《复合材料 Composites》课程教学资源（学习资料）第五章 陶瓷基复合材料_4Fabrication of super-high-strength microchanneled Al2O3

Availableonlineatwww.sciencedirect.com .e Science Direct Scripta materialia ELSEVIER cripta Materialia 61(2009)686-689 www.elsevier.com/locate/scriptam Fabrication of super-high-strength microchanneled Al2O3-ZrO2 ceramic composites with fibrous microstructure Byong Taek Lee"and Swapan Kumar Sarkar Biomedical Engineering and Materials Department, School of Medicine, Soonchunhyang University 366-1, Sangyoung-dong, Cheonan City 330-090, South Korea Received 11 May 2009: accepted 30 May 2009 Available online 9 june 2009 Continuously porous Al,Oxmonolinic-ZrO,)/tetragonal-zrO2 ceramic composites with tailored fine-scale fibrous microstruc- ture was fabricated by a multipass extrusion process. Precise control of the microstructure was achieved, giving AlO(m-zrO2) fibrous phase (3.5 um)in the t-ZrO2 phase(1. 5 um thickness ). The pore size and porosity was 175 um and 36%. The rounded micro- channels and fine geometric control of the fibrous microstructure result in composites with a remarkably high bending strength value of 588 MPa which is 2- to 6-fold higher than their monolithic counterparts o 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved Keywords: Extrusion; Ceramics: Ceramic matrix composites; Microstructure Due to their excellent mechanical, thermal and mesoscale. Template and replica methods with fugitive chemical properties, Al2O3-ZrO2 composites are utilized pore-forming agents have been reported for different as advanced materials in many industrial applications ceramics [9, 10]. However, the size and distribution of such as high-temperature structural materials, refractory pores was not homogeneous and the mechanical proper materials, cutting and grinding tools, high-temperature ties were also very poor. Another problem with this type corrosion-resistant materials, various electric and elec- of porous body is that the tronic applications, supportive biomaterials for human on a higher extent of porosity, which significantly de- hard tissue, high-performance filters and membranes, grades the mechanical property of the ceramics due to var- and supports for catalysts, etc. [1, 2]. However, like other ious internal flaws in the dense zone of the composites. ceramics, they have an intrinsic drawback of low fracture Porous Al2O3 with hierarchical micromorphologies and toughness. Many approaches to improve the fracture a well-oriented biomorphic structure has been fabricated toughness as well as other mechanical properties of the with a pyrolyzed carbon template and subsequent al va AlO3-ZrO, systems, including significant strengthening, por infiltration and oxidation [ll]. The porosity was high have been reported. Microporous Al2O3-ZrO2 ceramics but the material lacked significant structural strength have been studied for various applications including hot Moreover, as it depends on the pyrolysis of a natura gas and molten metal filtration, membrane support, cata- plant, the flexibility of realizing tailored micromorphol- lyst support, refractory applications and biomaterials ogy and porosity is restricted applications [3-7]. The functional behavior and use of A relatively new kind of highly ordered porous mem- these porous materials largely depends on the pore size brane of Al_, was fabricated by an anodizing method and extent of porosity. Various methods have been used [12]. However, this process is only applicable to thin to date to make monolithic or composite porous struc monolithic membranes and is not suitable for bulk tures from AlO3 and ZrO2 ceramics. Mesoporous materials or composites. Furthermore, the pore size is Al2O3 or Zro, membrane has been produced by the restricted to a very narrow range, mainly in the sub sol-gel method [8] but the bulk property of the material micrometer level. The strength considerations for these was poor and porosity was restricted to only a nano-or nanoporous ceramics have rarely been discussed. In both the microporous and mesoporous materials, the usability and performance of the porous composites largely Corresponding author. Tel. +82 41 570 2427: fax: +82 41 577 depend on their mechanical strength. Fabrication of 2415: e-mail: Ibt(@sch. ac kr porous ceramics with significant mechanical strength 1359-6462/S.see front matter 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. do: 10. 1016/j.scriptamat. 2009.05.047

Fabrication of super-high-strength microchanneled Al2O3–ZrO2 ceramic composites with fibrous microstructure Byong Taek Lee* and Swapan Kumar Sarkar Biomedical Engineering and Materials Department, School of Medicine, Soonchunhyang University, 366-1, Sangyoung-dong, Cheonan City 330-090, South Korea Received 11 May 2009; accepted 30 May 2009 Available online 9 June 2009 Continuously porous Al2O3–(monolinic-ZrO2)/tetragonal-ZrO2 ceramic composites with tailored fine-scale fibrous microstruc￾ture was fabricated by a multipass extrusion process. Precise control of the microstructure was achieved, giving Al2O3–(m-ZrO2) fibrous phase (3.5 lm) in the t-ZrO2 phase (1.5 lm thickness). The pore size and porosity was 175 lm and 36%. The rounded micro￾channels and fine geometric control of the fibrous microstructure result in composites with a remarkably high bending strength value of 588 MPa which is 2- to 6-fold higher than their monolithic counterparts. 