Biomaterials 24 (2003) 31153123

Scaffold development using selective laser sintering of polyetheretherketonehydroxyapatite biocomposite blends

K.H. Tana, C.K. Chuaa,*,1, K.F. Leonga, C.M. Cheaha, P. Cheangb, M.S. Abu Bakarb, S.W. Chac
a

School of Mechanical and Production Engineering, Nanyang Technological University, Singapore 639798, Singapore b School of Materials Engineering, Nanyang Technological University, Singapore 639798, Singapore c National Institute of Education, Nanyang Technological University, Singapore 639798, Singapore Received 27 November 2002; accepted 20 February 2003

Abstract In tissue engineering (TE), temporary three-dimensional scaffolds are essential to guide cell proliferation and to maintain native phenotypes in regenerating biologic tissues or organs. To create the scaffolds, rapid prototyping (RP) techniques are emerging as fabrication techniques of choice as they are capable of overcoming many of the limitations encountered with conventional manualbased fabrication processes. In this research, RP fabrication of solvent free porous polymeric and composite scaffolds was investigated. Biomaterials such as polyetheretherketone (PEEK) and hydroxyapatite (HA) were experimentally processed on a commercial selective laser sintering (SLS) RP system. The SLS technique is highly advantageous as it provides good user control over the microstructures of created scaffolds by adjusting the SLS process parameters. Different weight percentage (wt%) compositions of physically mixed PEEK/HA powder blends were sintered to assess their suitability for SLS processing. Microstructural assessments of the scaffolds were conducted using electron microscopy. The results ascertained the potential of SLSfabricated TE scaffolds. r 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Tissue engineering; Scaffolds; Polyetheretherketone; Hydroxyapatite; Selective laser sintering

1. Introduction Scaffold-guided tissue engineering (TE) has been developed to regenerate specic and functional human tissues or organs [14]. As the scaffolds form the platform for cells to develop and to be organised into tissues and organs. TE scaffolds should facilitate the colonisation of cells and possess properties and characteristics that enhance cell attachment, proliferation, migration and expression of native phenotypes. Scaffold characteristics and properties such as porosity, surface area to volume ratio, pore size, pore interconnectivity, structural strength, shape (or overall geometry) and biocompatibility [59] are often considered to be critical
*Corresponding author. Tel.: +65-6790-4897; fax: +65-6795-7329. E-mail address: mckchua@ntu.edu.sg (C.K. Chua). 1 Present address: 50 Nanyang Avenue, School of Mechanical and Production Engineering, Nanyang Technological University, Singapore 639798, Singapore.

factors in their design and fabrication. The limitations and difculties encountered with conventional fabrication techniques for producing scaffolds with pores that are appropriately and consistently sized and thoroughly interconnected led to the research for alternative fabrication methods that are capable of providing the user with some control over the formation of the scaffolds internal microstructure. In order to facilitate the reorganisation of cells through the secretion of the cells native matrix, reproducible biocompatible scaffolds with unique geometries are needed. Many natural, e.g., collagen and chitin, and synthetic biomaterials, e.g., poly(a-hydroxyesters) and poly(anhydrides), [1012], have been widely and successfully used as scaffolding materials because of their good cell-tissue biocompatibility and processability [13]. However despite the suitability of the scaffolding materials [14], scaffolds produced to date are far from ideal due to the limitations encounter with the fabrication techniques [15,16].

0142-9612/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0142-9612(03)00131-5

