New Dispersants for the Preparation of Y2O3 Spherical Powders by Homogeneous Precipitation at High Temperature①

ZHANG D-Ci GUO Wng HUANG Ji-Qun CAO Yong-Ge②

?

New Dispersants for the Preparation of Y2O3Spherical Powders by Homogeneous Precipitation at High Temperature①

ZHANG Da-Caia, bGUO WangaHUANG Ji-QuanaCAO Yong-Gea②

a(350002)b(100049)

Spherical monodispersed and submicron-sized Y2O3powders were successfully synthesized through the urea homogeneous precipitation method adding PVA, PVP, or PVA and PVP compound (PVA/PVP) as the dispersant which generated no impurity phases after calcining. The productivity is up to 60% at 107 ± 2 ℃for 3.5 h in an oil bath. The structure, phase composition and evolution, morphology and specific surface area of Y2O3precursor and the calcined powders were explored by means of XRD, TG/DTA, FTIR, SEM, TEM and Micropore analyzer (BET). The spherical particle size of the powders calcined at 900 ℃ for 2 h was 330～350 nm. In this study, 15.5 Wt.% PVA, 8.5 Wt.% PVP or the mixture of both is in favor of enhancing the dispersity of the products. Based on what we have already achieved, it is rather significant to advance this research.

PVA, PVP, PVA/PVP, productivity;

1 INTRODUCTION

Possessing the high melting point, high thermal conductivity, chemical stability[1, 2], transparency over a wide wavelength range from violet to infrared light[3, 4], and low phonon energy, yttria (Y2O3) is a promising material for infrared domes, optical matrix for scintillation, high temperature windows, and component of semiconductor. Moreover, rare earth-doped Y2O3has been considered as a candidate host material for solid-state lasers for many years[5-7], which has attracted much attention of researchers.

As we know, the transmittance and laser per- formance of Y2O3transparent ceramics have greatly depended on the properties of initial Y2O3powders, such as purity, morphology, dispersity, particle size and size distribution. Based on this, spherical, monodispersed, fine and pure Y2O3powders are ideal choices for making Y2O3ceramics with high densification and transmittance. Great numbers of different techniques have been developed to synthesize rare earth oxide (REO) powders, like solid phase synthesis[8, 9], hydrothermal method[10, 11], sol-gel[12, 13], coprecipitation[14, 15]and homogeneous precipitation[16, 17]. Among these methods, homo- geneous precipitation (HP) is the optimal way to attain desired Y2O3powders. However, few papers have reported the productivity by HP. As a matter of fact, this method is so low-yield, which is partly due to its low reaction temperature in an aqueous medium (～90 ℃)[17], that its practical value is confined. Thus, improving reaction temperature is a significant means to increase powders yield with keeping other reaction conditions immutable. Nevertheless, the higher reaction temperature is, the bigger particle size of the precipitated precursor is, and the more severe agglomeration is. So, it is necessary to add dispersants to the solution to obtain dispersive and fine precursors. Traditionally, ammo- nium sulfate ((NH4)2SO4) is used as a dispersant because of the comparatively high electronegativity and decomposition temperature (～1100 ℃) of the sulfate ion (SO42-)[18-20]. (NH4)2SO4has indeed contributed to improving the dispersity of precursors and decreasing the size of particles at relatively low reaction temperature in an liquidous medium[21]. However, when increasing the reaction temperature to some degree, we found (NH4)2SO4causes more severe aggregation. Thus it is needed to substitute new dispersants for ammonium sulfate to improve dispersity of products and increasing productivity at high reaction temperature. After testing dozens of additives, we found poly(vinyl alcohol) (PVA) and poly(vinyl pyrrolidone) (PVP) are beneficial to improving the decentrality of products as surfac- tants.

