CERAMICS INTERNATIONAL SEVIER Ceramics International 30(2004)843-852 www.elsevier.com/locate/ceramint Electrophoretic deposition of alumina and zirconia I Single-component systems Karel Maca", Hynek Hadraba, Jaroslav Cihlar Department of Ceramics, Brno University of Technolog 616 69 Brno, Czech Republic Received 30 May 2003: received in revised form 17 August 2003; accepted 25 September 2003 Available online 20 March 2004 Abstract Electrophoretic deposition of Al,, and ZrO, suspended in isopropanol in the presence of monochloroacetic acid and polyvinylbuty under constant-current conditions was studied. The deposition of ceramic particles occurred on the anode. An optimum deposition in terms of surface flatness, deposit density and thickness was found for 15 wt. of monochloroacetic acid in isopropanol. The electrophoretic mobility of alumina particles in this suspension was determined from deposition kinetics. The sintering behaviour, hardness, fracture toughness and ending strength of Al2O3 and ZrO2 electrophoretic deposits were comparable with the values measured for the specimens prepared from identical powders by injection moulding and cold isostatic pressing or with the values given in the literature C 2003 Elsevier Ltd and Techna Group S.r. l. All rights reserved Keywords: B Composites, D. Al2O3; D ZrO2; Electrophoretic deposition 1. Introduction time suspension stability is the limiting factor with respect to obtaining a deposit of high density [10] Electrophoretic deposition is an experimentally unde- Zhitomirski [ll] described electrophoretic manding and comparatively cheap technique enabling the 12O3 and ZrO2 powders in propanol without admixtures formation of deposits from particles of nanometric di- and at voltages from 50 to 200V, he deposited within mensions [1, 2] to particles of several micrometers in size 10-300 s a deposit of 1-10 um in thickness. By contrast, 3]. Electrophoretic deposition can yield layers of a wide Harbach and Nienburg [5,6] described a ZrO2 suspension range of thickness, from several nanometers [4] to several in pure isopropanol and ethanol as unstable. The authors millimetres [5, 6] obtained better results using ZrO2 dispersed in ethanol Usually, electrophoretic deposition of Al2O3 and ZrO2 and isopropanol stabilized with 4-hydroxy benzoic acid, in aqueous medium is described in the literature [7, 8]. The polyethylenimine and acrylate-acrylamid copolymer. The advantage of using aqueous suspensions for electrophoretic same authors deposited AlO3 tubes stabilized with trioxy- deposition lies in the easy preparation of stable suspensions decanoic acid in isopropanol. will et al. [ 12] stabilized Zroz of ceramic particles and the ease of controlling the charge powder for deposition in ethanol using polyethylenimine on particle surface by changing the ph value of aqueous Ishihara and coworkers [ 13, 14] tested a great number of sol- medium [9]. However, using aqueous solutions results in vents from the viewpoint of the surface quality of the Zro2 the electrolysis of water, with gaseous hydrogen generated deposit. They obtained the best results when using acety- on the cathode and oxygen on the anode. These gases form lactone, in which he stabilized the zrO2 particles by means bubbles in the deposit, reduce its density and prevent the of iodine. Chen and Liu [15] used iodine to stabilize Zroz deposit from adhering to the electrode[7]. On the other side particles in a mixture of one volume part of ethanol( which he disadvantage of using suspensions with organic solvents improved the resistance of the deposit to cracking during is that the particles are more difficult to stabilize; at the same drying) and three volume parts of acetone(which improved the surface quality of the deposit). Basu et al. [16] deposited dried ZrO2 from anhydrous medium of glacial acetic acid. Corresponding author. Fax: +420-541143202 Most of the references describe electrophoretic E-mail address: maca @umi. fme. vutbr cz(K. Maca) ition at a constant voltage [5,8,9, 11, 15, 17-19]. In this 0272-8842/$30.00 0 2003 Elsevier Ltd and Techna Group S r l. All rights reserved doi:10.1016/ ceramist2003.09.021
