Availableonlineatwww.sciencedirect.com Science Direct E噩≈RS ELSEVIER Joumal of the European Ceramic Society 30(2010)1195-1202 www.elsevier.comlocate/jeurceramsoc Fabrication of textured alumina by orienting template particles during electrophoretic deposition Li Zhang, Jef vleugels, Omer Van der biest K ULeunen, Department of Metallurgy and Materials Engineering, Kasteelpark Arenberg 44, B-3001 Heverlee, belgium Available online 16 July 2009 orientation during EPD was examined with respect to the impact of the electric field force, gravity and hydrodynamic force in two different deposition cells with vertically or horizontally positioned deposition electrodes. A sharp(00 1)fbre texture'was obtained after templated grain growth during sintering of a deposit formed from a stirred 5 vol% platelet containing suspension in a vertical deposition cell. The texture was characterized by means of the Lotgering factor, texture index and electron backscattering diffraction(EBSD) 2009 Elsevier Ltd. All rights reserved. Keywords: Suspensions; Platelets; Al2 O3: Texture Introduction This work shows the possibility to align platelet template dur- ing electrophoretic deposition(EPD). EPD is a colloidal proces Textured materials have been investigated extensively in the wherein particles suspended in a stable suspension are deposited past few years due to their improved electronic and structural on one of the electrodes by applying an electric field. The sim- properties. If polycrystalline particles could be aligned, they plicity in experimental set-up leads to low equipment cost. EPD may be able to exhibit the anisotropic characteristics typical is also a fast compacting technique and the deposit thickness of single crystals. For instance, as an important engineer- can be easily controlled. EPD has the capability to form com- ing ceramic, textured a-alumina is widely investigated and the plex shapes and patterns. Moreover, EPD can be applied on a property of this textured a-alumina was reported to be signif- wide range of substrates such as porous ceramics, conductive icantly enhanced 24 Textured materials can be produced by a polymers or metals. 0. I variety of techniques including magnetic orientation and tem- In this study, hexagonal a-alumina platelet templates were growth(TGG). -6Templated grain growth is widely aligned in a fine matrix alumina powder by EPD. After tem- investigated. 4-6 Large and anisotropically shaped template par- plated grain growth, a highly textured material was developed ticles are homogeneously aligned in a fine matrix powder during The mechanism of platelet alignment during deposition was compacting. During sintering, textured material was obtained by investigated considering the impact of orientation forces grain growth in the anisotropic direction of the aligned template particles. The template alignment in the green compact is a crit ical factor in TGG since it determines the final texturisation. 2. Experimental procedure Templates could be aligned by various powder consolidation techniques including uniaxial pressing, slip casting, tape cast- High purity fine a-alumina powder(0.3 um, Baikowski ing, gel casting, centrifugal casting and extrusion. - In those grade SM8, France)was used as matrix powder, whilst hexag- cases, the active forces orientate the template grains during the onal alumina platelets(10-15 um in diameter, -0.5 um thick, powder consolidation process ELF Atochem, France)were used as templates. The morphology of the alumina matrix powder and platelet template is shown in milled for 24 h on a multi-directional mixer(type Turbula, WAB Corresponding author. Tel: +32 16 321777: fax: +32 16 321992 Switzerland) at 70rpm in absolute ethanol(99.9% Merck Bel E-mail address: LiZhang@mtm kuleuven. be(L. Zhang gium) with 0.5 vol% de-ionised water. Zirconia milling balls 0955-2219 front matter@ 2009 Elsevier Ltd. All rights reserved. doi: 10. 1016/j-jeurceramsoc. 2009.06.026