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Extrusion; Ceramics; Ceramic matrix composites; Microstructure Due to their excellent mechanical, thermal and chemical properties, Al2O3–ZrO2 composites are utilized as advanced materials in many industrial applications such as high-temperature structural materials, refractory materials, cutting and grinding tools, high-temperature corrosion-resistant materials, various electric and elec￾tronic applications, supportive biomaterials for human hard tissue, high-performance filters and membranes, and supports for catalysts, etc. [1,2]. However, like other ceramics, they have an intrinsic drawback of low fracture toughness. Many approaches to improve the fracture toughness as well as other mechanical properties of the Al2O3–ZrO2 systems, including significant strengthening, have been reported. Microporous Al2O3–ZrO2 ceramics have been studied for various applications including hot gas and molten metal filtration, membrane support, cata￾lyst support, refractory applications and biomaterials applications [3–7]. The functional behavior and use of these porous materials largely depends on the pore size and extent of porosity. Various methods have been used to date to make monolithic or composite porous struc￾tures from Al2O3 and ZrO2 ceramics. Mesoporous Al2O3 or ZrO2 membrane has been produced by the sol–gel method [8], but the bulk property of the material was poor and porosity was restricted to only a nano- or mesoscale. Template and replica methods with fugitive pore-forming agents have been reported for different ceramics [9,10]. However, the size and distribution of pores was not homogeneous and the mechanical proper￾ties were also very poor. Another problem with this type of porous body is that the interconnectivity of pores relies on a higher extent of porosity, which significantly de￾grades the mechanical property of the ceramics due to var￾ious internal flaws in the dense zone of the composites. Porous Al2O3 with hierarchical micromorphologies and a well-oriented biomorphic structure has been fabricated with a pyrolyzed carbon template and subsequent Al va￾por infiltration and oxidation [11]. The porosity was high but the material lacked significant structural strength. Moreover, as it depends on the pyrolysis of a natural plant, the flexibility of realizing tailored micromorphol￾ogy and porosity is restricted. A relatively new kind of highly ordered porous mem￾brane of Al2O3 was fabricated by an anodizing method [12]. However, this process is only applicable to thin monolithic membranes and is not suitable for bulk materials or composites. Furthermore, the pore size is restricted to a very narrow range, mainly in the sub￾micrometer level. The strength considerations for these nanoporous ceramics have rarely been discussed. In both the microporous and mesoporous materials, the usability and performance of the porous composites largely depend on their mechanical strength. Fabrication of porous ceramics with significant mechanical strength 1359-6462/$ - see front matter 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.scriptamat.2009.05.047 * Corresponding author. Tel.: +82 41 570 2427; fax: +82 41 577 2415; e-mail: lbt@sch.ac.kr Available online at www.sciencedirect.com Scripta Materialia 61 (2009) 686–689 www.elsevier.com/locate/scriptamat

B. T. Lee, S. K Sarkar/ Scripta Materialia 61(2009)686-689 and microporosity presents another concern requiring t-ZrO2(TZ-3Y, Tosoh, Japan) with a particle size of further investigation. A through-channel porosity with 80 nm. Carbon powder(<15 um, Aldrich, USA)was defined geometry can be seen as a compromise for this used as a pore-forming agent. Al2O3 25% m-ZrO2 strength and porosity consideration. It is also highly was ball milled in ethanol and dried. t-ZrO2, AlO3- desirable in case of applications associated with fluid flow (m-ZrO2), and carbon powder were separately shear or porosity where high interconnectivity is required. mixed with a polymer binder, ethylene vinyl acetate Channelled microporosity with improved strength can (EVA)(Elvax 250 and Elvax 210, Dupont, USA), in a act as a holder or container of other kinds of mesoporous heated blender( Shina Platec Co., Suwon, South Korea) system and may find a range of applications with stearic acid as a lubricant In the first step of the Recently, a multipass extrusion process has been experiment, a t-ZrO2 shell was made by warm pressing investigated with respect to the fabrication of a porous and the Al2O3 core was fabricated by extrusion to make body [13-15]. Ordered microchanneled porous bodies a 30 mm diameter feed roll. The t-ZrO2/Al,O3 volume with monolithic ceramics were fabricated. However ratio was chosen to be 60/40. The feed roll was then ex their mechanical properties were significantly lower truded in a cylindrical die at 120 C to obtain 3.5 mm diameter filaments. These were the first-pass core/shell rable or higher compared to those yielded by conven- filaments, and the core/ shell arrangement of the filament tional porous composite fabrication processes such as accounts for the core/shell microstructure