L-C. Kang et al./ Materials Letters 59(2005)69-73 71 (b) Fig. 2. SEM micrographs of the porous Al2O, body after buming out and sintering processes. a-second pass, b-third pass, and arged image of the forms a uniform distribution of carbon in the Al,O3 matrix Fig. 4(a) shows the spindle-shaped human osteoblast (Fig. I(d)). Although the rod shape of carbon particle is not like MG-63 cells cultured on 6-well plate. The MG-63 observed due to sectioning problem, the diameter calculated cells were originated from a human osteosarcoma and have considering the extrusion ratio is approximately 50 um. To osteoblast characteristics. The osteoblasts are known as sum up, the Fig. I indicates that the ceramic microstructure bone-forming cells of developing and mature bone. The aving more than two phases can be easily controlled using top surface of second passed fibrous monolithic Al2O3 the fibrous monolithic process porous bodies were fully covered with osteoblasts at 14 Fig 2 shows SEM microstructure of porous Al2O3 body ys(Fig. 4(b). after burning out the EVA under N2 atmosphere followed by MG-63 cells were attached to top, inside, and bottom sintering in air. It is seen that the microstructure of sintered surface of Al2O3 porous body and grew well(Fig. 5). The body presents the same pattern with the one as extruded in cells grown on top surface of the porous body showed Fig. 1. The carbon/EVA (rod parts in Fig. 1) was fully highly condensed circular growth around the margin of the removed by the burning out process, and then the pores porous body(Fig. 5(a)). The high roughness formed along were formed. The pores produced in the second passed and inside the continuous pores [15] probably contributed to sintered body are about 200 um in diameter and ar promote the attachment, spreading, and growth of cells omogenously embedded in the Al2O3 matrix(Fig. 2(a)). Surface roughness dependence of the cells was well Similar to the as-extruded microstructure in Fig. 1(c), the reported by Boyan et al. [16]. The cells grown on bottom microstructure in the third passed and sintered body is surface of the porous body showed epithelial-like morpho- impossible to distinguish under the same low magnification, logic changes and grew faster than that on the top surface shown in Fig. 2(b). But the enlarged microstructure(Fig.(Fig. 5(c). The cells grew up through inside porous bodies 2(c))shows that approximately 40-um pores were formed in The pore size of third passed Al2O3 porous ody was the third filament. The reduction in the size after sintering however, too small to make the cells grow. Thus, it is not generally occurred In order to investigate a variation of composition or hase during the whole fibrous monolithic process, XRD characterization was conducted, as shown in Fig 3(a-f. It is not strange to detect Al2O3(Fig 3(a)) and carbon(Fig 3(b)) from the raw Al2O3 powder and carbon powder, f)ALOvC-sintered respectively. Fig 3(c)shows Al_O3/EVA mixture consists of ALO, C-2nd BI only Al2O3 due to the amorphous phase of EVA. However, Al2O3/EVA-carbon/EVA composite filament presents peaks for Al2O3 and carbon after the first binder burning out (Ist d) Al,O.C-Ist BBo BO), whereas the carbon was fully removed after the second binder burning out process(2nd BBO), as shown in Fig. 3(d) and (e), respectively. The binder burning out process is usually divided into two thermal steps. The first step(lst BBO)is for a transformation of EVa to carbon a)ALyO, powder under N2 atmosphere, while the second is to remove the carbons which were initially combined as a pore-forming agent as well as transformed from the eva during the first BO [15. The fibrous monolithic Al2O3/EVA-carbon/EVA composite finally is composed of only Al2O3 after sintering, Fig. 3. XRD profile of Al2O3-C composites: raw Al2O, powder(a),raw carbon powder(b), Al2O3/poly osite(c), AlO,/polymer compo- thus the porous Al O3 sintered body having the purity can site after first binder burning out(d), Al,/polymer composite after second be used as a bioinert implant binder burning out(e), and the sintered body (f).forms