Available online at www.sciencedirect.com Journal of the European Ceramic Society 30 (2010) 1195–1202 Fabrication of textured alumina by orienting template particles during electrophoretic deposition Li Zhang ∗, Jef Vleugels, Omer Van der Biest K.U.Leuven, Department of Metallurgy and Materials Engineering, Kasteelpark Arenberg 44, B-3001 Heverlee, Belgium Available online 16 July 2009 Abstract (0 0 1)-Textured -alumina has been processed by electrophoretic deposition (EPD) and templated grain growth. The mechanism of platelet template orientation during EPD was examined with respect to the impact of the electric field force, gravity and hydrodynamic force in two different deposition cells with vertically or horizontally positioned deposition electrodes. A sharp (0 0 1) ‘fibre texture’ was obtained after templated grain growth during sintering of a deposit formed from a stirred 5 vol% platelet containing suspension in a vertical deposition cell. The texture was characterized by means of the Lotgering factor, texture index and electron backscattering diffraction (EBSD). © 2009 Elsevier Ltd. All rights reserved. Keywords: Suspensions; Platelets; Al2O3; Texture 1. Introduction Textured materials have been investigated extensively in the past few years due to their improved electronic and structural properties. If polycrystalline particles could be aligned, they may be able to exhibit the anisotropic characteristics typical of single crystals.1 For instance, as an important engineering ceramic, textured -alumina is widely investigated and the property of this textured -alumina was reported to be significantly enhanced.2–4 Textured materials can be produced by a variety of techniques including magnetic orientation and templated grain growth (TGG).3–6 Templated grain growth is widely investigated.4–6 Large and anisotropically shaped template particles are homogeneously aligned in a fine matrix powder during compacting. During sintering, textured material was obtained by grain growth in the anisotropic direction of the aligned template particles. The template alignment in the green compact is a critical factor in TGG since it determines the final texturisation. Templates could be aligned by various powder consolidation techniques including uniaxial pressing, slip casting, tape casting, gel casting, centrifugal casting and extrusion.5–9 In those cases, the active forces orientate the template grains during the powder consolidation process. ∗ Corresponding author. Tel.: +32 16 321777; fax: +32 16 321992. E-mail address: Li.Zhang@mtm.kuleuven.be (L. Zhang). This work shows the possibility to align platelet template during electrophoretic deposition (EPD). EPD is a colloidal process wherein particles suspended in a stable suspension are deposited on one of the electrodes by applying an electric field. The simplicity in experimental set-up leads to low equipment cost. EPD is also a fast compacting technique and the deposit thickness can be easily controlled. EPD has the capability to form complex shapes and patterns. Moreover, EPD can be applied on a wide range of substrates such as porous ceramics, conductive polymers or metals.10,11 In this study, hexagonal -alumina platelet templates were aligned in a fine matrix alumina powder by EPD. After templated grain growth, a highly textured material was developed. The mechanism of platelet alignment during deposition was investigated considering the impact of orientation forces. 2. Experimental procedure High purity fine -alumina powder (∼0.3m, Baikowski grade SM8, France) was used as matrix powder, whilst hexagonal alumina platelets (10–15m in diameter, ∼0.5m thick, ELF Atochem, France) were used as templates. The morphology of the alumina matrix powder and platelet template is shown in Fig. 1. Fine alumina matrix powder suspensions were initially milled for 24 h on a multi-directional mixer (type Turbula, WAB, Switzerland) at 70 rpm in absolute ethanol (99.9% Merck Belgium) with 0.5 vol% de-ionised water. Zirconia milling balls 0955-2219/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jeurceramsoc.2009.06.026
L Zhang et al. Joumal of the European Ceramic Society 30(2010)1195-1202 clectrode Fig. 2. EPD cell with horizontally positioned deposition electrode counter deposition electrode electrode magnetic stirrer bar To estimate the influence of the electric field force, gravity and hydrodynamic force on the alignment of the platelets, two different set-ups were used, i.e., with a vertical and horizontal EPD cell. Three configurations were studied in a slightly mod- .V Spot Magn Det WD Exp ified horizontal cell as summarised in Table 1. The standard horizontal EPD cell is schematically presented in Fig. 2. In case of an applied suspension flow(configuration 2 in Table 1),a Fig. 1. SEM image of (a) as received alumina grade SM8 and(b) platelets suspension was flown through the deposition cell by means of a suspension circulation system driven by a peristaltic pump Tosoh grade TZ-3Y) with a diameter of 5 mm were added to The distance between the flat disc shaped electrodes was 3. 5 cm the polyethylene container to facilitate breaking the agglomer- and the electrodes had a diameter of 3. 