of the frame emplate [4] foaming [5] and leaching methods [6]. For region in the final porous composites. Sixty-one fila- all porous materials, monolithic ceramics or composites ments were then reloaded in the same die and extruded with random distributions of each constituent phase to obtain second-pass filaments. These second-pass have been investigated thus far. Microstructural modifi- core/shell filaments are 3.5 mm in diameter and contain cation in the frame region of the macroporous body has 61 core/shells in a close-packed arrangement. Polymer not been investigated previously. It is very difficult mixed carbon powder was extruded with the same die tailor the microstructure in the frame using existing fab- at 120C to obtain 3.5 mm diameter filaments rication methods. The present study is thought to be the To make a porous body, the carbon filaments and the first attempt to tailor the microstructure of a pore frame, second-pass core/shell Al2O3(m-ZrO2)t-zrO2 fila in approach that could dramatically improve the ments were arranged in the previously used steel die. mechanical properties of Al2O3-ZrO2 porous composite In the die, two outer layers were comprised of second- systems pass core/shell filaments and the inner layers were filled n this work, a fibrous monolithic process is used for with carbon filaments. The arrangement was compacted the first time to tailor the frame region of a microchan and extruded at 120oC at an extrusion rate of neled macroporous body, yielding a highly ordered and 8 mm min. The obtained 3. 5 mm diameter filament unidirectional fibrous microstructure resembling a bam- contained an inner carbon core and a ceramic shell with boo-like biomimetic microstructure(see Figure 1). The characteristic core/shell arrangements. This is termed frame of the porous body was made with Al2O3(tetrag the first-pass green porous body. Again, 61 onal-ZrO2) fibrous phase homogeneously distributed ir of these filaments were arranged in the steel die and ex continuous t-ZrO2 phase. The bending strength of the truded yielding the final second-pass green porous fila Al2O3 and 2-fold relative to ta a6-rold over that of ments. Figure I shows the schematic representation of having the same degree of porosity [13, 14]. The same ceramIcs he polymer binder was removed from the concept of tailoring the frame region of the porous body green body with a slow heating rate under a flowing can be extended to other types of microstructures such nitrogen atmosphere and then pore former carbon wa as laminated and fibrous laminated types. Using this burnt out at 1000C in an air atmosphere. Sintering method, the fibrous microstructure along with the pore of the samples was carried out in an air atmosphere at size and porosity can be controlled with great flexibility. 1450 and 1500C for 2 h. All the m-ZrO2 was trans- This method is attractive for incorporating various func formed to t-ZrO, after sintering. The density of the sin- tionalities by selection of microstructure and materials tered Al2O3 t-ZrO2)/t-ZrO2 porous bodies was and by innovative design measured using the weight dimension method. Bending The starting powders were AlO3(AKP-50, Sumim- strength was measured by a four-point bending test oto, Japan) with a particle size of 300 nm, monoclinic- The sintered bodies were used without any prior prepa rO2TZ-0Y, Tosoh, Japan), and yttria-stabilized ration, as the surface was smooth and the samples were ylindrical. The microstructure of the composites wa examined by scanning electron microscopy (SEM JALOr(s%em-ZrO)I JSM 6335F, JEOL, Tokyo, Japan)and transmission ●,●杰● electron microscopy (TEM, JEM2010, JEOL, Tokyo, Japan). Figure 2a shows the sintered first-pass porous body which is cylindrical with a circular pore inside. Prior to burn-out, the green body's outer diameter was 3.5 mm, which was reduced to 2.75 mm after sintering are/sDeⅡl Feed roll for 1 Green 1 pss Green 2 ps In the frame region the microstructure was fibrous, run ning unidirectionally through the composites, with a Figure 1. Schematic representation of the microstructure development. uniform distribution of Al,Ox(t-ZrO2) fiber he

and microporosity presents another concern requiring further investigation. A through-channel porosity with defined geometry can be seen as a compromise for this strength and porosity consideration. It is also highly desirable in case of applications associated with fluid flow or porosity where high interconnectivity is required. Channelled microporosity with improved strength can act as a holder or container of other kinds of mesoporous system and may find a range of applications. Recently, a multipass extrusion process has been investigated with respect to the fabrication of a porous body [13–15]. Ordered microchanneled porous bodies with monolithic ceramics were fabricated. However, their mechanical properties were significantly lower compared to those of the dense body, although compa￾rable or higher compared to those yielded by conven￾tional porous composite fabrication processes such as template [4], foaming [5] and leaching methods [6]. For all porous materials, monolithic ceramics or composites with random distributions of each constituent phase have been investigated thus far. Microstructural