is the bright field TEM image and corresponding SAED pattern of the annealed Dy2 O3 -TiO2 powder mixtures that were previously milled for 96 h at 200 rpm. Indexing and analyzing the ring-shaped SAED pattern taken from the area denoted by the letter “B” indicates that the particles are composed of cubic Dy2 O3 and pyrochlore Dy2 Ti2 O7 , which is accord with the XRD results of the powder mixtures annealed at 1000 ◦ C for 3 h as shown in Figure 7. This experimental result is different from that observed in the annealed powder mixtures milled for 96 h at 500 rpm. Materials 2017, 10, 19 11 of 13

Figure 8. The bright field TEM images and corresponding SAED patterns of the ball-milled Dy2O3Figure 8. The bright field TEM images and corresponding SAED patterns of the ball-milled Dy2 O3 -TiO2 TiO2 powder mixtures annealed at◦ 1000 °C for 3 h, (a,b) the powder mixtures are previously milled powder mixtures annealed at 1000 C for 3 h, (a,b) the powder mixtures are previously milled for 96 h for 96 h at 500 rpm; (c,d) the powder mixtures are previously milled for 96 h at 200 rpm. at 500 rpm; (c,d) the powder mixtures are previously milled for 96 h at 200 rpm.

4. Conclusions 4. Conclusions The microstructural evolution of Dy2O3-TiO2 powder mixtures during ball milling and postThe microstructural evolution of Dy2 O3 -TiO2 powder mixtures during ball milling and milled annealing was investigated using TEM, SEM, XRD, and DSC. The conclusions can be made as post-milled annealing was investigated using TEM, SEM, XRD, and DSC. The conclusions can be made follows: as follows: 1. The ball-milling parameters had a great effect on ball milling and the subsequent annealing process. 2. At 500 rpm rotation speeds, the mixtures were fined, homogenized, nanocrystallized, and then completely amorphized, and the crystal structure of Dy2O3 was transformed from cubic to monoclinic. The amorphous transformation resulted from monoclinic Dy2O3, not from cubic Dy2O3. However, at 200 rpm rotation speeds, the Dy2O3-TiO2 powder mixtures were only homogenized, and the polymorphic transformation of Dy2O3 from cubic to monoclinic was not observed. Meanwhile, the powder mixtures did not transform to the amorphous phase. 3. The powder mixtures milled for 96 h at 500 rpm were annealed for 3 h at a temperature range of

Materials 2017, 10, 19

1. 2.

3.

11 of 12

The ball-milling parameters had a great effect on ball milling and the subsequent annealing process. At 500 rpm rotation speeds, the mixtures were fined, homogenized, nanocrystallized, and then completely amorphized, and the crystal structure of Dy2 O3 was transformed from cubic to monoclinic. The amorphous transformation resulted from monoclinic Dy2 O3 , not from cubic Dy2 O3 . However, at 200 rpm rotation speeds, the Dy2 O3 -TiO2 powder mixtures were only homogenized, and the polymorphic transformation of Dy2 O3 from cubic to monoclinic was not observed. Meanwhile, the powder mixtures did not transform to the amorphous phase. The powder mixtures milled for 96 h at 500 rpm were annealed for 3 h at a temperature range of 800 to 1000 ◦ C. An intermediate phase with a crystal structure similar to that of cubic Dy2 TiO5 was synthesized, which was a metastable phase that transformed to orthorhombic Dy2 TiO5 when the annealing temperature was above 1050 ◦ C. However, the powder mixtures milled for 96 h at 200 rpm did not transform to the amorphous phase. The annealing behavior showed that the chemical reaction of Dy2 O3 with TiO2 initially formed pyrochlore Dy2 Ti2 O7 , and then Dy2 Ti2 O7 reacted with the remaining Dy2 O3 to form orthorhombic Dy2 TiO5 .