7 cm. The suspension ates. Afterwards, platelets (5 wt% relative to the total amount in the reservoir was stirred to avoid sedimentation. In addition of powder) were added to the ball-milled suspension, whilst the flow-through system also decreases the sedimentation of the n-butylamine (99.5% Acros Belgium, 3. 4 vol% relative to the suspension in the cell with the fluid flow. Homogeneous deposits suspension volume)was added to negatively charge the particles were made by pumping 200 ml suspension at 1 ml/s through the and Dolapix Ce-64(Zschimmer Schwarz, Germany, 1. I wt% deposition cell. The suspension circulation system was not used relative to the powder mass)was added as dispersant. For com- in the other two configurations (3 and 4 of Table 1) where only parison, a suspension without platelets was also prepared. This 50 ml suspension was used to fill the EPD cell. Configuration suspension was magnetically stirred for 60 min, ultrasonicated 4 and 3 form respectively depositions along and in the oppo- in an ultrasonic bath(Branson 2510)for 15 min and magnetically site direction of the gravity force, in which the perforated top stirred for another 15 min electrode( Fig. 2)was replaced with a normal one. EPD was con- Density, Lotgering factor and texture index for different EPD configurations Green density [%] Sintered density [%I ening factor index Vertical Vertical stirred 594 0.49 18.32 Horizontal Bottom Flow through Horizontal 8.12 Horizontal Bottom Stagnant Stagnant
1196 L. Zhang et al. / Journal of the European Ceramic Society 30 (2010) 1195–1202 Fig. 1. SEM image of (a) as-received alumina grade SM8 and (b) platelets. (Tosoh grade TZ-3Y) with a diameter of 5 mm were added to the polyethylene container to facilitate breaking the agglomerates. Afterwards, platelets (5 wt% relative to the total amount of powder) were added to the ball-milled suspension, whilst n-butylamine (99.5% Acros Belgium, 3.4 vol% relative to the suspension volume) was added to negatively charge the particles and Dolapix Ce-64 (Zschimmer & Schwarz, Germany, 1.1 wt% relative to the powder mass) was added as dispersant. For comparison, a suspension without platelets was also prepared. This suspension was magnetically stirred for 60 min, ultrasonicated in an ultrasonic bath (Branson 2510) for 15 min and magnetically stirred for another 15 min. Fig. 2. EPD cell with horizontally positioned deposition electrode. Fig. 3. EPD cell with vertically positioned deposition electrode. To estimate the influence of the electric field force, gravity and hydrodynamic force on the alignment of the platelets, two different set-ups were used, i.e., with a vertical and horizontal EPD cell. Three configurations were studied in a slightly modified horizontal cell, as summarised in Table 1. The standard horizontal EPD cell is schematically presented in Fig. 2. In case of an applied suspension flow (configuration 2 in Table 1), a suspension was flown through the deposition cell by means of a suspension circulation system driven by a peristaltic pump. The distance between the flat disc shaped electrodes was 3.5 cm and the electrodes had a diameter of 3.7 cm. The suspension in the reservoir was stirred to avoid sedimentation. In addition, the flow-through system also decreases the sedimentation of the suspension in the cell with the fluid flow. Homogeneous deposits were made by pumping 200 ml suspension at 1 ml/s through the deposition cell. The suspension circulation system was not used in the other two configurations (3 and 4 of Table 1) where only 50 ml suspension was used to fill the EPD cell. Configuration 4 and 3 form respectively depositions along and in the opposite direction of the gravity force, in which the perforated top electrode (Fig. 2) was replaced with a normal one. EPD was conTable 1 Density, Lotgering factor and texture index for different EPD configurations. Configuration Cell Deposition Suspension Green density [%] Sintered density [%] Lotgering factor Texture index 1 Vertical Vertical stirred 59.4 97.9 0.49 18.32 2 Horizontal Bottom Flow through 58.9 97.5 0.01 1.60 3 Horizontal Top Stagnant 62.7 99.7 0.21 8.12 4 Horizontal Bottom Stagnant 61.1 98.0 0.02 2.52 5 Vertical Vertical Stagnant – – 0.12 8.05