Ceramics International 30 (2004) 843–852 Electrophoretic deposition of alumina and zirconia I. Single-component systems Karel Maca∗, Hynek Hadraba, Jaroslav Cihlar Department of Ceramics, Brno University of Technology, 616 69 Brno, Czech Republic Received 30 May 2003; received in revised form 17 August 2003; accepted 25 September 2003 Available online 20 March 2004 Abstract Electrophoretic deposition of Al2O3 and ZrO2 suspended in isopropanol in the presence of monochloroacetic acid and polyvinylbutyral under constant-current conditions was studied. The deposition of ceramic particles occurred on the anode. An optimum deposition in terms of surface flatness, deposit density and thickness was found for 15 wt.% of monochloroacetic acid in isopropanol. The electrophoretic mobility of alumina particles in this suspension was determined from deposition kinetics. The sintering behaviour, hardness, fracture toughness and bending strength of Al2O3 and ZrO2 electrophoretic deposits were comparable with the values measured for the specimens prepared from identical powders by injection moulding and cold isostatic pressing or with the values given in the literature. © 2003 Elsevier Ltd and Techna Group S.r.l. All rights reserved. Keywords: B. Composites; D. Al2O3; D. ZrO2; Electrophoretic deposition 1. Introduction Electrophoretic deposition is an experimentally undemanding and comparatively cheap technique enabling the formation of deposits from particles of nanometric dimensions [1,2] to particles of several micrometers in size [3]. Electrophoretic deposition can yield layers of a wide range of thickness, from several nanometers [4] to several millimetres [5,6]. Usually, electrophoretic deposition of Al2O3 and ZrO2 in aqueous medium is described in the literature [7,8]. The advantage of using aqueous suspensions for electrophoretic deposition lies in the easy preparation of stable suspensions of ceramic particles and the ease of controlling the charge on particle surface by changing the pH value of aqueous medium [9]. However, using aqueous solutions results in the electrolysis of water, with gaseous hydrogen generated on the cathode and oxygen on the anode. These gases form bubbles in the deposit, reduce its density and prevent the deposit from adhering to the electrode [7]. On the other side the disadvantage of using suspensions with organic solvents is that the particles are more difficult to stabilize; at the same ∗ Corresponding author. Fax: +420-541143202. E-mail address: maca@umi.fme.vutbr.cz (K. Maca). time suspension stability is the limiting factor with respect to obtaining a deposit of high density [10]. Zhitomirski [11] described electrophoretic deposition of Al2O3 and ZrO2 powders in propanol without admixtures, and at voltages from 50 to 200 V, he deposited within 10–300 s a deposit of 1–10 m in thickness. By contrast, Harbach and Nienburg [5,6] described a ZrO2 suspension in pure isopropanol and ethanol as unstable. The authors obtained better results using ZrO2 dispersed in ethanol and isopropanol stabilized with 4-hydroxybenzoic acid, polyethylenimine and acrylate-acrylamid copolymer. The same authors deposited Al2O3 tubes stabilized with trioxydecanoic acid in isopropanol. Will et al. [12] stabilized ZrO2 powder for deposition in ethanol using polyethylenimine. Ishihara and coworkers [13,14] tested a great number of solvents from the viewpoint of the surface quality of the ZrO2 deposit. They obtained the best results when using acetylacetone, in which he stabilized the ZrO2 particles by means of iodine. Chen and Liu [15] used iodine to stabilize ZrO2 particles in a mixture of one volume part of ethanol (which improved the resistance of the deposit to cracking during drying) and three volume parts of acetone (which improved the surface quality of the deposit). Basu et al. [16] deposited dried ZrO2 from anhydrous medium of glacial acetic acid. Most of the references describe electrophoretic deposition at a constant voltage [5,8,9,11,15,17–19]. In this 0272-8842/$30.00 © 2003 Elsevier Ltd and Techna Group S.r.l. All rights reserved. doi:10.1016/j.ceramint.2003.09.021
844 K. Maca et al. /Ceramics International 30(2004)843-852 case, however, the weight increment of the deposit with timizing suspension composition with all MCAA and PVB time declines rapidly or even stops completely because the concentration levels was conducted with type HP Al2O3 density of electric current decreases due to increasing re- All MCAA concentration levels were also used with type sistance of the system. This problem does not appear in the 3YS ZrO2. The study of suspension sedimentation, sintering constant-current mode [6, 201 kinetics and mechanical properties of deposits was carried It has been established recently that in isopropanol suspen- out on only one selected concentration level of MCAA and sions stabilized with monochloroacetic acid the Al2O3 and PVB. The exact composition of all the suspensions used is ZrO particles have a negative charge [21]. Electrophoretic gIven in Table 2 deposition, which in such cases occurs on the anode, has not so far been described for these systems in the literature. 2. 