3116

K.H. Tan et al. / Biomaterials 24 (2003) 31153123

1.1. Applications of RP techniques for tissue engineering Rapid prototyping (RP) is a key group of prototyping technologies with the ability to rapidly fabricate complex three-dimensional (3-D) physical structures [17]. RP rely on the use of model data created in a computer aided design (CAD) solid modelling environment to generate 3-D physical objects. RP systems process CAD data by mathematically slicing the computer model of the nal desired object into thin even layers. The RP systems then utilise the slice data to replicate or reconstruct a physical object through layerby-layer manufacturing whereby solid material layers are fabricated, stacked vertically and bonded to the previous layer to give rise to the physical object. In scaffold production, conventional fabrication techniques such as ber bonding, gas foaming, emulsion freeze drying, melt moulding, membrane lamination, solvent casting, etc. [13,1821] have been developed and used for producing TE scaffolds which have been tested and applied with varying degree of success. Despite the wide variety of techniques available, conventionally produced scaffolds lack consistency and reproducibility in structural and mechanical properties that are generally required in TE applications. In addition, the extensive use of organic solvents with most of these techniques are not desirable as residual traces left behind after processing can result in adverse toxic effects in vitro and elicit inammatory responses in vivo [13]. Due to such limitations, the potential in using RP is tremendous as most RP techniques are able to fabricate intricate 3D reproducible structures without the use of organic solvent. RP techniques that have been explored or being developed for scaffold fabrication can be classied under three different categories, namely, solid

[22,23], liquid [24,25] or powder-based [26,27] techniques depending on the physical form of the material stocks or building materials utilised in these processes. The research presented in this paper explored the viability of using a powder-based RP technique, Selective Laser Sintering (SLS) for fabricating polymeric and composite TE scaffolds with desirable macro- and micro-structural characteristics. 1.2. Selective laser sintering The SLS technique employs a carbon dioxide laser beam to sinter thin layers of powdered polymeric materials to form solid three-dimensional objects. The object is built layer-by-layer from CAD data les in the industry-standard (STL le format). During SLS fabrication, the laser beam is selectively scanned over the powder surface following the cross-sectional proles carried by the slice data. The interaction of the laser beam with the powder raises the powder temperature to the point of melting and causes the particles to be fused together to form a solid mass. Subsequent layers are built directly on top of previously sintered layers with new layers of powder being deposited via a roller on top of the previously sintered layer. Fig. 1 shows the process chain of the SLS technique. As commercially available SLS modelling materials used by SLS are non biocompatible in nature, SLS application in TE scaffold production remains limited. Lee et al. [28,29] looked into the development of bioceramic scaffolds that can aid the regeneration of hard tissues for recovery of bone defects and injuries via laser sintering of hydroxyapatite (HA) powders that were coated with a secondary polymeric binder, poly(methylmethacrylate) (PMMA). A slurry comprising

Fig. 1. Schematic layout of the SLS process.

K.H. Tan et al. / Biomaterials 24 (2003) 31153123

3117

of predetermined mixing ratios of ceramic particles and PMMA was sprayed dried to obtain the PMMA-coated HA powders. In addition, diluted methanol was used in the fabrication process. Since the use of organic solvents is highly undesirable for the processing of TE scaffolds as mentioned, the research work presented in this paper circumvents the use of solvents by utilising pure biopolymer powders (polyetheretherketone, PEEK) and physically blended mixtures of PEEK and HA powders. The feasibility of sintering such powder blends and the inuence of SLS process parameters on the sintering quality and resulting microstructure of the sintered specimens were studied.

powders to achieve powder blends having 10, 20, 30 and 40 wt% HA contents. As PEEK is bioinert, the addition of HA particles will increase the bioactivity of the overall implant. The prepared powder blends were processed on a commercial SLS system, Sinterstation 2500 (3D Systems Inc., Valencia, CA) [33], to produce test specimens. Except for changing the process parameters on the operating software of the SLS system, no modications were made on the SLS system for processing the new materials. Since PEEK has a much lower melting point compared to HA, it is possible to induce sintering of PEEK at temperatures near Tg and to bind and partly expose the HA particles within the sintered PEEK matrix. 2.3. Design and fabrication of test specimens The test specimens fabricated on the SLS were designed as a circular disc with a diameter and thickness of 12.0 and 0.5 mm, respectively, for microscopic examination and biocompatibility assessments. The discs were designed so as to t snugly into the wells of a standard 24-well plate. The CAD model of the test specimen was generated using a standard CAD software, ProENGINEER Ver. 2000i (Parametric Technology Corp., Needham, MA) and exported in the STL le format for uploading into the SLS system. Since the quality and the degree of sintering of the build are dependent on the SLS process parameters, a study was conducted to determine the optimal SLS processing parameters for PEEK. The inuence of three main SLS process parameters on the degree of sintering achieve with PEEK were investigated, namely, laser power, part bed temperature and scan speed. In the research conducted by Leong et al. [34] and Nelson [35] on the laser-sintering of polymer powders, these three mentioned parameters were ascertained to be mainly responsible for the amount of laser irradiation per unit area received by the powders (i.e., energy density or Andrews Number [35]) during fabrication on a SLS system. Since the energy density affects the degree of sintering encountered by the exposed powders, specimens with different porosity can be obtained by varying SLS process parameters. The settings for the SLS process parameters were kept at their default values except for laser power and part bed temperature, which was set at 8 W and 110 C, respectively. Pure PEEK was then subjected to laser-sintering at a different laser power and temperature settings to determine the optimal settings for the sintering of PEEK. By varying the 2 process parameters, ve groups of specimens comprising of different blends of PEEK and HA were fabricated. The range of laser power and the part bed temperature settings for fabricating the various specimens are tabulated in Table 1.