In this research, taking a certain amount of PVA and PVP as the dispersants, we successfully synthe- sized pure uniform spherical Y2O3powders through a urea homogeneous precipitation method at high temperature (107 ± 2 ℃) for the oil bath. The pro- ductivity is up to 60%, which is much higher than the one at 90 ℃ (～10%). PVA and PVP are expected to have wide applications in preparing other rare earth oxide (REO) powders with high quality.

2 EXPERIMENTAL

2. 1 Materials and preparation of Y2O3powders

Yttria (99.99%), nitric acid (99.999%), urea (99.999%), (NH4)2SO4(99.997%), ethanol (analy- tical grade), PVA (average MW 67000, 88% alcoholyzed) and PVP (average MW 24000) were used as the raw starting materials. According to our previous studies, the appropriate additions of PVA and PVP are 15.5 Wt.% and 8.5 Wt.% of raw Y2O3powders, respectively. So, for comparison, Y2O3precursors were prepared by HP technique according to the following five different means: (1) without any dispersants; (2) only 5Wt.% (NH4)2SO4as the dispersant (weighing 5 Wt.% of raw Y2O3pow- ders)[21]; (3) only 15.5 Wt.% PVA as the dispersant; (4) only 8.5 Wt.% PVP as the dispersant; (5) the mixture of PVA (15.5 Wt.%) and PVP (8.5 Wt.%) as the dispersant. Firstly, proper amounts of raw Y2O3powders were dissolved in dilute nitric acid to make a nitrate solution in which the concentration of Y3+was kept at 0.04 mol/L. The concentration ratio of urea and Y3+was controlled at 15:1. Then, ethanol, 20 vol.% of the solution, and the relevant dispersant was added to the solution and the pH was adjusted to 6.0. After being stirred at room temperature for 1 h for homogenization, the mixed solution was heated in a Dimethyl silicone bath whose temperature was 107 ± 2 ℃ and held for 3.5 h. In fact, the temperature of the mixed solution was kept at 85 ± 2 ℃ after 0.5 h. Next, the solution was naturally cooled down to room temperature. Finally, the precursor was centrifuged and washed repeatedly with deionized water and ethanol to completely remove by-products of the reaction. After rinsing, the precursor was dried at 65 ℃ for 12 h and calcined at 900 ℃ for 2 h in a tube furnace in O2.

2. 2 Physical measurements

Phase identification was examined by X-ray diffraction analysis (XRD, MiniFlex600, Rigaku, Japan) and thermal analysis of the precursors was performed via thermo-gravimetric/differential thermal analysis (TG/DTA, STA449F3, Netzsch, Germany). Fourier transform infrared spectroscopy (FTIR, Vertex70, Bruker, Germany) of the pre- cursors was determined at room temperature and the specific surface area of the powders calcined at 900 ℃ was measured with Micropore analyzer (Asap 2020 C+M, Micromeritics, USA) by BET method in N2. The morphology of calcined powders was characterized through scanning electron microscope (SEM, JSM-6700F, JEOL, Japan) and Transmission electron microscope (TEM, JEM-2010, JEOL, Japan).

3 RESULTS AND DISCUSSION

To illustrate the effects of dispersants concisely, all dispersants (or additives) mentioned in the following discussion just only refer to the ones added to the solution in liquid-phase reactions.

3. 1 IR spectroscopy

According to the FTIR spectra of the virgin and various dispersant-added precursors in Fig. 1, we can get some important information about the effect of dispersants on the products. The wide peaks at ～3373 cm-1are due to O–H bend and the peaks at about 648, 696, and 756 cm-1are attributed to the split non-planar bending vibration of CO32-. The peak at ～1125 cm-1results from the residual SO42-after rinsing. The peaks at about 1408 and 1520 cm-1are assigned to the split anti-symmetrical stretching vibration of CO32-that is weakened by the absorbed PVA, as shown in the inset. Maybe PVA’s hydro- philic groups, -CH2OH-, improved the stretching symmetry of CO32-by absorbing the surrounding electrons. From the above analysis, the precursor is reasonably considered as the basic carbonate with crystal water and the formula is Y+2y(OH)3x(CO3)3y·nH2O.