3. Electrophoretic deposition The aim of the present work was therefore to prepare Al2O3 and ZrO2 deposits from isopropanol suspensions stabilized A schematic section through a horizontal electrophoretic with monochloroacetic acid. Attention was mainly focused cell is given in Fig. 1. The electrodes made of stainless on optimizing the suspension composition and describing steel with polished surface were at a constant distance of he sintering kinetics, resultant microstructure and mechan- 26 mm, their effective areas were in the shape of a trapezoid of 15 cm. The volume of suspension in the electrophoretic cell was 80 mL A stabilized source controlled by microcom- puter(E815, Consort, Belgium) was uised as the voltage and 2. Experimental current source. All experiments were conducted in the 5 mA constant-current mode 21. Materials The depositions employed to optimize the suspension composition(Section 3. 1) were performed by continuous The Al2O3 and ZrO2 ceramic powder materials used for electrophoretic deposition lasting 30 min electrophoretic deposition are summarized in Table 1. Iso- When studying the effect of suspension sedimentation on propanol(p a, Onex, Czech Republic) was used as the dis- the course of deposition(Section 3. 2), interrupted deposi persion medium for the preparation of the suspensions of tions taking a total of 60-80 min were performed. With ev- Al2O3 and ZrOz powders. Monochloroacetic acid(MCAA; ery interruption of the deposition the deposit thickness w p.a., Lachema, Czech Republic)was used as the stabilizing measured, namely at 18, 38 and 58 mm below the suspen- and dispersing aid and polyvinylbutyral(PVB, Butvar B79, sion level and subsequently the suspension was or was not Monsanto, USA) was used as a binder. The water content stirred In the case of deposition without stirring the drop in in the suspensions was reduced to minimum by drying the the suspension top level (which contrasted with the volume ceramic powders at 130C/2 h while isopropanol was dried of clear solvent without ceramic powder-the supernatant) over calcium for several hours and subsequently distilled un- was read during the deposition time. These readings were til its water content was less than 0.01%(as shown by gas used to establish the mean velocity of sedimentation(us) chromatography analysis The resultant shape of the deposit was determined by mea- suring its thickness at 22 equidistant vertical depths 2. 2. Suspension composition Deposits of a larger thickness, designed for the study f sintering (Section 3.3) and mechanical The suspensions were prepared by mixing 15 wt% of (Section 3. 4), were obtained by interrupted deposition, i.e Al2O or ZrO2 powder in 85 wt. of the liquid phase that a sequence of consecutive depositions from one suspension, was prepared by dissolving 1, 2, 5, 15 or 25 wt. of MCAa with the suspension stirred between every two depositions and (in the case of Al2O3 suspension)0, 1, 2 or 4 wt% of Every partial deposition lasted 5 min, the total deposition PVB in isopropanol. A complete factor experiment for op- time was 120-160 min Table 1 Ceramic powder materials used for electrophoretic deposition Marking Material Manufacturer Manufacturers marking Mean particle size(um)a Mean agglomerate Malakoff Ind (USA) RC HP DBM 0.47 UFX Malakoff Ind (USA) RC UFX DBM 0.33 ZrO, d Tosoh (Japan) TZ-3Y 3YS ZrO, d Tosoh(Japan) TZ-3YS Established by laser diffraction(Horiba LA-500, Japan) c With an addition of 0.05 wt% Mgo d Stabilized by 3 mol. Y203
844 K. Maca et al. / Ceramics International 30 (2004) 843–852 case, however, the weight increment of the deposit with time declines rapidly or even stops completely because the density of electric current decreases due to increasing resistance of the system. This problem does not appear in the constant-current mode [6,20]. It has been established recently that in isopropanol suspensions stabilized with monochloroacetic acid the Al2O3 and ZrO2 particles have a negative charge [21]. Electrophoretic deposition, which in such cases occurs on the anode, has not so far been described for these systems in the literature. The aim of the present work was therefore to prepare Al2O3 and ZrO2 deposits from isopropanol suspensions stabilized with monochloroacetic acid. Attention was mainly focused on optimizing the suspension composition and describing the sintering kinetics, resultant microstructure and mechanical properties of these deposits. 2. Experimental 2.1. Materials The Al2O3 and ZrO2 ceramic powder materials used for electrophoretic deposition are summarized in Table 1. Isopropanol (p.a., Onex, Czech Republic) was used as the dispersion medium for the preparation of the suspensions of Al2O3 and ZrO2 powders. Monochloroacetic acid (MCAA; p.a., Lachema, Czech Republic) was used as the stabilizing and dispersing aid and polyvinylbutyral (PVB; Butvar B79, Monsanto, USA) was used as a binder. The water content in the suspensions was reduced to minimum by drying the ceramic powders at 130 ◦C/2 h while isopropanol was dried over calcium for several hours and subsequently distilled until its water content was less than 0.01% (as shown by gas chromatography analysis). 