a uniform distribution of carbon in the Al2O3 matrix (Fig. 1(d)). Although the rod shape of carbon particle is not observed due to sectioning problem, the diameter calculated considering the extrusion ratio is approximately 50 Am. To sum up, the Fig. 1 indicates that the ceramic microstructure having more than two phases can be easily controlled using the fibrous monolithic process. Fig. 2 shows SEM microstructure of porous Al2O3 body after burning out the EVA under N2 atmosphere followed by sintering in air. It is seen that the microstructure of sintered body presents the same pattern with the one as extruded in Fig. 1. The carbon/EVA (rod parts in Fig. 1) was fully removed by the burning out process, and then the pores were formed. The pores produced in the second passed and sintered body are about 200 Am in diameter and are homogenously embedded in the Al2O3 matrix (Fig. 2(a)). Similar to the as-extruded microstructure in Fig. 1(c), the microstructure in the third passed and sintered body is impossible to distinguish under the same low magnification, as shown in Fig. 2(b). But the enlarged microstructure (Fig. 2(c)) shows that approximately 40-Am pores were formed in the third filament. The reduction in the size after sintering generally occurred. In order to investigate a variation of composition or phase during the whole fibrous monolithic process, XRD characterization was conducted, as shown in Fig. 3(a)–(f). It is not strange to detect Al2O3 (Fig. 3(a)) and carbon (Fig. 3(b)) from the raw Al2O3 powder and carbon powder, respectively. Fig. 3(c) shows Al2O3/EVA mixture consists of only Al2O3 due to the amorphous phase of EVA. However, Al2O3/EVA–carbon/EVA composite filament presents peaks for Al2O3 and carbon after the first binder burning out (1st BBO), whereas the carbon was fully removed after the second binder burning out process (2nd BBO), as shown in Fig. 3(d) and (e), respectively. The binder burning out process is usually divided into two thermal steps. The first step (1st BBO) is for a transformation of EVA to carbon under N2 atmosphere, while the second is to remove the carbons which were initially combined as a pore-forming agent as well as transformed from the EVA during the first BBO [15]. The fibrous monolithic Al2O3/EVA–carbon/EVA composite finally is composed of only Al2O3 after sintering, thus the porous Al2O3 sintered body having the purity can be used as a bioinert implant. Fig. 4(a) shows the spindle-shaped human osteoblast￾like MG-63 cells cultured on 6-well plate. The MG-63 cells were originated from a human osteosarcoma and have osteoblast characteristics. The osteoblasts are known as bone-forming cells of developing and mature bone. The top surface of second passed fibrous monolithic Al2O3 porous bodies were fully covered with osteoblasts at 14 days (Fig. 4(b)). MG-63 cells were attached to top, inside, and bottom surface of Al2O3 porous body and grew well (Fig. 5). The cells grown on top surface of the porous body showed highly condensed circular growth around the margin of the porous body (Fig. 5(a)). The high roughness formed along inside the continuous pores [15] probably contributed to promote the attachment, spreading, and growth of cells. Surface roughness dependence of the cells was well reported by Boyan et al. [16]. The cells grown on bottom surface of the porous body showed epithelial-like morpho￾logic changes and grew faster than that on the top surface (Fig. 5(c)). The cells grew up through inside porous bodies. The pore size of third passed Al2O3 porous body was, however, too small to make the cells grow. Thus, it is not Fig. 2. SEM micrographs of the porous Al2O3 body after burning out and sintering processes. a—second pass, b—third pass, and c—enlarged image of the third pass. Fig. 3. XRD profile of Al2O3-C composites: raw Al2O3 powder (a), raw carbon powder (b), Al2O3/polymer composite (c), Al2O3/polymer compo￾site after first binder burning out (d), Al2O3/polymer composite after second binder burning out (e), and the sintered body (f). I.-C. Kang et al. / Materials Letters 59 (2005) 69–73 71