modifi- cation in the frame region of the macroporous body has not been investigated previously. It is very difficult to tailor the microstructure in the frame using existing fab￾rication methods. The present study is thought to be the first attempt to tailor the microstructure of a pore frame, an approach that could dramatically improve the mechanical properties of Al2O3–ZrO2 porous composite systems. In this work, a fibrous monolithic process is used for the first time to tailor the frame region of a microchan￾neled macroporous body, yielding a highly ordered and unidirectional fibrous microstructure resembling a bam￾boo-like biomimetic microstructure (see Figure 1). The frame of the porous body was made with Al2O3–(tetrag￾onal-ZrO2) fibrous phase homogeneously distributed in continuous t-ZrO2 phase. The bending strength of the porous composites was improved 6-fold over that of Al2O3 and 2-fold relative to that of ZrO2 porous body having the same degree of porosity [13,14]. The same concept of tailoring the frame region of the porous body can be extended to other types of microstructures such as laminated and fibrous laminated types. Using this method, the fibrous microstructure along with the pore size and porosity can be controlled with great flexibility. This method is attractive for incorporating various func￾tionalities by selection of microstructure and materials and by innovative design. The starting powders were Al2O3 (AKP-50, Sumim￾oto, Japan) with a particle size of 300 nm, monoclinic￾ZrO2 (TZ-0Y, Tosoh, Japan), and yttria-stabilized t-ZrO2 (TZ-3Y, Tosoh, Japan) with a particle size of 80 nm. Carbon powder (<15 lm, Aldrich, USA) was used as a pore-forming agent. Al2O3 + 25% m-ZrO2 was ball milled in ethanol and dried. t-ZrO2, Al2O3– (m-ZrO2), and carbon powder were separately shear￾mixed with a polymer binder, ethylene vinyl acetate (EVA) (Elvax 250 and Elvax 210, Dupont, USA), in a heated blender (Shina Platec. Co., Suwon, South Korea) with stearic acid as a lubricant. In the first step of the experiment, a t-ZrO2 shell was made by warm pressing and the Al2O3 core was fabricated by extrusion to make a 30 mm diameter feed roll. The t-ZrO2/Al2O3 volume ratio was chosen to be 60/40. The feed roll was then ex￾truded in a cylindrical die at 120 C to obtain 3.5 mm diameter filaments. These were the first-pass core/shell filaments, and the core/shell arrangement of the filament accounts for the core/shell microstructure of the frame region in the final porous composites. Sixty-one fila￾ments were then reloaded in the same die and extruded to obtain second-pass filaments. These second-pass core/shell filaments are 3.5 mm in diameter and contain 61 core/shells in a close-packed arrangement. Polymer￾mixed carbon powder was extruded with the same die at 120 C to obtain 3.5 mm diameter filaments. To make a porous body, the carbon filaments and the second-pass core/shell Al2O3–(m-ZrO2)/t-ZrO2 fila￾ments were arranged in the previously used steel die. In the die, two outer layers were comprised of second￾pass core/shell filaments and the inner layers were filled with carbon filaments. The arrangement was compacted and extruded at 120 C at an extrusion rate of 8 mm min1 . The obtained 3.5 mm diameter filament contained an inner carbon core and a ceramic shell with characteristic core/shell arrangements. This is termed hereafter the first-pass green porous body. Again, 61 of these filaments were arranged in the steel die and ex￾truded, yielding the final second-pass green porous fila￾ments. Figure 1 shows the schematic representation of the microstructure development. To obtain porous ceramics, the polymer binder was removed from the green body with a slow heating rate under a flowing nitrogen atmosphere and then pore former carbon was burnt out at 1000 C in an air atmosphere. Sintering of the samples was carried out in an air atmosphere at 1450 and 1500 C for 2 h. All the m-ZrO2 was trans￾formed to t-ZrO2 after sintering. The density of the sin￾tered Al2O3–(t-ZrO2)/t-ZrO2 porous bodies was measured using the weight dimension method. Bending strength was measured by a four-point bending test. The sintered bodies were used without any prior prepa￾ration, as the surface was smooth and the samples were cylindrical. The microstructure of the composites was examined by scanning electron microscopy (SEM, JSM 6335F, JEOL, Tokyo, Japan) and transmission electron microscopy (TEM, JEM2010, JEOL, Tokyo, Japan). Figure 2a shows the sintered first-pass porous body, which is cylindrical with a circular pore inside. Prior to burn-out, the green body’s outer diameter was 3.5 mm, which was reduced to 2.75 mm after sintering. In the frame region the microstructure was fibrous, run￾ning unidirectionally through the composites, with a Figure 1. Schematic representation of the microstructure development. uniform distribution of Al2O3–(t-ZrO2) fiber in the B. T. Lee, S. K. Sarkar / Scripta Materialia 61 (2009) 686–689 687