2.2. Suspension composition The suspensions were prepared by mixing 15 wt.% of Al2O3 or ZrO2 powder in 85 wt.% of the liquid phase that was prepared by dissolving 1, 2, 5, 15 or 25 wt.% of MCAA and (in the case of Al2O3 suspension) 0, 1, 2 or 4 wt.% of PVB in isopropanol. A complete factor experiment for opTable 1 Ceramic powder materials used for electrophoretic deposition Marking Material Manufacturer Manufacturer’s marking Mean particle size (m)a Mean agglomerate size (m)b HP Al2O3 c Malakoff Ind. (USA) RC HP DBM 0.47 0.6 UFX Al2O3 c Malakoff Ind. (USA) RC UFX DBM 0.33 0.4 3Y ZrO2 d Tosoh (Japan) TZ-3Y 0.25 0.7 3YS ZrO2 d Tosoh (Japan) TZ-3YS 0.12 0.7 a Established from microphotographs. b Established by laser diffraction (Horiba LA-500, Japan). c With an addition of 0.05 wt.% MgO. d Stabilized by 3 mol.% Y2O3. timizing suspension composition with all MCAA and PVB concentration levels was conducted with type HP Al2O3. All MCAA concentration levels were also used with type 3YS ZrO2. The study of suspension sedimentation, sintering kinetics and mechanical properties of deposits was carried out on only one selected concentration level of MCAA and PVB. The exact composition of all the suspensions used is given in Table 2. 2.3. Electrophoretic deposition A schematic section through a horizontal electrophoretic cell is given in Fig. 1. The electrodes made of stainless steel with polished surface were at a constant distance of 26 mm, their effective areas were in the shape of a trapezoid of 15 cm2. The volume of suspension in the electrophoretic cell was 80 ml. A stabilized source controlled by microcomputer (E815, Consort, Belgium) was used as the voltage and current source. All experiments were conducted in the 5 mA constant-current mode. The depositions employed to optimize the suspension composition (Section 3.1) were performed by continuous electrophoretic deposition lasting 30 min. When studying the effect of suspension sedimentation on the course of deposition (Section 3.2), interrupted depositions taking a total of 60–80 min were performed. With every interruption of the deposition the deposit thickness was measured, namely at 18, 38 and 58 mm below the suspension level and subsequently the suspension was or was not stirred. In the case of deposition without stirring the drop in the suspension top level (which contrasted with the volume of clear solvent without ceramic powder—the supernatant) was read during the deposition time. These readings were used to establish the mean velocity of sedimentation (vs). The resultant shape of the deposit was determined by measuring its thickness at 22 equidistant vertical depths. Deposits of a larger thickness, designed for the study of sintering (Section 3.3) and mechanical properties (Section 3.4), were obtained by interrupted deposition, i.e. a sequence of consecutive depositions from one suspension, with the suspension stirred between every two depositions. Every partial deposition lasted 5 min, the total deposition time was 120–160 min
K. Maca et al. /Ceramics International 30(2004)843-852 Table 2 Al]O3(wt%) O Isopropanol(wt % MCAA(wt % VB(wt %) Used in section(s) HP-1-0 83.30 80.75 4.25 P-150 72.25 HP25-0 21.25 83.16 HP-2. 79.80 P-15-1 55555555555 71.4 12.60 HP-1-2 HP25-2 0.75 HP-1-4 555555 80.20 0.81 333333333 HP-25-4 3YS-5-0 3.1 3YS-15-0 00000 3.1,3.3,34 UFX-15-0 3Y-15-0 15 75 3.3.3.4 2. 4. Evaluation of deposit properties annealed(800C/1 h)and sintered (1500C/2 h)in ai All the deposits together with the electrodes were The quality of the deposit surface subsequent to drying first dried at room temperature for 24 h and at 70C for was evaluated visually. The relative density of the annealed I h, after which they were removed from the electrodes, deposit (Pgoo) was established from soaking capacity and the relative density of the sintered deposit(p1500) was estab- lished by the Archimedes method (EN 623-2). The pore size distribution was established by mercury porosimetry( Carlo supply Erba 200, Italy) The sintering process was monitored by a high-temperature dilatometer(L75/50, Linseis, Germany ) The relative length ge in the deposit as a function of time and temp was measured in two directions, namely in the direction of particle motion during electrophoretic deposition (longitu dinal direction-marked by index"L )and in perpendic ular direction to the particle motion during the deposit (transversal direction--marked by index"T). The dimen sionless coefficient of specimen shrinkage anisotropy (h) was calculated according to the relation where EL is the final deposit relative shrinkage in the longi tudinal direction, and ET is the final deposit relative shrin Fig. 1. Section through the cell used for electrophoretic deposition of age in the transversal direction. The coefficient of thermal expansion(CTE) was calculated from the cooling curves of