L Zhang et al. / Journal of the European Ceramic Society 30(2010)1195-1202 1197 ducted for 450s. The 50 ml vertical EPD cell, presented in Fig 3, 3. Results consisted of two vertically positioned electrodes with a separa- tion distance of 3.5 cm and a surface area of 9 cm. The edges 3.1. Vertical deposition from a stirred suspension of the deposition electrode were shielded by a non-conductive PTFE cover. Two configurations were studied in the vertical cell, The influence of the platelet template can be assessed by i.e., with a stagnant( configuration 5)and a magnetically stirred comparison with the random alumina powder based ceramic 250rpm)(configuration 1)suspension. The texture of the platelets containing material was confirmed Constant voltage anodic electrophoretic deposition was per- by XRD, as shown in Fig. 4. The diffraction spectrum of the formed with freshly prepared suspensions, using a F.U. G(type random ceramic(Fig 4(a)) is consistent with the JCPDS card MCN 1400-50)power supply. The pH* of the suspension before of alumina(card number 43-1484 ), which means that no signif EPD was 11.45 pH* denotes the operational pH for which a icant texture was formed in the sample. The(1 10)and (300) standard ph electrode was used to measure the pH in ethanol diffraction peak intensities are pronounced in the spectrum of suspensions. 2 The conductivity of the suspension at room tem- the sample, cross-sectioned perpendicular to the depositionelec- perature was 24.5 uS/cm, as measured by a conductivity sensor trode(perpendicular section, see Fig. 4(b)). Those peaks are Cond Level 2 type, WTW). After EPD, the deposit was care- stronger than in the random sample as well as in the sample fully removed from the suspension and dried in air. Afterwards, cross-sectioned parallel to the electrode(parallel section, see the deposit was sintered in air(Nabertherm furnace, Germany) Fig 4(c). The(006)and(00 12)peaks are hardly observed in at 1550C for I h with a heating rate of 10C/min. The green the perpendicular section and the random sample, whereas they and sintered density was measured in ethanol by the Archimedes are prominent in the parallel section. The(104)and(1010) method peaks are also stronger in the parallel section than in the random The microstructure of the polished and thermally etched sample and the perpendicular section. These results imply that surface of sintered samples was investigated by scanning elec- a large volume fraction of strong(00 1)textured alumina grains tron microscopy (SEM, XL30-FEG, FEL, Netherlands). Texture are formed in the platelet containing material. The c-axis is ori analysis was performed by X-ray diffraction(type Seifert ented perpendicular to the surface of the deposition electrode 3003), pole figure measurements (Siemens D500 Texture Based on the XRD spectrum, the Lotgering factor is calculated Stress) and Electron Back-Scatter Diffraction(EBSD, EDAX, to be 0.49, which implies a well-textured material. Netherlands) The microstructural anisotropy is confirmed by the SEM The Lotgering factor is widely used in literature to charac crographs shown in Fig. 5. SEM analysis at different loca- terize the texture degree of hexagonal a-alumina, 3, 4 and can tions in the sintered deposit revealed that the platelet particles be obtained from the X-ray diffraction pattern of a sample. The are nearly homogeneously distributed throughout the deposit Lotgering factor is defined as: The seeded platelets acted as templates for the grain growth C∑100/∑/(hA)-(∑P(0O∑P(hkD) during sintering. The basal planes of the grown platelets have f 1-∑(00/∑P(hkD) (1) been aligned parallel to the surface of the deposition electrode During sintering, the aligned platelet seeds grew very fast by with El(00D, the summation of all (00D peak intensities and means of coarsening, i.e by the consumption of the fine matrix 2I(hkl), the summation of all peak intensities in the spectrum alumina particles, resulting in a highly textured ceramic. Grain Superscript 0 corresponds to a random sample. Thef factor growth has been fast by means of grain boundary migration since changes between 0 and 1. A large f value implies a highly tex here is a substantial amount of pores trapped inside the grains tured material, since f=0 for a random sample and f=l for a Intergranular pores are also clearly observed in Fig. 5. The rel fully oriented material. ative density of this ceramic is 97. 9%0. Beside the intergranular In order to give a complete estimation of texture formation. the texture index was calculated depending on the orientation distribution function(ODF). The OdF was obtained from mea- sured pole figures by"Hexagonal ODF software system"(Dept MTM-K.U. Leuven). The texture index was used as an indica- tor of the sharpness of the texture, and is defined as the integral of the square of the ODF,f(g), over the entire orientation space L=∮Uf(g)2dg 010 The higher the value, the sharper the texture. A value close to 1 (c) (116) implies random texture. EbSd was used to examine the microstructure and 20304050607080 the regional gra 2 Theta 180 x 520 um- and the step size was 1 um. The grain orientation Fig. 4. X-ray diffraction pattern of sintered (a)a random alumina sample and maps and inverse pole figures were all generated from the experi- cross-sectioned template containing ceramic(b)perpendicular and(c)parallel mental data using commercial software (TSL OIM analysis 4.5). to the electrode surface