B. T. Lee, S. K Sarkar / Scripta Materialia 61(2009)686-689 larger particle size(<15 um), which means there was a difference in flow behavior of the pore-forming core and ceramic bound shell phase of the hot thermoplastic feed role. The irregular interface of the Al,O,m-ZrO2)core and t-ZrO2 shell and that of the rough pore surface are 100m mainly due to these reasons. From Figure 2d, it is seen that the t-zrO grain size was around 400 nm, whereas that of Al,O3 was in the sub-micrometer range. The inclu sion of t-ZrO, phase in the core inhibits substantial grain growth of Al,O3. Figure 2e is a bright-field TEM image of the Al,Ox(m-ZrO2)core, where the white and black con- trasts are Al2O3 and Zro2 particles, respectively. They were homogeneously dispersed and densely packed. Even though Al2O3-m-ZrO2) was initially used in the core twin structures or strain field were not observed. In the Al2O3-(m-ZrO2)system, we previously observed a strong strain field near the m-Zro2 particle due to the tetrago nal-monoclinic phase transformation during the cooling time of the sintering process [16]. This also introduced Figure 2.(a) Cross-sectional SEM image of first-pass porous body. (b) Cross-sectional SEM image of the second-pass porous body. (c) microcracks near the Al O3 and m-ZrO interface However, in this system there were no microcracks or Enlarged image of the pore region. (d) Enlarged image of frame region. twin defects near the Al,O3/zrO2 interface. Figure 2f (e and f) TEM images of the frame region. shows the high-resolution(HR)TEM and electron dif- fraction pattern of the [11 1]zone axis, which was taken t-zro2 matrix. Along the radial direction, there were from Figure 2e. From the electron difraction pattern it around 18 unit cells with Al2O3t-ZrO2) fiber, as de- was confirmed that even though m-ZrO2 was used gned in the green stage during fabrication. Figure 2b initially, t-ZrO2 was obtained after sintering. A few edge shows a cross-sectional SEM image of a sintered sec- dislocations can be observed in the hrtEM image. The ond-pass porous body. The pores are almost circular arrow indicates the low-angle tilt boundary, and a in shape and equidistant. The geometry and distribution periodical interfacial dislocation can be observed in the of the pores are two important factors for the ultimate 1 Table I presents the characteristic dimensions of the mechanical performance of the porous body With irreg- ular shape and a non-uniform distribution, the mechan- sintered porous bodies. In the second-pass porous body ical strength degrades more rapidly with the extent of the pore diameter was around 175 um. The frame thick porosity due to the tortuous and skewed pore frame ness of the porous body was around 110 um. Between where stress concentration zones, flaws or crack initia- two pores there were roughly 36 Al2O3t-ZrO2) fibers tion zones readily arise. In the current processing meth- distributed in the t-ZrO2 matrix. In the second-pass por od, the geometries of the pores and frames were ous body the microstructure in the frame region was significantly improved. The uniform distribution and highly refined, as seen from the AlO3(t-ZrO2)core shape of the pore, and thus uniformity of the frame and t-ZrO2 shell. The microstructure was arbitrarily de- thickness without much irregular curvature (except for signed in the present form. It can be changed very easily the microlevel surface roughness), resulted in improved by changing only the feed role design Changes in com- mechanical strength of the composites. Figure 2c shows position in the Al2O3-t-ZrO2) core and t-ZrO2 shell a closer view of the pores. It can be seen that the pore is phase can be easily achieved by changing their composi- surrounded by a fine fibrous microstructure, as in the tion in the green stage. Fiber dimension and density per first-pass but with significantly reduced dimension. Fig- unit area can also be controlled precisely by controlling ure 2d shows an enlarged image of the frame region the dimension and number of dense Al2O3t-ZrO2)/t The Al,O,t-ZrO2) cores, appearing with dark ZrO, core/shell filaments in the feed role for the first contrast. had a diameter of around 3.5 whereas pass porous body he t-ZrO2 matrix thickness was in the range of 1-2 um ure 3 shows al section SEM images of Figure 3a is an enlarged view in the t-zro2 matrix. No bulk defects, such as shrinkage of the pore region Figure 3b is a SEM image of the pore cavities or large cracks or any other flaws, were observed surface. The pore surface was highly rough with a bimo- in the composites. The Al,m-ZrO2)cores were some- dal distribution of roughness. It was entirely covered what irregular in shape instead of the original circular shape due to the repeated extrusion of the green compos- ites(the core/shell experienced four iterations of the Table 1. Characteristic dimensions of the sintered composite extrusion process). The slight difference in the rheology of the green polymer-bound Al2O3(m-ZrO2)and First pass Second pass(um) t-ZrO2, due to different mixing conditions and particle Pore diameter 144 size of the ceramics contributed to minute differential Frame thickness flow behavior while undergoing a reduction ratio of Al,O3 fiber diameter 71: 1. The polymer-bound carbon has a comparatively t-ZrO, shell thickness