K. Maca et al. / Ceramics International 30 (2004) 843–852 845 Table 2 Composition of suspensions Suspension Al2O3 (wt.%) ZrO2 (wt.%) Isopropanol (wt.%) MCAA (wt.%) PVB (wt.%) Used in section(s) HP-1-0 15 0 84.15 0.85 0 3.1 HP-2-0 15 0 83.30 1.70 0 3.1 HP-5-0 15 0 80.75 4.25 0 3.1 HP-15-0 15 0 72.25 12.75 0 3.1, 3.2, 3.3, 3.4 HP-25-0 15 0 63.75 21.25 0 3.1 HP-1-1 15 0 83.16 0.84 1 3.1 HP-2-1 15 0 82.32 1.68 1 3.1 HP-5-1 15 0 79.80 4.20 1 3.1 HP-15-1 15 0 71.40 12.60 1 3.1 HP-25-1 15 0 63.00 21.00 1 3.1 HP-1-2 15 0 82.17 0.83 2 3.1 HP-2-2 15 0 81.34 1.66 2 3.1 HP-5-2 15 0 78.85 4.15 2 3.1 HP-15-2 15 0 70.56 12.45 2 3.1 HP-25-2 15 0 62.25 20.75 2 3.1 HP-1-4 15 0 80.20 0.81 4 3.1 HP-2-4 15 0 79.38 1.62 4 3.1 HP-5-4 15 0 76.95 4.05 4 3.1 HP-15-4 15 0 68.85 12.15 4 3.1 HP-25-4 15 0 60.75 20.25 4 3.1 3YS-1-0 0 15 84.15 0.85 0 3.1 3YS-2-0 0 15 83.30 1.70 0 3.1 3YS-5-0 0 15 80.75 4.25 0 3.1 3YS-15-0 0 15 72.25 12.75 0 3.1, 3.3, 3.4 3YS-25-0 0 15 63.75 21.25 0 3.1 UFX-15-0 15 0 72.25 12.75 0 3.3, 3.4 3Y-15-0 0 15 72.25 12.75 0 3.3, 3.4 2.4. Evaluation of deposit properties All the deposits together with the electrodes were first dried at room temperature for 24 h and at 70 ◦C for 1 h, after which they were removed from the electrodes, Fig. 1. Section through the cell used for electrophoretic deposition of ceramic materials. annealed (800 ◦C/1 h) and sintered (1500 ◦C/2 h) in air atmosphere. The quality of the deposit surface subsequent to drying was evaluated visually. The relative density of the annealed deposit (ρ800) was established from soaking capacity and the relative density of the sintered deposit (ρ1500) was established by the Archimedes method (EN 623-2). The pore size distribution was established by mercury porosimetry (Carlo Erba 200, Italy). The sintering process was monitored by a high-temperature dilatometer (L75/50, Linseis, Germany). The relative length change in the deposit as a function of time and temperature was measured in two directions, namely in the direction of particle motion during electrophoretic deposition (longitudinal direction—marked by index “L”) and in perpendicular direction to the particle motion during the deposition (transversal direction—marked by index “T”). The dimensionless coefficient of specimen shrinkage anisotropy (k) was calculated according to the relation: k = εT εL (1) where εL is the final deposit relative shrinkage in the longitudinal direction, and εT is the final deposit relative shrinkage in the transversal direction. The coefficient of thermal expansion (CTE) was calculated from the cooling curves of
K. Maca et al. /Ceramics International 30(2004)843-852 sintered deposits. The temperature dependence of relative specimen density(Prel (n)) was calculated from the values F of final densities and from the shrinkage curve ED(first, 5 PVB CONCENTRATION [ wt%1 ◇4w%(HPA2o Polished specimens of sintered deposits were thermally ·0w%(3Yszo2) etched and made conductive by applying three layers of gold coating and then studied by SEM (Philips XL30, the Nether lands). The grain size of sintered ceramics(dG )was deter- a ined by computer image analysis(Atlas Software, Tescan, 9 20 Czech Republic)from microphotographs of etched deposit ui sections. The value of equivalent diameter, calculated from he area of grains, was used as the mean grain size(dG) The indentation fracture toughness of sintered deposits MCAA CONCENTRATION [wt%] was measured by the indentation method (according to Japanese Industrial Standard JIs R 1607). The Vickers Fig. 2. Dependence of electric field intensity during deposition of alumina and zirconia on the concentration of MCAA and PvB in the suspension. hardness tester was used as the indentor. The fracture toughness(Kic) of materials was calculated using the val polyvinylbutyral in the suspension(see Fig. 2). The drop in ues of the length of indentation crack(c), indentation force the intensity of electric field with increasing concentration (F=98N), Youngs modulus of ZrO2(Ey 380 GP of monochloroacetic acid was probably due to the increas- [19])and Al2O3(Ey= 210 GPa [19]) and hardness(HV according to the relation given in JISR 1607. The hardness suspensions increased from 2. x 10-4Sm-I(HP-1-0)[21] of material was determined from the size of the diagonal of to 23. 