2. Materials and methods 2.1. Polyetheretherketone (PEEK) and hydroxyapatite (HA) powders PEEK exists as a semi-crystalline polymer at room temperature and possess pairs of ether linkages in its chain backbone. Its relatively high melting point, TmE343 C, and glass transition temperature, TgE143 C, makes it a suitable and stable polymer to be processed at high temperatures. Its mechanical properties remain stable at high temperature of about 200 C for prolonged periods of time [30]. PEEK exhibits excellent chemical resistance and is almost insoluble in most solvents. An added advantage of PEEK in biomedical application is that being radio opaque, the implanted scaffolds allow for further investigation by means of X-ray. Furthermore, it can be readily sterilised in a steam autoclave or by radiation without encountering any signicant degradation in its inherent properties. PEEK powder with a specied average particle size of 25 mm marketed under the brand name PeeKTM 150XF (Victrex Plc, Lancashire, UK) was used for the purpose of this research. PeeKTM 150XF is a low viscosity grade polymer that is used mainly for powder coating and as llers. The cytotoxicity of PEEK has been reported elsewhere [31,32]. HA powders used in this research are sold under the brand name CAMCERAM II HA (Cam Implants B.V., Netherlands). The HA powders meet the ASTM F 1185-88 requirements and have a particle size distribution with at least 90 wt% below 60 mm, as determined by Coulter Counter analysis. The average material density is specied as 3.05 g/cm3. 2.2. Physical blending of PEEK and HA powders Mixtures of PEEK/HA powders were produced by physically blending pure PEEK and HA powders in different weight percentages using a roller-mixer. In producing the blends, PEEK is the base material and HA powders were added and dispersed into the PEEK

3118

K.H. Tan et al. / Biomaterials 24 (2003) 31153123

Table 1 Specimens groups for pure PEEK and PEEK/HA polymer blends Laser Power (W) Temperature ( C) 110 9 10 12 14 16 HAPEEK HAPEEK HAPEEK HAPEEK 0100 0100 0100 0100 140 HAPEEK 0100 HAPEEK 0100 HAPEEK 0100, HAPEEK 1090, HAPEEK 2080, HAPEEK 3070, HAPEEK 4060 HAPEEK 0100 HAPEEK 0100 HAPEEK 0100 0100Pure PEEK 109010 wt% HA-90 wt% 208020 wt% HA-80 wt% 307030 wt% HA-70 wt% 406040 wt% HA-60 wt% PEEK PEEK PEEK PEEK

With the Tg value determined, the part bed temperature (which is the temperature by which the powders are preheated to before laser sintering) was set at an initial trial setting of 110 C to prevent unwanted coagulation and hardening of the powders. 3.2. Microscopic examination of powder stocks Microscopic examinations were carried out on the individual powder materials prior to blending, after blending and on the polymer blend produced by lasersintering. The results of the examinations are presented in the following sections. 3.2.1. HA and PEEK powder before blending Figs. 2a and b are micrographs taken for the asreceived PEEK and HA powders, respectively, prior to blending. As observed in these gures, the PEEK powders are irregular in shape compared to the spherical HA powders. This distinct morphological features between the two materials make the identication of HA particles in the sintered PEEK matrix much simpler.