L. Zhang et al. / Journal of the European Ceramic Society 30 (2010) 1195–1202 1197 ducted for 450 s. The 50 ml vertical EPD cell, presented in Fig. 3, consisted of two vertically positioned electrodes with a separation distance of 3.5 cm and a surface area of 9 cm2. The edges of the deposition electrode were shielded by a non-conductive PTFE cover. Two configurations were studied in the vertical cell, i.e., with a stagnant (configuration 5) and a magnetically stirred (∼250 rpm) (configuration 1) suspension. Constant voltage anodic electrophoretic deposition was performed with freshly prepared suspensions, using a F.U.G. (type MCN 1400-50) power supply. The pH* of the suspension before EPD was 11.45. pH* denotes the operational pH for which a standard pH electrode was used to measure the pH in ethanol suspensions.12 The conductivity of the suspension at room temperature was 24.5S/cm, as measured by a conductivity sensor (Cond Level 2 type, WTW). After EPD, the deposit was carefully removed from the suspension and dried in air. Afterwards, the deposit was sintered in air (Nabertherm furnace, Germany) at 1550 ◦C for 1 h with a heating rate of 10 ◦C/min. The green and sintered density was measured in ethanol by the Archimedes method. The microstructure of the polished and thermally etched surface of sintered samples was investigated by scanning electron microscopy (SEM, XL30-FEG, FEI, Netherlands). Texture analysis was performed by X-ray diffraction (type Seifert 3003), pole figure measurements (Siemens D500 Texture & Stress) and Electron Back-Scatter Diffraction (EBSD, EDAX, Netherlands). The Lotgering factor is widely used in literature to characterize the texture degree of hexagonal -alumina,13,14 and can be obtained from the X-ray diffraction pattern of a sample. The Lotgering factor is defined as: f = I(00l)/ I(hkl) − I0(00l)/ I0(hkl) 1 − I0(00l)/ I0(hkl) (1) with ΣI(0 0 l), the summation of all (0 0 l) peak intensities and ΣI(hkl), the summation of all peak intensities in the spectrum. Superscript 0 corresponds to a random sample. The f factor changes between 0 and 1. A large f value implies a highly textured material, since f = 0 for a random sample and f = 1 for a fully oriented material. In order to give a complete estimation of texture formation, the texture index was calculated depending on the orientation distribution function (ODF). The ODF was obtained from measured pole figures by “Hexagonal ODF software system” (Dept MTM—K.U.Leuven). The texture index was used as an indicator of the sharpness of the texture, and is defined as the integral of the square of the ODF, f(g), over the entire orientation space15: T.I. = [f (g)]2dg (2) The higher the value, the sharper the texture. A value close to 1 implies random texture. EBSD was used to examine the microstructure and the regional grain orientation. The examined area was 180 × 520m2 and the step size was 1 m. The grain orientation maps and inverse pole figures were all generated from the experimental data using commercial software (TSL OIM analysis 4.5). 3. Results 3.1. Vertical deposition from a stirred suspension The influence of the platelet template can be assessed by comparison with the random alumina powder based ceramic. The texture of the platelets containing material was confirmed by XRD, as shown in Fig. 4. The diffraction spectrum of the random ceramic (Fig. 4(a)) is consistent with the JCPDS card of alumina (card number 43-1484), which means that no significant texture was formed in the sample. The (1 1 0) and (3 0 0) diffraction peak intensities are pronounced in the spectrum of the sample, cross-sectioned perpendicular to the deposition electrode (perpendicular section, see Fig. 4(b)). Those peaks are stronger than in the random sample as well as in the sample cross-sectioned parallel to the electrode (parallel section, see Fig. 4(c)). The (0 0 6) and (0 0 12) peaks are hardly observed in the perpendicular section and the random sample, whereas they are prominent in the parallel section. The (1 0 4) and (1 0 10) peaks are also stronger in the parallel section than in the random sample and the perpendicular section. These results imply that a large volume fraction of strong (0 0 1) textured alumina grains are formed in the platelet containing material. The c-axis is oriented perpendicular to the surface of the deposition electrode. Based on the XRD spectrum, the Lotgering factor is calculated to be 0.49, which implies a well-textured material. The microstructural anisotropy is confirmed by the SEM micrographs shown in Fig. 5. SEM analysis at different locations in the sintered deposit revealed that the platelet particles are nearly homogeneously distributed throughout the deposits. The seeded platelets acted as templates for the grain growth during sintering. The basal planes of the grown platelets have been aligned parallel to the surface of the deposition electrode. During sintering, the aligned platelet seeds grew very fast by means of coarsening, i.e., by the consumption of the fine matrix alumina particles, resulting in a highly textured ceramic. Grain growth has been fast by means of grain boundary migration since there is a substantial amount of pores trapped inside the grains. Intergranular pores are also clearly observed in Fig. 5. The relative density of this ceramic is 97.9%. Beside the intergranular Fig. 4. X-ray diffraction pattern of sintered (a) a random alumina sample and cross-sectioned template containing ceramic (b) perpendicular and (c) parallel to the electrode surface.
l198 L Zhang et al. Joumal of the European Ceramic Society 30(2010)1195-1202 lated texture index of thi e Is 3.2. Horizontal downward deposition from a fowing suspension the bottom and suspension flowing through. The platelets in the deposit however were not well aligned as shown in the eBsd of the perper (Fig. 6(b)). As shown in Fig. 6(e), the peaks are relatively weak and are not homogenously distributed implying that the pris matic planes are not well aligned. The corresponding X-ray pole 如Em figure confirms a very weak texturing after grain growth with a limited texture index of 1.60 (b) 3.3. Deposition from a stagnant suspension The abov nat the cell ical factor for platelet alignment during EPD. The electric field Eg force, gravity and hydrodynamic force applied on the platelets are believed to be the main factors influencing platelet alignment The influence of these forces is studied using 3 additional con- figurations as summarised in Table 1. in order to investigate the electric field force effect, an upward deposition was performed the horizontal cell without suspension flowing through(con- figuration 3). In order to investigate the gravity effect, EPD was performed from a stagnant suspension in the horizontal cell(con- figuration 4). In order to study the impact of the hydrodynamic force, the suspension is stirred comparing with stagnant suspen sion(configuration 5)or pumped through the cell as described in Sections 3.1 and 3.2 Fig. 5. Textured alumina deposited in a horizontal cell, cross-sectioned (a) Table 1 summarises the Lotgering factor and texture index parallel and (b) perpendicular to the deposition electrode. measured on the parallel cross-sectioned sintered deposits obtained under different EPd cell configurations, as well as the porosity,the residual porosity also results from platelet particle green and sintered density. The green relative density of all the constrained sintering, as discussed in literature. 6, 7 Since th deposits was in the 58.9-62.7% range, which is quite compara template platelets are essential for texture development, all fur. ble. After sintering, quite dense ceramics with some remaining ther investigated grades were prepared from platelet containing porosity due to constrained sintering were obtained. The den- sification of the powder matrix is significantly retarded by the suspensions. presence of large inclusions, i.e., platelets, resulting in a lower EBSD of perpendicularly cross-sectioned grades was investi- densification as reported in literature. 16. I The sintered density gated to characterize the texture. The EBSD pattern and inverse pole figure(IPF) of the above material grade are presented is proportional to the green density of the samples, as shown in ig. 6(a)and (d). The platelets are well aligned parallel to th Table 1. EPD from a stirred suspension in a vertical cell(con- deposition electrode surface, positioned at the bottom of the figuration 1)results in the highest Lotgering factor and texture picture in Fig. 6(a), revealing that the c-axis has been well- index, which indicates that the best texture was formed. The extent of texturisation is quite different for the five investigated aligned perpendicular to the surface of the deposition electrode, configurations. The mechanism of texture formation is discussed in Fig. 6(d)is large and the peak is homogeneously distributed at the edge of the IPF, proving that the prismatic planes are well ori- ented in the normal direction. The a-and b-axes are not aligned. 