t-ZrO2 matrix. Along the radial direction, there were around 18 unit cells with Al2O3–(t-ZrO2) fiber, as de￾signed in the green stage during fabrication. Figure 2b shows a cross-sectional SEM image of a sintered sec￾ond-pass porous body. The pores are almost circular in shape and equidistant. The geometry and distribution of the pores are two important factors for the ultimate mechanical performance of the porous body. With irreg￾ular shape and a non-uniform distribution, the mechan￾ical strength degrades more rapidly with the extent of porosity due to the tortuous and skewed pore frame where stress concentration zones, flaws or crack initia￾tion zones readily arise. In the current processing meth￾od, the geometries of the pores and frames were significantly improved. The uniform distribution and shape of the pore, and thus uniformity of the frame thickness without much irregular curvature (except for the microlevel surface roughness), resulted in improved mechanical strength of the composites. Figure 2c shows a closer view of the pores. It can be seen that the pore is surrounded by a fine fibrous microstructure, as in the first-pass but with significantly reduced dimension. Fig￾ure 2d shows an enlarged image of the frame region. The Al2O3–(t-ZrO2) cores, appearing with dark contrast, had a diameter of around 3.5 lm, whereas the t-ZrO2 matrix thickness was in the range of 1–2 lm. The Al2O3–(m-ZrO2) cores were uniformly distributed in the t-ZrO2 matrix. No bulk defects, such as shrinkage cavities or large cracks or any other flaws, were observed in the composites. The Al2O3–(m-ZrO2) cores were some￾what irregular in shape instead of the original circular shape due to the repeated extrusion of the green compos￾ites (the core/shell experienced four iterations of the extrusion process). The slight difference in the rheology of the green polymer-bound Al2O3–(m-ZrO2) and t-ZrO2, due to different mixing conditions and particle size of the ceramics, contributed to minute differential flow behavior while undergoing a reduction ratio of 71:1. The polymer-bound carbon has a comparatively larger particle size (615 lm), which means there was a difference in flow behavior of the pore-forming core and ceramic bound shell phase of the hot thermoplastic feed role. The irregular interface of the Al2O3–(m-ZrO2) core and t-ZrO2 shell and that of the rough pore surface are mainly due to these reasons. From Figure 2d, it is seen that the t-ZrO2 grain size was around 400 nm, whereas that of Al2O3 was in the sub-micrometer range. The inclu￾sion of t-ZrO2 phase in the core inhibits substantial grain growth of Al2O3. Figure 2e is a bright-field TEM image of the Al2O3–(m-ZrO2) core, where the white and black con￾trasts are Al2O3 and ZrO2 particles, respectively. They were homogeneously dispersed and densely packed. Even though Al2O3–(m-ZrO2) was initially used in the core, twin structures or strain field were not observed. In the Al2O3–(m-ZrO2) system, we previously observed a strong strain field near the m-ZrO2 particle due to the tetrago￾nal–monoclinic phase transformation during the cooling time of the sintering process [16]. This also introduced microcracks near the Al2O3 and m-ZrO2 interface. However, in this system there were no microcracks or twin defects near the Al2O3/ZrO2 interface. Figure 2f shows the high-resolution (HR) TEM and electron dif￾fraction pattern of the [1 1 1] zone axis, which was taken from Figure 2e. From the electron diffraction pattern it was confirmed that even though m-ZrO2 was used initially, t-ZrO2 was obtained after sintering. A few edge dislocations can be observed in the HRTEM image. The arrow indicates the low-angle tilt boundary, and a periodical interfacial dislocation can be observed in the image. Table 1 presents the characteristic dimensions of the sintered porous bodies. In the second-pass porous body, the pore diameter was around 175 lm. The frame thick￾ness of the porous body was around 110 lm. Between two pores there were roughly 36 Al2O3–(t-ZrO2) fibers distributed in the t-ZrO2 matrix. In the second-pass por￾ous body the microstructure in the frame region was highly refined, as seen from the Al2O3–(t-ZrO2) core and t-ZrO2 shell. The microstructure was arbitrarily de￾signed in the present form. It can be changed very easily by changing only the feed role design. Changes in com￾position in the Al2O3–(t-ZrO2) core and t-ZrO2 shell phase can be easily achieved by changing their composi￾tion in the green stage. Fiber dimension and density per unit area can also be controlled precisely by controlling the dimension and number of dense Al2O3–(t-ZrO2)/t￾ZrO2 core/shell filaments in the feed role for the first￾pass porous body. Figure 3 shows longitudinal section SEM images of the sintered porous body. Figure 3a is an enlarged view of the pore region. Figure 3b is a SEM image of the pore surface. The pore surface was highly rough with a bimo￾dal distribution of roughness. It was entirely covered Table 1. Characteristic dimensions of the sintered composites. Pass no. First pass Second pass (lm) Pore diameter 1.44 mm 175 Frame thickness 600 lm 110 Al2O3 fiber diameter 26 lm 3.5 t-ZrO2 shell thickness 13 lm 1.7 Figure 2. (a) Cross-sectional SEM image of first-pass porous body. (b) Cross-sectional SEM image of the second-pass porous body. (c) Enlarged image of the pore region. (d) Enlarged image of frame region. (e and f) TEM images of the frame region. 688 B. T. Lee, S. K. Sarkar / Scripta Materialia 61 (2009) 686–689