18 20 22 24 26 28 HAPEEK HAPEEK HAPEEK HAPEEK HAPEEK

2.4. Thermal analysis of PEEK To assist in the proper selection of part bed temperature, the Tg and Tm values of the PEEK powder stock have to be determined. Differential scanning calorimetry (DSC) performed on a PerkinElmer Inc., DSC-7, thermal analyser was employed for the measurements. Repeated samples of PEEK powder (average weight=5.5 mg) were scanned from 50 C to 350 C at a linear ramp rate of 10 C/min, using nitrogen as a purge gas. 2.5. Microscopic examination Particle size analysis for both the PEEK and HA powders and microstructural characterisation of the SLS-fabricated specimens were carried out using a scanning electron microscope (SEM), JEOL JSM-5600 LV, to characterise the individual powders and to analyse the surface morphology and microstructure of the sintered specimens. All SLS-fabricated specimens were examined under high vacuum conditions at 12 kV and magnications of 500 times.

3. Results and discussion 3.1. Thermal properties of PEEK DSC results for PEEK indicated an averaged Tg and Tm of 143 C and 342 C, respectively, which were very close to the gures provided by the technical specications for the PeeKTM 150XF material (Section 2.1).

Fig. 2. Micrographs taken for the as-received (a) PEEK and (b) HA powders.

K.H. Tan et al. / Biomaterials 24 (2003) 31153123

3119

3.2.2. HA/PEEK powder blends before laser sintering Figs. 3ad presents the micrographs taken for the four different powder blends containing 1040 wt% HA, respectively, prior to laser sintering. In Figs. 3ad, HA particles can be observed in all the four different powder blends in increasing amounts with increased ratio of HA content. It is worthwhile to note that irregardless of the different quantities of powder used as samples for analysis in SEM, the HA particles were observed on all samples indicating that mixtures with good dispersion and distribution of HA were obtained. 3.2.3. Sintering of pure PEEK Before subjecting the powder blends to laser sintering, some preliminary sintering tests were carried out on pure PEEK powders to determine the range of suitable processing parameters to be used on the SLS system. Different specimens were produced at various parameter settings for part bed temperature and laser power. Instead of building a disc specimen with a thickness of 0.5 mm as described earlier, only one layer of material (0.l mm thick) was sintered to test the parameter settings. For all the experiments conducted, the scan

speed of the laser system was kept at the default value of 5080 mm/s (200 in/s). Figs. 4ad presents the micrographs taken of the microstructure of pure PEEK specimens that sintered at a part bed temperature setting of 110 C and at different laser power settings of 1016 W. Attempts to sinter the PEEK powder at a part bed temperature of 110 C and laser power settings below 10 W were unsuccessful as the irradiated laser energy was too low to result in the proper sintering of PEEK powder. As observed in the micrographs in Fig. 4, the formation of necks between powder particles becomes prominent in specimens fabricated at laser power settings of at least 12 W. However, due to the relatively low part bed temperature used, delamination was observed between the layers on the sintered PEEK specimens. Thus, in order to obtain better structural integrity, the use of a much higher part bed temperature setting was decided upon. For subsequent sintering experiments, the part bed temperature was raised to 130 C and the same set of values for the laser power settings were repeated. The results observed for specimens fabricated with the new set of parameters were similar to those obtained for the

Fig. 3. Micrograph of PEEK/HA powder blend before sintering for different weight composition; (a) 10 wt% HA, (b) 20 wt% HA, (c) 30 wt% HA, and (d) 40 wt% HA.

3120

K.H. Tan et al. / Biomaterials 24 (2003) 31153123

Fig. 4. Micrographs of pure PEEK specimens at a part bed temperature of 110 C and laser power settings of (a) 10 W, (b) 12 W, (c) 14 W, and (d)16 W.