4. Discussion i.e., the material has no preferred in-plane orientation(a-axis or 4.1. Influence of the electric field force b-axis orientation)as indicated by the fact that the peak intensity is homogeneously distributed at the edge of the IPF. The electric field force is the driving force for powder con- In order to investigate the global sample area, X-ray pole solidation during EPD. In order to investigate the impact of the figures were used to characterize the macro-texture. The calcu- electric field force on platelet alignment, EPD from a stagnant
1198 L. Zhang et al. / Journal of the European Ceramic Society 30 (2010) 1195–1202 Fig. 5. Textured alumina deposited in a horizontal cell, cross-sectioned (a) parallel and (b) perpendicular to the deposition electrode. porosity, the residual porosity also results from platelet particle constrained sintering, as discussed in literature.16,17 Since the template platelets are essential for texture development, all further investigated grades were prepared from platelet containing suspensions. EBSD of perpendicularly cross-sectioned grades was investigated to characterize the texture. The EBSD pattern and inverse pole figure (IPF) of the above material grade are presented in Fig. 6(a) and (d). The platelets are well aligned parallel to the deposition electrode surface, positioned at the bottom of the picture in Fig. 6(a), revealing that the c-axis has been wellaligned perpendicular to the surface of the deposition electrode, as shown in Fig. 6(a). The peak intensity of the [0 0 1] IPF shown in Fig. 6(d) is large and the peak is homogeneously distributed at the edge of the IPF, proving that the prismatic planes are well oriented in the normal direction. The a- and b-axes are not aligned, i.e., the material has no preferred in-plane orientation (a-axis or b-axis orientation) as indicated by the fact that the peak intensity is homogeneously distributed at the edge of the IPF. In order to investigate the global sample area, X-ray pole figures were used to characterize the macro-texture. The calculated texture index of this sample is 18.32, which implies a sharp texture formation. 3.2. Horizontal downward deposition from a flowing suspension The same suspension composition was used for EPD in the horizontal cell, shown in Fig. 2, with the deposition electrode at the bottom and suspension flowing through. The platelets in the deposit however were not well aligned as shown in the EBSD pattern of the perpendicular cross-section of the sintered ceramic (Fig. 6(b)). As shown in Fig. 6(e), the peaks are relatively weak and are not homogenously distributed implying that the prismatic planes are not well aligned. The corresponding X-ray pole figure confirms a very weak texturing after grain growth with a limited texture index of 1.60. 3.3. Deposition from a stagnant suspension The above results clearly show that the cell geometry is a critical factor for platelet alignment during EPD. The electric field force, gravity and hydrodynamic force applied on the platelets are believed to be the main factors influencing platelet alignment. The influence of these forces is studied using 3 additional con- figurations as summarised in Table 1. In order to investigate the electric field force effect, an upward deposition was performed in the horizontal cell without suspension flowing through (con- figuration 3). In order to investigate the gravity effect, EPD was performed from a stagnant suspension in the horizontal cell (con- figuration 4). In order to study the impact of the hydrodynamic force, the suspension is stirred comparing with stagnant suspension (configuration 5) or pumped through the cell as described in Sections 3.1 and 3.2. Table 1 summarises the Lotgering factor and texture index measured on the parallel cross-sectioned sintered deposits obtained under different EPD cell configurations, as well as the green and sintered density. The green relative density of all the deposits was in the 58.9–62.7% range, which is quite comparable. After sintering, quite dense ceramics with some remaining porosity due to constrained sintering were obtained. The densification of the powder matrix is significantly retarded by the presence of large inclusions, i.e., platelets, resulting in a lower densification as reported in literature.16,17 The sintered density is proportional to the green density of the samples, as shown in Table 1. EPD from a stirred suspension in a vertical cell (con- figuration 1) results in the highest Lotgering factor and texture index, which indicates that the best texture was formed. The extent of texturisation is quite different for the five investigated configurations. The mechanism of texture formation is discussed below, based on these experimental findings. 