Fig. 1. FTIR spectra of the precursors with different additives; Inset is the magnified image of FTIR spectra from 1230 to 1676 cm-1

3. 2 TG analyses

TG curves (Fig. 2) of the five different precursors show total weight loss of 39.6～41.4% up to 1050 ℃. Between 30～195 ℃, a weight of ～9% is mainly attributed to the evaporation of absorbed water and the release of molecular water. The subsequent loss from 195 to 800 ℃ is owing to the decomposition of Y(OH)3and Y2(CO3)3. The remainder of 58.6～60.4% does not change anymore between 800～1050 ℃, which reveals that Y(OH)3and Y2(CO3)3have decomposed into Y2O3utterly before 800 ℃. In addition, water loss of the precursor added with PVA, PVP, or the mixture of both is a little larger than the two others, seen in the inset. This is because the hydrophilic groups of PVA and PVP attached to the precursors absorbed more water.

3. 3 Structure description

Fig. 3a shows the XRD patterns of the virgin and various dispersant-added Y2O3powders calcined at 900 ℃ for 2 h. It reveals that, except for the (NH4)2SO4-added one, the precursors all completely transformed to Y2O3crystals at 900 ℃ and no other phases were detected. After calcining, the (NH4)2SO4-added powder has the impure phase Y2O2SO4because the decomposition temperature of SO42-is higher than 900 ℃. This result is consistent with the FTIR analysis (Fig. 1). Residual Y2O2SO4is rather pernicious for ceramics sintering, because itwill decompose and release gas during sintering process, then densification and transmittance of ceramics will be impaired severely. Structural refinement with the Rietveld method[22]using Fullprof Program was performed to analyze further the effect of PVA and PVP on the structure of the prepared powders. The results, seen in Table 1 and Fig. 3b, indicate good agreement between the observed XRD pattern and the calculated one based on pure Y2O3phase. This result means that after calcination the PVA/PVP decomposed and evapora- ted completely and brought forth no impurities. Besides, we can find that the ratio of O2-andY3+is larger than 3:2 (Table 1). The reason is that ambient O2entered the interstices of Y2O3crystals during the calcining process in O2.

Table 1. Rietveld Refinement Results of the PVA/PVP-added Y2O3Powder Calcined at 900 ℃ for 2 h Compared with the Pure Y2O3Powder

PowdersAtomsxyzOccupancyLattice constants Y1 (C3i)0.25000.25000.25000.1000a = b = c = 10.6039 ? Pure Y2O3Y2 (C2)0.467500.25000.3000a= b= g= 90° O10.10870.34780.11950.6000 Y1 (C3i)0.25000.25000.25000.0982a = b = c = 10.6091 ? PVA/PVP-added Y2O3Y2 (C2)0.468500.25000.3018a= b= g= 90° O10.10780.34480.11960.6085

Fig. 2. TG curves of the precursors with different additives; Inset is the magnified image of the selected part

(a)????????????(b)

Fig. 3. (a) XRD patterns of the powders calcined at 900 ℃ for 2 h with different additives. (b) Rietveld refinement pattern for the powder calcined at 900 ℃ for 2 h added with PVA/PVP as the dispersant; Inset is the unit cell structure of Y2O3revealing coordination environment of Y and O

