硅酸盐学报· 1620 ·2008年部分稳定氧化锆陶瓷的凝胶注模成型工艺仝建峰，陈大明(北京航空材料研究院，先进复合材料国防科技重点实验室，北京 100095)摘要：用流变学的方法研究了不同条件，如：固相含量、分散剂加入量、烧结助剂、增塑剂等对碱性部分稳定氧化锆(partially stabilized zirconia, PSZ)悬浮体的流变性的影响。

结果表明：分散剂含量对悬浮液的流变性能有明显影响，当PSZ固相体积含量为55%时，分散剂加入量(占固相含量的质量分数)应为0.4%。

当固相体积含量为50%～56%时，氧化锆碱性料浆呈现剪切变稀行为，具有较低的黏度(在剪切速率为10 s–1时，低于50mPa·s)。

氧化锆陶瓷碱性料浆(pH>7)在低的剪切速率(<100s–1)时，表现为剪切变稀。

凝胶注模法生产的PSZ陶瓷坯体的内部结构是均匀的。

关键词：部分稳定氧化锆；凝胶注模；流变特性中图分类号：O373 文献标识码：A 文章编号：0454–5648(2008)11–1620–05PREPARATION OF PARTIALLY STABILIZED ZIRCONIA CERAMIC BY AQUEOUS GELCASTINGTONG Jianfeng，CHEN Daming(The National Key Laboratory of Advanced Composite Materials, Beijing Institute of Aeronautical Materials,Beijing 100095, China, Beijing 81–3 100095, China)Abstract: A partially stabilized zirconia (PSZ) ceramic was prepared by aqueous gelcasting. The effects of the zeta potentials, solid loading, dispersant content and milling time on the PSZ suspension were studied. The dispersant content has a remarkable effect on the rheological properties of the suspension. The appropriate dispersant mass fraction for PSZ aqueous slurry with the solid loading of 55% in volume is 0.4%. All suspensions (50%−56% solid loading) exhibit a shear-thinning behavior and relatively low viscosity (less than 50mPa·s, at a shear rate of 10s–1), which is suitable for casting. The degree of shear thinning and the viscosity at high shear rates increase with the increasing of volume fraction of solid phase. As the milling time is prolonged, the viscosity of the suspension de-creases first, and then a plateau appears and the average diameter remains unchanged. When the milling time is shorter than 20h, the viscosity of the slurry decreases gradually as the time of milling is increased. After 20h milling, the viscosity of the slurry tends to be consistent. Therefore, the ball milling time should be equal to or more than 20h in order to obtain a stable suspension at equilibrium. The appropriate time for casting the slurry (idle time) can be controlled by the amounts of initiator and catalyst added to the slurry as well as by the processing temperature. According to micrographs, the gelcasting green body is homogeneous.Key words: partially stabilized zirconia; gelcasting; rheologic propertyGelcasting is an attractive new ceramic forming proc-ess for making high-quality, complex-shaped ceramic bodies.[1–6] Gelcasting has many distinct advantages compared with conventional ceramic forming processes such as dry pressing, slip casting, tape casting, and injec-tion molding, and it is a near-net-shape forming process. Its products have high green density, low levels of or-ganic additives and machinability in the green state due to a high strength.[7–15]Both non aqueous and aqueous solvents can be used in gelcasting. But aqueous systems are preferred be-cause the use of water as the solvent has many advan-tages, e.g., less departure from traditional ceramic processing and no environmental problems for disposal. In aqueous gelcasting, acrylamide and methylene bis-acrylamide are commonly used to make monomer solu-tions.[11–18]According to previous studies,[1–21 the composition of monomer solution, the amounts of initiator and catalyst added, and the temperature and humidity of the drying atmosphere are important processing parameters to be controlled for optimum gelcasting. In this work, a par收稿日期：2008–04–25。