previous specimens that were fabricated at a part bed temperature of 110 C. The part bed temperature was subsequently raised to 140 C, which is 3 C below the measured Tg for PEEK. A higher range of laser power settings from 9 W to 28 W was also employed. Micrographs taken for the different specimens produced from the new set of parameters are presented in Figs. 5af. Although specimens can be produced at a laser power setting of 9 W using the higher part bed temperature of 140 C, the specimens obtained were fragile. Observations made from the micrograph taken for a sample specimen (Fig. 5a) showed poor quality of sintering with little necking observed between the particles. However comparing the micrographs in Figs. 4 and 5, it could be deduced that the sintering of PEEK is more successful at higher part bed temperatures for each of the different laser power settings tested as evident from the quantity and prominence of neck formation between particles. When the laser power was increased to 12 W and higher at a part bed temperature setting of l40 C, the sintering results obtained appeared more promising as observed from Figs. 5cf. The degree of necking was more evident in specimens fabricated at part bed temperature of 140 C when compared to 110 C. However, delamination (edges opening up) was noted near the edge for specimens built using laser power settings lower than 16 W. Fig. 6 is a micrograph of a pure PEEK specimen built using 15 W at 140 C showing

signs of delamination. As delamination occurs when adjacent layers of material are not properly bonded to one another, the laser power was further increased gradually till 28 W to eliminate the problem. It was observed that the degree of melting of PEEK was more evident at higher laser power. Specimens produced at laser power settings of 28 W appeared charred and may have degraded and are thus not acceptable. As illustrated in Fig. 5e, the use of laser power settings of 22 W and above would result in the formation of microstructures with highly dense morphology. However, to allow the in growth of seeded cell, it is of interest to the authors to create porous specimens. As such, the most favourable SLS process parameter settings for building PEEK scaffolds should be 140 C for the part bed temperature and between 16 W to 21 W for the laser power setting.

3.2.4. HA/PEEK composite obtained from laser sintering Using the parameters determined previously, thin discs of the polymer blend comprising of PEEK and HA powders were laser-sintered for biocompatibility assessments and to be used as a comparison with pure PEEK. All specimens were fabricated at a part bed temperature and laser power setting of 140 C and 16 W, respectively. Promising results were obtained as for the sintering experiments as illustrated by Figs. 7ad, in which HA particles are circled. As noted from these gures, the HA

K.H. Tan et al. / Biomaterials 24 (2003) 31153123

3121

Fig. 5. Micrograph of sintered pure PEEK specimens produced at a part bed temperature of 140 C and laser power settings of (a) 9 W, (b) 12 W, (c) 16 W, (d) 20 W, (e) 24 W, and (f) 28 W.

Fig. 6. Micrograph of sintered pure PEEK taken at the edge showing delamination.

particles can be seen embedded and partially exposed in the PEEK matrix. The successful sintering of the test specimens indicates the potential of producing PEEK scaffolds on the SLS. In addition, the successful incorporation of HA into the polymer matrix will enhance the bioactivity of the specimens. However, it is noted that with a decrease in the percentage of PEEK in the composition of the powder blend, the fabricated disc specimens were fragile to handle and this fragility made it not practical to use in laser-sintering. In order to produce a structure with good integrity, it was proposed that the composition of HA in the mixture should be kept at 40 wt% HA and hence no further increase in the composition of HA in the PEEK/HA mixture was carried out.

3122

K.H. Tan et al. / Biomaterials 24 (2003) 31153123

Fig. 7. Micrograph of sintered polymer blend with; (a) 10 wt% HA, (b) 20 wt% HA, (c) 30 wt%, and (d) 40 wt% HA.

4. Conclusion The advantages of using SLS for TE scaffolds lie in the ability to control pore structure for biogenesis through the control of polymer content and the ability to construct complex three-dimensional structure for TE applications. Due to its ability to process various materials, the potential of using SLS is further enhanced as unconventional materials such as PEEK, which was used in this project, was tested to establish the feasibility of its usage for TE scaffolds. Three main parameters of SLS, namely the part bed temperature, laser power and scan speed were investigated to study its effect on the integrity of the test specimens fabricated for the purpose of biocompatibility of the materials used. With the scan speed kept constant and varying part bed temperature and laser power, several circular specimens measuring 12 mm in diameter and 0.5 mm thick were fabricated. The results obtained showed that a low part bed temperature should be complemented by a higher laser power. The research has also shown promising result of being able to laser sinter a high melting point polymer in a much lower temperature environment. Furthermore, the ability to incorporate different amounts of a

bioactive material, hydroxyapatite, into the polymer blend reiterated its viability for use in TE scaffolds, especially bone scaffolds as apatite is one of the naturally occurring components in human bones.