3. 4 Morphology characteristics

The SEM and TEM morphologies of the virgin and various dispersant-added Y2O3powders calcined at 900 ℃ for 2 h are presented in Fig. 4. Except for the (NH4)2SO4-added one (Fig. 4c, 4d), particles of all other powders are spherical because of the low concentration of Y3+(0.04 M), and the particle size of these powders is 330～350 nm. However, the size of the aggregates formed by the reunite of particles is much different. As Fig. 4 shows, the powder added with PVA or PVP as the dispersant is less agglomerative than the virgin one, so is the powder added with PVA/PVP. The mechanism of surfactants that consist of hydrophilic and hydrophobic groups has been extensively studied. PVA is a nonionic surfactant and PVP is a cationic one whose hydrophilic groups are positive after ionization. After the precursor particles were formed, the hydrophobic groups adhered to the surfaces of particles and the hydrophilic groups stretched into the liquidous medium (Fig. 5), which contributed to keeping particles from gathering with each other, namely, steric hindrance effect. In addition, because PVP’s hydrophilic groups are positive, it constituted charge layers on the surfaces which enhanced the repulsion between two particles. As to the (NH4)2SO4-added powder, the electrostatic effect of SO42-is not strong enough to keep particles apart from each other at high reaction temperature. Although its particle size is much smaller (Fig. 4d), particles agglomerated rather severely and many pores were formed in the aggregate so that it is quite hard to sinter for making transparent ceramics.

Fig. 4. SEM photographs (left) and TEM photographs (right) of the powders calcined at 900 ℃for 2 h. The additives were: virgin (a and b); (NH4)2SO4(c and d); PVA (e and f); PVP (g and h); PVA/PVP (i and j)

(a)????????????????? (b)

Fig. 5. Dispersive effect of PVA (a) and PVP (b)

3. 5 Special surface area measurement

To study further about the effect of PVA and PVP on improving dispersity of particles, we examined the specific surface area of calcined products via BET method. Fig. 6 shows that the specific surface area of the product taking PVA or PVP as the dispersant is larger than the virgin one’s by 20% approximately. What’s more, the specific surface area of the product using PVA/PVP as the dispersant is about 30% larger than that of the virgin one, which manifests PVA and PVP cooperated with each other in alleviating agglomeration. Considering the fact that the temperature of the oil bath is so high (107 ± 2 ℃) that particles are strongly inclined to gather, PVA and PVP are still promising dispersants, though their benefit for improving products’ dispersity is not very great. As for the (NH4)2SO4- added powder, its much larger specific surface area is owing to the smaller particles than others.

Fig. 6. Specific surface area of the powders calcined at 900 ℃ for 2 h with different additives measured by BET method

4 CONCLUSION

Monodispersed and submicron-sized Y2O3spherical powders were successfully prepared via the urea homogeneous precipitation method using PVA, PVP, or PVA/PVP as the dispersant. Both PVA and PVP improved the particle’s dispersion to a great degree. Furthermore, they were synergistic and engendered no impurity phases after calcining. The spherical particle sizes of the powders calcined at 900 ℃ for 2 h are 330～350 nm and the productivity is up to 60%. In this study, the contents of PVA and PVP were 15.5 Wt.% and 8.5 Wt.% of the raw Y2O3powders, respectively. Considering their vital role in weakening agglomeration, it is significant to move the study forward. For example, we can explore the effect of the proportion and average molecular weights of PVA and PVP on promoting dispersity of products. In sum, to make HP in producing Y2O3and other REO powders more meaningful and valuable, it is crucial to substitute new dispersants, such as PVA and PVP, for the traditional (NH4)2SO4at high reaction temperature in solution.

(1) Tsukuda, Y. Application for refractory and corrosion resistant materials.. 1988, 23, 456-460.

(2) Curtis, C. E. Properties of yttrium oxide ceramics.. 1957, 40, 274-278.

(3) Wickersheim, K. A.; Lefever, R. A. Infrared transmittance of crystalline yttrium oxide and related compounds.. 1964, 111, 47-51.

(4) Brecher, C.; Wei, G. C.; Rhodes, W. H. Point defects in optical ceramics: high-temperature absorption processes in lanthana-strengthened yttria.. 1990, 73, 1473-1488.

(5) Kong, J.; Lu, J.; Takaichi, K.; Uematsu, T.; Ueda, K.; Tang, D. Y.; Shen, D. Y.; Yagi, H.; Yanagitani, T.; Kaminskii, A. A. Diode-pumped Yb:Y2O3ceramic laser.. 2003, 82, 2556-2558.