修改稿收到日期：2008–07–03。

基金项目：国家自然科学基金(50672091)资助项目。

第一作者：仝建峰(1972—)，男，博士，高级工程师。

Received date:2008–04–25. Approved date: 2008–07–03. First author: TONG Jianfeng (1972–), male, doctor, senior engineer. E-mail: jftong@第36卷第11期2008年11月硅酸盐学报JOURNAL OF THE CHINESE CERAMIC SOCIETYVol. 36，No. 11November，2008仝建峰等：部分稳定氧化锆陶瓷的凝胶注模成型工艺· 1621 ·第36卷第11期tially stabilized zirconia (PSZ) ceramic was prepared by aqueous gelcasting. The effects of the zeta potentials, solid loading, dispersant content and milling time on the PSZ suspension were studied.1 Experimental procedure1.1 GelcastingThe PSZ powder used in the test was commercial grade, produced by Jiangxi Fanmeiya Co., Ltd. in China. Its average size was 0.4μm. The chemical composition in mass of the powder was >94% ZrO2, 5.2% Y2O3, <0.01% SiO2, <0.005% Fe2O3, <0.005% Na2O, <0.005% Al2O3, <0.02% Cl–1. Its average specific surface area was 12m2/g.The essential components of the gelcasting process were the reactive organic monomers: mono-functional acrylamide, C2H3CONH2 (AM), and difunctional, N, N- methylene bisacrylamide, (C2H3CONH)2CH2(MBAM). These monomers were dissolved in deionized water to give a premix solution. The premix solution undergoes free radical initiated vinyl polymerization in the presence of an initiator such as ammonium presulfate, (NH4)S2O8. The reaction is accelerated by heat or by the catalyst N,N,N,N-tetramethylenediamine (TEMED). The result-ing cross-linked polymer is an elastic hydrogel that serves as the binder.[1–6]The gelcasting process flow chart is shown in Fig.1. The monomers were dissolved in deionized water to make solution with a composition of 2.5% (mass fraction according to the mass of PSZ powder solid loading, the same below) AM, 0.125% MBAM. Gelcasting slurriesFig.1 Flow chart of preparing PSZ by gelcasting were prepared by ball-milling the PSZ powder in the monomer solution. 0.4% JA281 (a dispersant produced by Beijing Institute of Aeronautical Materials, Beijing, China) were added to the slurries. In the slurries the solid loading was about 55% (volume fraction, the same be-low), and the total amount of organic monomers was 3.2% for the PSZ powder. Both the initiator, 1% aqueous solution of ammonium persulphate, and the catalyst, 50% aqueous solution of TEMED, were added just before mold casting. After gelling, the disk-shaped samples were demolded and dried in lab atmosphere.1.2 Properties measurementThe zeta potentials at different suspension pH values were measured by a zeta potential meter. The rheological properties of the ceramic slurry were measured by a ro-tary rheometer (Model RV–30, US). The densities of the green bodies were calculated from the volume measured by a mercury displacement method. The dried green body was cut into sample bars with a width of 6mm, a height of 5mm and a length of 36mm. Their room temperature bending strength was examined by the three-point flexure test with a span of 30mm and a cross head speed of 0.5 mm/min. The microscopic morphology was observed by a scanning electron microscope (SEM, Model Jonel–2000).2 Results and discussion2.1 Effects of pH value on zeta potential of PSZsuspensionThe zeta potentials at different pH values of the sus-pensions are shown in Fig.2. After dispersant is added, the isoelectric point (pH iep value) will decrease from 7.5 to 7.0. At lower pH, far from the pH iep value, the PSZ particles have a high positive zeta potential and are col-loidal stable. In the vicinity of the pH iep value, the parti-cles have a low zeta potential that may be either positive (pH<pH iep) or negative (pH>pH iep). The suspensions pre-pared within this region are colloidal unstable, consisting of large agglomerates that are likely to lead to porous gelcast ceramics. At higher pH values, far from the pH iepFig.2 Zeta potentials of PSZ as a function of pH value硅酸盐学报· 1622 ·2008年value, the particles have a high negative zeta potentialand are colloidal stable. The profiles in Fig.2 show thatthe zeta potential is maximized at pH=11, which can pro-vide better colloidal stability.2.2 Rheology characteristic2.2.1 Effect of dispersant on the slurry rheology For the PSZ aqueous slurry with JA281 as dispersant, the effects of dispersant content on the rheology of the slurry are shown in Fig.3. As seen in Fig.3, the viscosity of the slurry has the minimum value at a dispersant mass frac-tion range of 0.35%–0.40%, beyond which the slurry viscosity increases again. Therefore, the appropriate dis-persant concentration for PSZ aqueous slurry is 0.40%.Fig.3 Variation of slurry viscosity with mass fraction ofdispersant JA281It is well known that the dispersant is a polyunsatu- rated ester molecule, and its dispersive capability is not only determined by steric stabilization, but also the stateof the dispersant in the surface of the particles. The three states of dispersant on the surface of the particles are shown in Fig.4.