8x 10-4Sm-(HP-25-0)[21]. The relation between the hardness tester tip indentation in the surface of material velocity of particle motion(u), electrophoretic mobility (u) caused by indentation force F and electric field intensity(E) is given by the equation The four-point flexure strength(EN 843-1)was only measured for Al2O3 deposits(type HP). The test speci U=×E. mens were cut with a diamond disk from deposits anneale at 1000"C/i h. The test specimens were then sintered at particles at an MCAA concentration of more than 0.85 wt% is independent of its concentration [211, the velocity of 2mm x 2.5 mm x 25 mm and the tensile side was polished the particles in the suspension decreased with decreasing with diamond paste of l um grain size. Ber electric field intensity and thus with ng mcaa con- was measured on a universal test machine(Z020, Zwick centration. The slower motion of particles probably led to Germany)with a jig for bending strength measurement their more effective arrangement on the anode, which was confirmed by the higher green density of deposits with a higher MCAA content(see Fig 3). The final relative deposit 3. Results and discussion 3.1. Optimization of suspension composition 5 冷D The suspensions with type HP AlzO3 and type 3YS ZrO2 vere used for the optimization experiments( see Table 2) The Al2O3 and ZrOz particles were charged negatively and he layer was deposited on the anode, in the same way as in a 62 green density (p work of Cihlar et al. [211 PVB CONCENTRATION [wt%] The smoothness of deposit surface, deposit thickness and density were affected by suspension composition 口1w%(FPA2) ( the contents of monochloroacetic acid and polyvinylbu- △2w%(AO) tyral). The deposit had a smooth surface if the deposition was from a suspension containing more than 4 wt. of monochloroacetic acid, irrespective of the PVB concentra- tion. This phenomenon was probably due to the effect of MCAA CONCENTRATION [wt% MCAA concentration on electrophoretic behaviour of the Fig 3. Dependence of the relative density of alumina deposit after an- suspension. The intensity of electric field during deposition nealing and sintering on the concentration of MCAA and PVB in the was dependent on the contents of monochloroacetic acid and
846 K. Maca et al. / Ceramics International 30 (2004) 843–852 sintered deposits. The temperature dependence of relative specimen density (ρrel(T)) was calculated from the values of final densities and from the shrinkage curve ε(T) (first, the specimen elongation due to simple thermal expansion of the material was subtracted from the elongation values measured). Polished specimens of sintered deposits were thermally etched and made conductive by applying three layers of gold coating and then studied by SEM (Philips XL30, the Netherlands). The grain size of sintered ceramics (dG) was determined by computer image analysis (Atlas Software, Tescan, Czech Republic) from microphotographs of etched deposit sections. The value of equivalent diameter, calculated from the area of grains, was used as the mean grain size (dG). The indentation fracture toughness of sintered deposits was measured by the indentation method (according to Japanese Industrial Standard JIS R 1607). The Vickers hardness tester was used as the indentor. The fracture toughness (KIC) of materials was calculated using the values of the length of indentation crack (c), indentation force (F = 98 N), Young’s modulus of ZrO2 (EY = 380 GPa [19]) and Al2O3 (EY = 210 GPa [19]) and hardness (HV) according to the relation given in JIS R 1607. The hardness of material was determined from the size of the diagonal of the hardness tester tip indentation in the surface of material caused by indentation force F. The four-point flexure strength (EN 843-1) was only measured for Al2O3 deposits (type HP). The test specimens were cut with a diamond disk from deposits annealed at 1000 ◦C/1 h. The test specimens were then sintered at 1500 ◦C/2 h. The sintered specimens were ground to be 2 mm × 2.5 mm × 25 mm and the tensile side was polished with diamond paste of 1 m grain size. Bending strength was measured on a universal test machine (Z020, Zwick, Germany) with a jig for bending strength measurement. 3. Results and discussion 3.1. Optimization of suspension composition The suspensions with type HP Al2O3 and type 3YS ZrO2 were used for the optimization experiments (see Table 2). The Al2O3 and ZrO2 particles were charged negatively and the layer was deposited on the anode, in the same way as in work of Cihlar et al. [21]. The smoothness of deposit surface, deposit thickness and density were affected by suspension composition (the contents of monochloroacetic acid and polyvinylbutyral). The deposit had a smooth surface if the deposition was from a suspension containing more than 4 wt.