(6) Boulon, G.; Lupei, V. Energy transfer and cooperative processes in Yb3+-doped cubic sesquioxide laser ceramics and crystals.. 2007, 125, 45-54.

(7) Lu, J. R.; Takaichi, K.; Uematsu, T.; Shirakawa, A.; Musha, M.; Ueda, K.; Yagi, H.; Yanagitani, T.; Kaminskii, A. A. Yb3+:Y2O3ceramics-a novel solid state laser material.s. 2002, 41, 1373-1375.

(8) Ikesue, A.; Kamata, K.; Yoshida, K. Synthesis of transparent Nd doped HfO2-Y2O3ceramics using HIP.c. 1996, 79, 359-364.

(9) Li, W. J.; Zhou, S. M.; Liu, N. Effect of additives on optical characteristic of thulium doped yttria transparent ceramics. J.. 2010, 32, 971-974.

(10) Sharma, P. K.; Jilavi, M. H.; Nab, R. Seeding effect in hydrothermal synthesis of nanosize yttria.. 1998, 17, 823-825.

(11) Li, Q. S.; Feng, C. H.; Jiao, Q. Z. Shape-controlled synthesis of yttria nanocrystals under hydrothermal conditions.J.2004, 201, 3055-3059.

(12) Dupont, A.; Parent, C.; Garrec, B. L. Size and morphology control of Y2O3nanopowders via a sol-gel route.. 2003, 171, 152-160.

(13) Mangalaraja, R. V.; Ramama, K. V. S.; Ravi, J. Synthesis of nanocrystalline yttria by microwave-assisted citrate-gel decomposition technique.. 2008, 197, 292-295.

(14) Saito, N.; Matsuda, S.; Ikegami, T. Fabrication of transparent yttria ceramics at low temperature using carbonate derived powder.. 1998, 81, 2023-2028.

(15) Wen, L.; Sun, X. D.; Lu, Q. Synthesis of yttria nanopowders for transparent yttria ceramics... 2006, 29, 239-245.

(16) Li, J. G.; Li, X. D.; Sun, X. D. Uniform colloidal spheres for (Y1-xGd)2O3(= 0-1): formation mechanism, compositional impacts, and physicochemical properties of the oxides.. 2008, 20, 2274-2281.

(17) Qin, X. P.; Zhou, G. H.; Yang, H.; Yang, Y.; Zhang, J.; Wang, S. W. Synthesis and upconversion luminescence of monodispersed, submicron-sized Er3+:Y2O3spherical phosphors.. 2010, 493, 672-677.

(18) Gong, H.; Tang, D. Y.; Huang, H. Effect of grain size on the sinterability of yttria nanopowders synthesized by carbonate-precipitation process.. 2008, 112, 423-426.

(19) Ikegami, T.; Li, J. G.; Mori, T. Fabrication of transparent yttria ceramics by the low-temperature synthesis of yttrium hydroxide.. 2002, 85, 1725-1729.

(20) Wen, L.; Sun, X. D.; Xiu, Z. M. Synthesis of nanocrystalline yttria powder and fabrication of transparent YAG ceramics.. 2004, 24, 2681-2688.

(21) Qin, H. M.; Liu, H.; Sang, Y. H.; Zhang, X. L.; Lv, Y. H.; Wang, J. Y. Influence of the synthesis conditions on preparation of yttria powders by urea precipitation method.. 2011, 40, 1455-1459.

(22) Rietveld, H. M. A profile refinement method for nuclear and magnetic structures.. 1969, 2, 65-71.

① This work was supported by the National Natural Science Foundation of China (91022035) and the Center for Advanced Materials,Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences

② Corresponding author. Cao Yong-Ge. E-mail: caoyongge@fjirsm.ac.cn

10.14102/j.cnki.0254-5861.2011-0722

18 March 2015; accepted 5 May 2015 (ICSD 155173)