[22] When the state of the dispersant on the surface of the particles is saturated adsorption, the dis-persant capability is good; however, both unsaturated adsorption and over-saturated adsorption are bad for the dispersing of particles in the solvent.Fig.4 The state of dispersant on the surface of the particles2.2.2 Effect of solid loading on the slurry rheologyFigure 5 (a) gives the plots of apparent viscosity versus applied shear rate of stable suspensions after millingFig.5 Rheological behavior of the suspensions with different solid loadings (in volume, the same below) after ballmilling for 24hfor 24h. It can be seen that all suspensions (50%−56% solid loading) exhibit a shear-thinning behavior and rela-tively low viscosity (less than 50mPa·s for 10s–1), and are suitable for casting. Figure 5 (b) shows the plots of the relationship of shear stress versus the applied shear rate of stable suspensions after milling for 24h. It can be seen that the suspensions with 50%−56% solid loading do not possess thixotropy.General concentrated colloidal stable suspensions ex-hibit shear-thinning behavior in steady shearing because of a perturbation of the suspension structure by shear-ing.[8–16] At lower shear rates, the suspension structure is close to equilibrium because the thermal motion domi-nates over the viscous forces. At higher shear rates, the viscous forces affect the suspension structure and shear thinning occurs. At very high shear rates, the viscous forces have a dominant affect and the viscosity can be considered as a measurement of the resistance to flow of a suspension with a completely hydrodynamically con-trolled structure. The degree of shear thinning and the viscosity at higher shear rates increase with the increas-ing of the volume fraction of the solid.2.2.3 Effect of milling time on the slurry rheologyTypical plots of apparent viscosity versus different times仝建峰 等：部分稳定氧化锆陶瓷的凝胶注模成型工艺· 1623 ·第36卷第11期of ball milling for different volume fractions of solid are shown in Fig.6. When the milling time is shorter than 20 h, the viscosity of the slurries decreases gradually with the increase of the milling time. This shows that the absorption of the dispersant on parti-cles cannot reach equilibrium and the suspension is unstable until the ball milling time is equal to or more than 20 h. After 20 h milling, the viscosity of the slurry tends to be consistent. Therefore, the ball milling time should be equal to or more than 20 h to obtain a stable suspension atequilibrium.Fig.6 Effect of milling time on the slurry viscosity of the sus-pensions with different solid loadings2.3 Gelling2.3.1 The effect of initiator, catalyst and temperature on gelling rate In the present study, the initiation of po-lymerization in premixes is determined by changes in solution temperature, since the reaction is exothermic. The process was monitored in terms of idle time, t idle , the time between the addition of the initiator, or the initia-tor/catalyst and the commencement of polymerization. This is equivalent to the time used for casting the slurry during processing. Figure 7 shows the change in tem-perature of a premix solution during gelling and indicates the idle time.Fig.7 Effect of content of the catalyst on the idle timeIni.—Initiator amount; Cat.