4. Discussion 4.1. Influence of the electric field force The electric field force is the driving force for powder consolidation during EPD. In order to investigate the impact of the electric field force on platelet alignment, EPD from a stagnant
L Zhang et al. / Journal of the European Ceramic Society 30(2010)1195-1202 喝() 0001 1100 1210 0110 (d)0001 2110 max=2.5 2213 1100 12001101210 2110 max=1.638 0001 100 1 01101210 min=0.191 70 um Fig. 6. EBSD data on a textured alumina. EBSD pattern of the perpendicular cross-section of the sample formed(a)in vertical cell with stirred suspension; (b)in orizontal cell with suspension fowing through; (c) corresponding colour coded map and (d)[001] IPF for(a), and(e)100 1]IPF for(b) suspension in a horizontal cell with a deposition electrode on the top(configuration 3)was investigated. In this case, the down- ward gravity force on the platelets is counteracted by the upward electric field force and the platelet alignment in the deposit is mainly induced by the electric field force. As shown in Fig. 7, the basal planes of the platelets are well aligned parallel to the depo- sition electrode surface. The electric field force can orientate platelets in two possible ways, i.e., during electrophoresis or One possible mechanism is that the electric field force aligns the platelets during electrophoresis due to the charge distribution on the platelet surfaces. The electrical charge on the basal plane of the platelets is different from that on the side plane due to t large difference in surface area. Under the present experimental conditions at pH 1l. 4, all platelet surfaces are negatively charged although the natural charge density could be different between the basal plane and prismatic plane. 8-20 The electric field force 10 am applied on the basal plane is therefore larger than on the pris- matic plane. The platelet alignment mechanism may depend on Fig. 7. SEM micrograph of a perpendicularly cross-sectioned ceramic obtained the polarisation of the electrical double layer in the electric field by upward EPD in a horizontal cell without suspension flowing through
L. Zhang et al. / Journal of the European Ceramic Society 30 (2010) 1195–1202 1199 Fig. 6. EBSD data on a textured alumina. EBSD pattern of the perpendicular cross-section of the sample formed (a) in vertical cell with stirred suspension; (b) in horizontal cell with suspension flowing through; (c) corresponding colour coded map and (d) [0 0 1] IPF for (a), and (e) [0 0 1] IPF for (b). suspension in a horizontal cell with a deposition electrode on the top (configuration 3) was investigated. In this case, the downward gravity force on the platelets is counteracted by the upward electric field force and the platelet alignment in the deposit is mainly induced by the electric field force. As shown in Fig. 7, the basal planes of the platelets are well aligned parallel to the deposition electrode surface. The electric field force can orientate the platelets in two possible ways, i.e., during electrophoresis or upon deposition. One possible mechanism is that the electric field force aligns the platelets during electrophoresis due to the charge distribution on the platelet surfaces. The electrical charge on the basal plane of the platelets is different from that on the side plane due to the large difference in surface area. Under the present experimental conditions at pH 11.4, all platelet surfaces are negatively charged although the natural charge density could be different between the basal plane and prismatic plane.18–20 The electric field force applied on the basal plane is therefore larger than on the prismatic plane. The platelet alignment mechanism may depend on the polarisation of the electrical double layer in the electric field, Fig. 7. SEM micrograph of a perpendicularly cross-sectioned ceramic obtained by upward EPD in a horizontal cell without suspension flowing through