% of monochloroacetic acid, irrespective of the PVB concentration. This phenomenon was probably due to the effect of MCAA concentration on electrophoretic behaviour of the suspension. The intensity of electric field during deposition was dependent on the contents of monochloroacetic acid and Fig. 2. Dependence of electric field intensity during deposition of alumina and zirconia on the concentration of MCAA and PVB in the suspension. polyvinylbutyral in the suspension (see Fig. 2). The drop in the intensity of electric field with increasing concentration of monochloroacetic acid was probably due to the increasing conductivity of the solution, which in the case of Al2O3 suspensions increased from 2.9×10−4 S m−1 (HP-1-0) [21] to 23.8 × 10−4 S m−1 (HP-25-0) [21]. The relation between velocity of particle motion (v), electrophoretic mobility (µ) and electric field intensity (E) is given by the equation: v = µ × E. (2) Since electrophoretic mobility of the Al2O3 and ZrO2 particles at an MCAA concentration of more than 0.85 wt.% is independent of its concentration [21], the velocity of the particles in the suspension decreased with decreasing electric field intensity and thus with increasing MCAA concentration. The slower motion of particles probably led to their more effective arrangement on the anode, which was confirmed by the higher green density of deposits with a higher MCAA content (see Fig. 3). The final relative deposit Fig. 3. Dependence of the relative density of alumina deposit after annealing and sintering on the concentration of MCAA and PVB in the suspension
K. Maca et al. /Ceramics International 30(2004)843-852 3.2. Suspension sedimentation and its elimination PVB CONCENTRATION [wt%! O Ow%(HP AL-o The deposits obtained in the previous part of the work Wt% (HP Al,o were of non-constant thickness. which increased contin- ◇4w%(HPAO uously towards the bottom of electrophoretic cell. The reason lay in suspension sedimentation due to earth's grav- itation, which in the case of horizontal arrangement of the electrophoretic cell is perpendicular to the direction of elec- trophoretic force. The following experiments, which wer conducted with the HP-15-0 suspension on the basis of the results of the preceding section, had as their aim the de scription and elimination of suspension sedimentation and Its consequences. MCAA CONCENTRATION The first to be studied was deposition with the interrupted deposition mode, during which the deposit thickness was Fig. 4. Dependence of alumina and zirconia deposit weight on the con measured continuously but the suspension was not stirred centration of MCAA and PVB in the suspension. The dependence of deposit thickness on time at three depths below the suspension level is given in the graph of Fig. 5. It density was ca. 99.5% and was independent of both Pvb can be seen from the graph that after some time the deposit and MCAA concentration(provided MCAA concentration thickness values differed with the place of measurement be- exceeded 4 wt % cause sedimentation occurred in the suspension As can be seen from Fig. 2, the addition of polyvinyl To prevent the sedimentation of powder, the suspension butyral to the suspension led to a slight increase in electric had to be stirred. Since after 5 min of deposition without stir current intensity during deposition. As established in work ring no differences in deposit thickness could be observed of Cihlar et al. [21], increasing polyvinylbutyral content re- at the individual levels of measuring(see Fig. 5), the 5-min duced the electrophoretic mobility of particles, which led interval was chosen as appropriate for stirring the suspen- to a lower weight of the deposit. This result has also been sion. Fig. 6 gives the dependence of deposit thickness proved in the present work as can be seen from Fig. 4, which time as obtained for interrupted deposition with suspension illustrates the dependence of deposit weight on MCAA and stirring. It can be seen that the deposition was uniform at al PVB concentrations. The electrophoretic mobility of parti- the three levels observed In the case that particle concentra cles is given the Henry eq tion in the suspension in vertical direction does not change (the suspension does not sediment or it is stirred)the de ( pendence of deposit weight(m)on time(O) is given by the Zhang equation [24](mo is the initial weight of powder in where e is the permittivity of the medium, s is the zeta suspension and d is the distance between electrodes otential, n is the dynamic viscosity of the medium, and f(Rk-)is the function of the product of particle radius(R) m=mo(1 -e-uE/d) and the thickness of electric double layer(k-). The drop in the electrophoretic mobility of particles in the suspension was due to the reduction in the s-potential of particles and the increase in the viscosity of the medium caused by the content of polyvinylbutyral in the suspension [21] The addition of polyvinylbutyral had no positive effect on E-d.58 mm below supension le either the final deposit density(see Fig. 3)or the flatness of g their surfaces. Since it reduced the electrophoretic mobility 2 of particles in the suspension, it had a negative effect on deposit weight(Fig 4) From the viewpoint of the flatness of der which is important, for example, in the deposition of layered composites[23], the optimum amounts were 15 and 25 wt% of MCAA in isopropanol. At a concentration of 15 wt % of MCAA, the deposit yield was, of course, higher(see Fig. 4) On the basis of the above results the suspensions containing no PVB and with a 15 wt. content of mCAa in the liquid phase(i.e. 12.75 wt. in the suspension) were evaluated as Fig. 5. Dependence of alumina( type HP)deposit thickness on deposition optimum suspensions time and on the vertical position(non-stirred suspension