Acknowledgments: The work was supported by the National Natural Science Foundation of China through Grant No. 11305136. Author Contributions: Guang Ran conceived and designed the experiments; Jinhua Huang performed the experiments and analyzed the data; Qiang Shen conducted the TEM experiment; Jinhua Huang, Jianxin Lin, Penghui Lei, and Xina Wang wrote the paper under the supervision of Guang Ran. All authors contributed to the scientific discussion of the results and reviewed the manuscript. Conflicts of Interest: The authors declare no conflicts of interest.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Ran, G.; Zhou, J.E.; Xi, S.Q.; Li, P.L. Formation of nanocrystalline and amorphous phase of Al-Pb-Si-Sn-Cu powder during mechanical alloying. Mater. Sci. Eng. A 2006, 416, 45–50. Eckert, J.; Schultz, L.; Urban, K. Formation of quasicrystals by mechanical alloying. Appl. Phys. Lett. 1989, 55, 117–119. [CrossRef] Luo, Y.; Ran, G.; Chen, N.J.; Shen, Q.; Zhang, Y.L. Microstructural evolution, thermodynamics and kinetics of Mo-Tm2 O3 powder mixtures during ball milling. Materials 2016, 9, 834. [CrossRef] Kimura, H.; Kimura, M.; Takada, F. Development of an extremely high energy ball mill for solid state amorphizing transformations. J. Less Common Met. 1988, 40, 113–118. [CrossRef] Gilman, P.S.; Benjamin, J.S. Mechanical alloying. Annu. Rev. Mater. Sci. 1983, 13, 279–300. [CrossRef] Kong, L.B.; Ma, J.; Zhu, W.; Tan, O.K. Preparation of PMN-PT ceramics via a high-energy ball milling process. J. Alloys Compd. 2002, 335, 290–296. [CrossRef] G-Martinez, G.; MGonzalez, L.G.; Escalante-García, J.I.; Fuentes, A.F. Phase evolution induced by mechanical milling in Ln2 O3 :TiO2 mixtures (Ln = Gd and Dy). Powder Technol. 2005, 152, 72–78. [CrossRef] Fuentes, A.F.; Takacs, L. Preparation of multicomponent oxides by mechanochemical methods. J. Mater. Sci. 2013, 48, 598–611. [CrossRef] Suryanarayana, C. Mechanical alloying and milling. Prog. Mater. Sci. 2001, 46, 1–184. [CrossRef] Khakpour, Z.; Youzbashi, A.A.; Maghsoudipour, A.; Ahmadi, K. Synthesis of nanosized gadolinium doped ceria solid solution by high energy ball milling. Powder Technol. 2011, 214, 117–121. [CrossRef] Huang, J.H.; Ran, G.; Liu, T.J.; Shen, Q.; Li, N. Microstructure and Physical Properties of Tb2 TiO5 Neutron Absorber Synthesized by Ball Milling and Sintering. J. Mater. Eng. Perform. 2016, 25, 4266–4273. [CrossRef] Risovany, V.D.; Varlashova, E.E.; Suslov, D.N. Dysprosium titanate as an absorber material for control rods. J. Nucl. Mat. 2000, 281, 84–89. [CrossRef] Risovany, V.D.; Klochkov, E.P.; Varlashova, E.E. Hafnium and dysprosium titanate based control rods for thermal water-cooled reactors. At. Energy 1996, 81, 764–769. [CrossRef] Kermit, W.; Theilacker, J.S. Neutron Absorber Materials for Reactor Control; Naval Reactors Division of Reactor Development United States Atomic Energy Commission: Washington, DC, USA, 1962. Petrova, M.A.; Novikova, A.S.; Grebenshchikov, R.G. Polymorphism of rare earth titanates of Ln2 TiO5 composition. Izv. Akad. Nauk SSSR. Neorg. Mater. 1982, 18, 287–291.

Materials 2017, 10, 19

16. 17.

18. 19. 20. 21.

22. 23. 24. 25. 26. 27.

12 of 12

Jung, C.H.; Kim, C.J.; Lee, S.J. Synthesis and sintering studies on Dy2 TiO5 prepared by polymercarrier chemical process. J. Nucl. Mater. 2006, 354, 137–142. [CrossRef] Panneerselvam, G.; Venkata Krishnan, R.; Antony, M.P.; Nagarajam, K.; Vasudevan, T.; Vasuadeva Rao, P.R. Thermophysical measurements on dysprosium and gadolinium titanates. J. Nucl. Mater. 2004, 327, 220–225. [CrossRef] Kim, J.S.; Kim, H.S.; Jeong, C.Y. Phase study on the Dyx Tiy Oz pellets. In Proceedings of the Transactions of the Korean Nuclear Society Autumn Meeting, Busan, Korea, 27–28 October 2005. Sinha, A.; Prakash, B. Development of Dysprosium Titanate Based Ceramics. J. Am. Ceram. Soc. 2005, 88, 1064–1066. [CrossRef] Suryanarayana, C.; Norton, M.G. X-ray Diffraction: A Practical Approach; Plenum Press: New York, NY, USA, 1998. Stubiˇcar, M.; Bermanec, V.; Stubiˇcar, N.; Dudrnovski, D.; Drumes, D. Microstructure evolution of an equimolar powder mixture of ZrO2 -TiO2 during high-energy ball-milling and post-annealing. J. Alloys Compd. 2001, 316, 316–320. [CrossRef] Gajovi´c, A.; Djerdj, I.; Furi´c, K.; Schlögl, R.; Su, D.S. Preparation of nanostructured ZrTiO4 by solid state reaction in equimolar mixture of TiO2 and ZrO2 . J. Am. Ceram. Soc. 2006, 89, 2196–2205. [CrossRef] Moreno, K.J.; Rodrigo, R.S.; Fuentes, A.F. Direct synthesis of A2 (Ti(1−y) Zry )2 O7 (A = Gd3+ , Y3+ ) solid solutions by ball milling constituent oxides. J. Alloys Compd. 2005, 390, 230–235. [CrossRef] Koch, C.C.; Cavin, O.B. Preparation of “amorphous”Ni60Nb40 by mechanical alloying. Appl. Phys. Lett. 1986, 49, 146–148. [CrossRef] Khor, K.A.; Li, Y. Effect of mechanical alloying on the reaction sintering of ZrSiO4 and Al2 O3 . Mater. Sci. Eng. A 1998, 256, 271–279. [CrossRef] Huang, J.H.; Lu, J.Q.; Ran, G.; Chen, N.J.; Qu, P.D. Formation of nanocrystalline and amorphization phase of Fe-Dy2 O3 powder mixtures induced by ball milling. J. Mater. Res. 2016. [CrossRef] Shcherbakova, L.G.; Kolesnikov, A.V.; Breusov, O.N. Investigation of TiO2 -Ln2 O3 systems under shock wave action. Inorg. Mater. 1979, 15, 1724–1729. © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).