—Catalyst amount.The idle time can be controlled by the concentration ofthe reagents and the temperature. In this study, the idle time ranged from 5 to 60 min. Figure 7 shows the varia-tion of idle time with the amount of initiator and catalyst, and Fig.8 shows the variation of idle time with initiator and solution temperature. These data show clearly that polymerization is accelerated by an increase in initiator,catalyst or temperature.Fig.8 Effect of contents of the initiator and temperature on theidle time2.3.2 The density and strength of gelcast green bodies The effect of polymer content on the density and strength of the gelcast sample is shown in Fig.9. The maximum density is obtained at a polymer content of ～6% in mass. The green strength, on the other hand, increases more or less linearly with the binder content. The maximum relative density of the green body ob-tained is 58%; at this density, the strength is 46 MPa.Fig.9 Effect of binder content in mass on the density andstrength of green bodies2.3.3 Micrograph of gel-cast green bodies Figure 10 shows the SEM photographs of a dry body. It can be seen that powders are connected by slender polymer chains, which are responsible for the strength of the green body. It can also be seen that the micrographs of the edge and center section were homogeneous.硅酸盐学报· 1624 ·2008年Fig.10 SEM photographs of gelcast green bodies3 Conclusions(1) Aqueous gelcast PSZ slurry has low viscosity at high solid loading.(2) The characteristics of the slurry are connected with the dispersant, solid loading and milling time.(3) The density and strength of gelcast green bodies can be controlled by varying the monomer concentration.(4) The time required for casting the slurry (idle time) can be controlled by the amounts of initiator and catalyst added to the slurry as well as by the processing tempera-ture.(5) The microstructure of the gelcast green is homo-geneous.Reference:[1] JANNEY M A, OMETETE O O. Gelcasting of ceramic powders [P].US Patent, 5145908. 1992–05–12.[2] OMETETE O O, JANNEY M A, STREHLOW R A. Gelcasting–a newceramic forming process [J]. Ceram Bull, 1991, 70(10): 1641–1648. [3] JANNEY M A. Method for molding ceramic powders [P]. US Patent,4894194. 1990–03–15.[4] LANGE F F. V ALAMAKANNI B V. Method of preparing dense ce-ramic product [P]. US Patent, 5188780. 1991–11–13.[5] TIEGS T N,WITTMER D E. Method of preparing a high solids content,low viscosity ceramic slurry [P]. US Patent, 5456877. 1994–09–12. [6] GAUCKLER L J,GRAULE T. Method for forming green ceramics [P].U.S. Patent, 5788891. 1996–11–02.[7] MORISSETTE S L. Chemorheology of aqueous-based alumina-poly(vinyl alcohol) gelcasting suspensions [J]. J Am Ceram Soc, 1999, 82(3): 521–527.[8] YANEZ J A. Shear modulus and yield stress measurement of attractivealumina particle networks in aqueous slurries [J]. J Am Ceram Soc, 1996, 79(11): 2917–1921.[9] HIDDER P C, GRAULE T J, GAUCKLER L J. Citric acid–a dispers-ant for aqueous alumina suspensions [J]. J Am Ceram Soc, 1996, 79(7): 1857–1867.[10] REED J S. Principles of Ceramics Processing [M]. Second Ed. NewYork: John Wiley and Sons, Inc. 1995: 96–97.[11] YOUNG A C, OMETETE O O, JANNEY M A. Gelcasting of alumina[J]. J Am Ceram Soc, 1991, 74(3): 612–616.[12] MARIA J P, KIGGANS J O, TIEGS T N. Gelcasting of sinteredreaction bonded silicon nitride for improved mechanical properties [J].Ceram Eng Sci Proc, 1995, 16(5): 1071–1074.[13] CARISEY T, WERTH A L, BRANDON D G. Control of texture inAl2O3 by gelcasting [J]. J Eur Ceram Soc, 1995, 15(1): 1–8.[14] BASKIN D M, ZIMMERMAN M H, FABER K T. Forming single-pHas a laminates via the gelcasting technique [J]. J Am Ceram Soc, 1991, 74(3): 612–615.[15] DUNCAN J S. Introduction to Colloid and Surface Chemistry [M].London: Bitterworth & Co, 1980: 35–36.[16] MOONEY M. The viscosity of concentrated suspension of sphericalparticles[J]. J Coll Sci, 1956, (6): 162–165.[17] CESARANO Ⅲ J, AKSAY I A. Processing of highly concentratedaqueous α-Al2O3 suspensions with poly (methacrylic acid) polyelec-trolyte [J]. J Am Ceram Soc, 1990, 71(4): 250–254.[18] OVERBEEK H V. On the theory of the stabilization of dispersions byadsorbed macromolecules, I statistics of the change of some configura-tional properties of adsorbed macromolecules on the approach of an impenetrable interface [J]. J Phys Chem, 1971, 75(1): 65–68.[19] OVERBEEK H V. On the theory of the stabilization of dispersions byadsorbed macromolecules, Ⅱ interaction between two flat particles [J]. J Phys Chem, 1971, 75(10): 2094–2097.[20] STRBLE L J, ZUKOSKI C F, MAITLAND G C, et al. Flow andmicrostructure of dense suspension [J]. Mater Res Soc, 1992, 72(4): 123–127.[21] BASKIN D M, ZIMMERMAN M H, FABER K T. Forming sin-gle-phase laminates via the gelcasting technique [J]. J Am Ceram Soc, 1991, 74(3): 612–616.[22] OVERBEEK H V. On the theory of the stabilization of dispersions byadsorbed macromolecules,Ⅱinteraction between two flat particles [J].J Phys Chem, 1971, 75(10): 2094–2097.。