B. T. Lee, S. K Sarkar/ Scripta Materialia 61(2009)686-689 Table 2. Material properties of the composite Sintering Relative temperature(°C) strength(MP 1450 63.77±0.0 479±24 1500 64.27±0.16 588±32 Table 2 shows the material properties of the po ody sintered at different temperatures. The bending strength of the porous ceramics was remarkably im- proved in the porous body and sintered at 1500C,it was 588 MPa. The bending strength value was also remarkably higher compared to that of monolithic por- Figure 3.(a)Longitudinal section SEM image of second-pass porou ous ceramics of Al]O3(90 MPa) and porous t-ZrO body. (b) SEM image of the pore surface. (c) SEM image of the fibro (270 MPa)made by a similar extrusion process and hav pore region.(d) Enlarged SEM image of frame region. ing the same value of porosity [13, 14]. In conclusion, fibrous frame porous Al,O3(t-ZrO2)/ t-ZrO2 composites were fabricated by a multipass extru- with t-zrO2 phase, as evident from the white contrast of sion process. The composite frame of the porous body the image. The thin layer of t-ZrO2 on the pore surface has a very fine and homogeneous distribution of was strongly attached to the matrix phase. This extreme AlO3 t-ZrO2) fiber in a t-ZrO2 matrix. The Al2O t roughness along with preferential pore size is favorable ZrO2)fibers were around 3.5 um and the t-zrO2 matrix for a range of applications. For biomaterials applica- had a dimension of 1-2 um between the Al,Ox(t-ZrO2) tions, the pores allow proliferation of bone cells and fibers. The pores were around 175 um in diameter. The the rough pore surface allows a microlevel mechanical pore was in circular channel form and the surface was terlocking that can enhance attachment of the bone rough. The relative density was around 64%, whereas cells. It should be much easier to integrate a functional the actual density was less then 3.5 g cm(see Table coating or incorporate another microsystem inside these 2). The bending strength of the porous bodies was im- ores. Figure 3c is a longitudinal section SEM image of proved significantly and the highest value was observed the frame region. Unidirectionally aligned continuous fi- to be 588.62 MPa when sintered at 1500C. The porous bers of Al, O,(t-ZrO2) can be clearly observed. The dis- body did not show any sign of bulk defects continuities of a few fibers seen in the image are simply overlap of the matrix phase, and immediately beneath []B.T. Lee, A Nishiyama, K. Hiraga, Mater. Trans. JIM 34 them continuous fibers can be found. Figure 3d is an en larged image from Figure 3c 2A. H D. Aza, J. Chevalier, G. Fantozzi, M. Schehl, R The interface between the Al,O3t-ZrO2)cores and t- Torrecillas, Biomaterials 23(3)(2002)937 ZrO, matrix in the pore frame was distinct except in some 3]H. Chen, J. Gu, J. Shi. Adv. Mater. 17(2005)2010 instances where the t-ZrO2 phase of the matrix and that of 4V. Biasini, M. Parasporo, A. Bellosi, Thin Solid Film 297 997)207 the Al,Oxt-ZrOo,) core were continuous due to signifi 5]K. Maca, P. Dobsak, A.R. Boccaccini, Ceram. Int. 27 cant grain growth during sintering. The ZrO2 phase was (2001)577 homogeneously dispersed in the Al_O3 phase. This not 6 M.H. Bocanegra-Bernal, D D.L. Torre, J. Mater. Sci. 37 only inhibited the grain growth of both phases but also 2002)4947 introduced microcracking in the grain boundary of [7J. Chevalier, Biomaterials 27(2006)535 AlO3 and t-ZrO2 phase, which can lead to improved frac [8]V. Gonzalez- Pena, C. Marquez-Alvarez, I. Diaz, M ture toughness [9]. This is because of the differences in the Grande, T. Blasco, J. Perez-Pariente, Micropor Mesop coefficient of thermal expansion and elasticity of these two Mater.80(2005)17 materials. During the furnace-cooling period after sinter- 9A.K. Gain, H.Y. Song. B.T. Lee, Scr. Mater. 54(12) ing, this mismatch creates a strain field near the grain [10]H. Sieber, Mater. Sci. Eng. A 412(1-2)(2005)43 boundary and microcracks appear. The mismatch also [l1]CR. Rambo, H. Sieber, Adv. Mater. 17(8)(2005)1088 imposes residual stress on the core and matrix phases, [12]H. Masuda, K. Fukuda, Science 268(1995)1466. the core being in compression and the matrix in tensio [3]BT. Lee, I.C. Kang, S.H. Cho, H.Y. Song, J. Am. Ceram.Soc.88(8)(2005)2262 crack tip zones and reducing the crack propagation en- [14]AK. Gain, B.T. Lee, Mater. Sci Eng. A 419(1-2)(2006) ergy [16]. It was reported that the difference in the degree of residual stress increases the crack deflection and hence [5]BT. Lee, K H. Kim, H.C. Youn, H.Y. Song, J. Am. the fracture toughness [17]. Miyazaki et al. proposed that Ceram.Soc.90(2)(2007)62 [16 B T Lee, K H Lee, K. Hiraga, Scr. Mater. 38(1998)1101 the efect of a crack deflection on the toughness is almost [17]T. Adachi, T. Sekino, T. Kusunose, T. Nakayama, A proportional to the variation in residual stress across the Hikasa, Y.H. Choa, K. Niihara, J. Ceram Soc. Jpn. Ill interface[18]. From these two findings it can be inferred (2003)4 that the fracture toughness can be effectively increased [18]H Miyazaki, Y. Yoshizawa, K. Hirao, J. Eur. Ceram in this system Soc.26(16)(2006)3539