K. Maca et al. / Ceramics International 30 (2004) 843–852 847 Fig. 4. Dependence of alumina and zirconia deposit weight on the concentration of MCAA and PVB in the suspension. density was ca. 99.5% and was independent of both PVB and MCAA concentration (provided MCAA concentration exceeded 4 wt.%). As can be seen from Fig. 2, the addition of polyvinylbutyral to the suspension led to a slight increase in electric current intensity during deposition. As established in work of Cihlar et al. [21], increasing polyvinylbutyral content reduced the electrophoretic mobility of particles, which led to a lower weight of the deposit. This result has also been proved in the present work as can be seen from Fig. 4, which illustrates the dependence of deposit weight on MCAA and PVB concentrations. The electrophoretic mobility of particles is given by the Henry equation [22]: µ = 2 3 εζ η f(Rκ−1) (3) where ε is the permittivity of the medium, ζ is the zeta potential, η is the dynamic viscosity of the medium, and f(Rk−1) is the function of the product of particle radius (R) and the thickness of electric double layer (k−1). The drop in the electrophoretic mobility of particles in the suspension was due to the reduction in the ζ-potential of particles and the increase in the viscosity of the medium caused by the content of polyvinylbutyral in the suspension [21]. The addition of polyvinylbutyral had no positive effect on either the final deposit density (see Fig. 3) or the flatness of their surfaces. Since it reduced the electrophoretic mobility of particles in the suspension, it had a negative effect on deposit weight (Fig. 4). From the viewpoint of the flatness of deposit surface, which is important, for example, in the deposition of layered composites [23], the optimum amounts were 15 and 25 wt.% of MCAA in isopropanol. At a concentration of 15 wt.% of MCAA, the deposit yield was, of course, higher (see Fig. 4). On the basis of the above results the suspensions containing no PVB and with a 15 wt.% content of MCAA in the liquid phase (i.e. 12.75 wt.% in the suspension) were evaluated as optimum suspensions. 3.2. Suspension sedimentation and its elimination The deposits obtained in the previous part of the work were of non-constant thickness, which increased continuously towards the bottom of electrophoretic cell. The reason lay in suspension sedimentation due to earth’s gravitation, which in the case of horizontal arrangement of the electrophoretic cell is perpendicular to the direction of electrophoretic force. The following experiments, which were conducted with the HP-15-0 suspension on the basis of the results of the preceding section, had as their aim the description and elimination of suspension sedimentation and its consequences. The first to be studied was deposition with the interrupted deposition mode, during which the deposit thickness was measured continuously but the suspension was not stirred. The dependence of deposit thickness on time at three depths below the suspension level is given in the graph of Fig. 5. It can be seen from the graph that after some time the deposit thickness values differed with the place of measurement because sedimentation occurred in the suspension. To prevent the sedimentation of powder, the suspension had to be stirred. Since after 5 min of deposition without stirring no differences in deposit thickness could be observed at the individual levels of measuring (see Fig. 5), the 5-min interval was chosen as appropriate for stirring the suspension. Fig. 6 gives the dependence of deposit thickness on time as obtained for interrupted deposition with suspension stirring. It can be seen that the deposition was uniform at all the three levels observed. In the case that particle concentration in the suspension in vertical direction does not change (the suspension does not sediment or it is stirred) the dependence of deposit weight (m) on time (t) is given by the Zhang equation [24] (m0 is the initial weight of powder in suspension and d is the distance between electrodes): m = m0(1 − e−(µE/d)t) (4) Fig. 5. Dependence of alumina (type HP) deposit thickness on deposition time and on the vertical position (non-stirred suspension)