with t-ZrO2 phase, as evident from the white contrast of the image. The thin layer of t-ZrO2 on the pore surface was strongly attached to the matrix phase. This extreme roughness along with preferential pore size is favorable for a range of applications. For biomaterials applica￾tions, the pores allow proliferation of bone cells and the rough pore surface allows a microlevel mechanical interlocking that can enhance attachment of the bone cells. It should be much easier to integrate a functional coating or incorporate another microsystem inside these pores. Figure 3c is a longitudinal section SEM image of the frame region. Unidirectionally aligned continuous fi- bers of Al2O3–(t-ZrO2) can be clearly observed. The dis￾continuities of a few fibers seen in the image are simply overlap of the matrix phase, and immediately beneath them continuous fibers can be found. Figure 3d is an en￾larged image from Figure 3c. The interface between the Al2O3–(t-ZrO2) cores and t￾ZrO2 matrix in the pore frame was distinct except in some instances where the t-ZrO2 phase of the matrix and that of the Al2O3–(t-ZrO2) core were continuous due to signifi- cant grain growth during sintering. The ZrO2 phase was homogeneously dispersed in the Al2O3 phase. This not only inhibited the grain growth of both phases but also introduced microcracking in the grain boundary of Al2O3 and t-ZrO2 phase, which can lead to improved frac￾ture toughness [9]. This is because of the differences in the coefficient of thermal expansion and elasticity of these two materials. During the furnace-cooling period after sinter￾ing, this mismatch creates a strain field near the grain boundary and microcracks appear. The mismatch also imposes residual stress on the core and matrix phases, the core being in compression and the matrix in tension. The microcracks play an important role in deflecting the crack tip zones and reducing the crack propagation en￾ergy [16]. It was reported that the difference in the degree of residual stress increases the crack deflection and hence the fracture toughness [17]. Miyazaki et al. proposed that the effect of a crack deflection on the toughness is almost proportional to the variation in residual stress across the interface [18]. From these two findings it can be inferred that the fracture toughness can be effectively increased in this system. Table 2 shows the material properties of the porous body sintered at different temperatures. The bending strength of the porous ceramics was remarkably im￾proved in the porous body and sintered at 1500 C, it was 588 MPa. The bending strength value was also remarkably higher compared to that of monolithic por￾ous ceramics of Al2O3 (90 MPa) and porous t-ZrO2 (270 MPa) made by a similar extrusion process and hav￾ing the same value of porosity [13,14]. In conclusion, fibrous frame porous Al2O3–(t-ZrO2)/ t-ZrO2 composites were fabricated by a multipass extru￾sion process. The composite frame of the porous body has a very fine and homogeneous distribution of Al2O3–(t-ZrO2) fiber in a t-ZrO2 matrix. The Al2O3–(t￾ZrO2) fibers were around 3.5 lm and the t-ZrO2 matrix had a dimension of 1–2 lm between the Al2O3–(t-ZrO2) fibers. The pores were around 175 lm in diameter. The pore was in circular channel form and the surface was rough. The relative density was around 64%, whereas the actual density was less then 3.5 g cm1 (see Table 2). The bending strength of the porous bodies was im￾proved significantly and the highest value was observed to be 588.62 MPa when sintered at 1500 C. The porous body did not show any sign of bulk defects. [1] B.T. Lee, A. Nishiyama, K. Hiraga, Mater. Trans. JIM 34 (88) (1993) 682. [2] A.H.D. Aza, J. Chevalier, G. Fantozzi, M. Schehl, R. Torrecillas, Biomaterials 23 (3) (2002) 937. [3] H. Chen, J. Gu, J. Shi. Adv. Mater. 17 (2005) 2010. [4] V. Biasini, M. Parasporo, A. Bellosi, Thin Solid Film 297 (1997) 207. [5] K. Maca, P. Dobsak, A.R. Boccaccini, Ceram. Int. 27 (2001) 577. [6] M.H. Bocanegra-Bernal, D.D.L. Torre, J. Mater. Sci. 37 (2002) 4947. [7] J. Chevalier, Biomaterials 27 (2006) 535. [8] V. Gonza´lez-Pen˜a, C. Ma´rquez-Alvarez, I. Dı´az, M. Grande, T. Blasco, J. Pe´rez-Pariente, Micropor. Mesopor. Mater. 80 (2005) 173. [9] A.K. Gain, H.Y. Song, B.T. Lee, Scr. Mater. 54 (12) (2006) 2081. [10] H. Sieber, Mater. Sci. Eng. A 412 (1–2) (2005) 43. [11] C.R. Rambo, H. Sieber, Adv. Mater. 17 (8) (2005) 1088. [12] H. Masuda, K. Fukuda, Science 268 (1995) 1466. [13] B.T. Lee, I.C. Kang, S.H. Cho, H.Y. Song, J. Am. Ceram. Soc. 88 (8) (2005) 2262. [14] A.K. Gain, B.T. Lee, Mater. Sci. Eng. A 419 (1–2) (2006) 269. [15] B.T. Lee, K.H. Kim, H.C. Youn, H.Y. Song, J. Am. Ceram. Soc. 90 (2) (2007) 629. [16] B.T. Lee, K.H. Lee, K. Hiraga, Scr. Mater. 38 (1998) 1101. [17] T. Adachi, T. Sekino, T. Kusunose, T. Nakayama, A. Hikasa, Y.H. Choa, K. Niihara, J. Ceram. Soc. Jpn. 111 (2003) 4. [18] H. Miyazaki, Y. Yoshizawa, K. Hirao, J. Eur. Ceram. Soc. 26 (16) (2006) 3539. Table 2. Material properties of the composite. Sintering temperature (C) Relative density (%) Bending strength (MPa) 1450 63.77 ± 0.05 479 ± 24 1500 64.27 ± 0.16 588 ± 32 Figure 3. (a) Longitudinal section SEM image of second-pass porous body. (b) SEM image of the pore surface. (c) SEM image of the fibrous pore region. (d) Enlarged SEM image of frame region. B. T. Lee, S. K. Sarkar / Scripta Materialia 61 (2009) 686–689 689