Acknowledgements The authors would like to acknowledge the nancial support of this project (ARC 18/97) from the Agency for Science, Technology and Research (A-STAR), Singapore.

References
[1] Patrick Jr CW, Mikos AG, Mcintire LV. Prospectus of tissue engineering. In: Patrick Jr CW, Mikos AG, Mcintire LV, editors. Frontiers in tissue engineering. New York: Pergamon, 1998. p. 35. [2] Hutmacher DW. Scaffolds in tissue engineering bone and cartilage. Biomaterials 2000;21:252943. [3] Kim SS, Utsunomiya H, Koski JA, Wu BM, Cima MJ, Sohn J, Mukai K, Grifth LG, Vacanti JP. Survival and function of hepatocytes on a novel three-dimensional synthetic biodegradable

K.H. Tan et al. / Biomaterials 24 (2003) 31153123 polymeric scaffold with an intrinsic network of channels. Ann Surg 1998;228:813. Baksh D, Davies JE, Kim S. Three-dimensional matrices of calcium polyphosphates support bone growth in vitro and in vivo. J Mater Sci: Mater Med 1998;9:7438. Yang SF, Leong KF, Du ZH, Chua CK. The design of scaffolds for use in tissue engineering: Part 1Traditional factors. Tissue Eng 2001;7(6):67990. Robinson BP, Hollinger JO, Szachowicz EH, Brekke J. Calvarial bone repair with porous d,l -polylactide. Otolaryng Head Neck 1995;112(6):70713. Ma PX, Zhang R. Synthetic nano-scale brous extracellular matrix. J Biomed Mater Res 1999;46:60. Boyan BD, Hummert TW, Dean DD, Schwartz Z. Role of material surfaces in regulating bone and cartilage cell response. Biomaterials 1996;17:13746. Zeltinger J, Sherwood JK, Graham DA, Mueller R, Grifth LG. Effect of pore size and void fraction on cellular adhesion, proliferation, and matrix deposition. Tissue Eng 2001;7(5): 55772. Weinberg CB, Bell E. A blood vessel model constructed from collagen and cultured vascular cells. Science 1986;231:397400. Cima LG, Langer R, Vacanti JP. Polymers for tissue and organ culture. J Bioact Compat Polym 1991;6:23240. Zhao F, Yin Y, Lu WW, Leong JC, Zhang W, Zhang J, Zhang M, Yao K. Preparation and histological evaluation of biomimetic three-dimensional hydroxyapatite/chitosangelatin network composite scaffolds. Biomaterials 2002;23:322734. Thomson RC, Shung AK, Yaszemski MJ, Mikos AG. Polymer scaffold processing. In: Lanza R, Langer R, Chick W, editors. Principles of tissue engineering. 2nd ed. Austin, TX: R.G. Landes, 2000. p. 25161. Kohn J, Langer R. Bioresorbable and bioderodible materials. In: Ratner BD, Hoffman AS, Schhoen FJ, Lemons JE, editors. Biomaterials science: an introduction to materials in medicine. New York: Academic Press, Inc, 1996. p. 6473. Lo H, Kadiyala S, Guggino SE, Leong KW. Poly(l-lactic acid) foams with cell seeding and controlled-release capacity. J Biomed Mater Res 1996;30(4):47584. Xiong Z, Yan Y, Zhang R, Sun L. Fabrication of porous poly(l-lactic acid) scaffolds for bone tissue engineering via precise extrusion. Scr Mater 2001;45:7739. Chua CK, Leong KF. Rapid prototypingprinciples and applications in manufacturing. Singapore: Wiley, 1997. Murphy WL, Dennis RG, Kileny JL, Mooney DJ. Salt fusion: an approach to improve pore interconnectivity within tissue engineering scaffolds. Tissue Eng 2002;8(1):4352. Mikos AG, Bao Y, Cima LG, Ingber DE, Vacanti JP, Langer R. Preparation of polyglycolic acid bonded ber structures for cell attachment and transplantation. J Biomed Mater Res 1993;27: 1839. Mooney DJ, Baldwin DF, Suh NP, Vacanti JP, Langer R. Novel approach to fabricate porous sponges of poly (d, l -lactic-co-

3123

[4]

[21] [22]

[5]

[6]

[23]

[7] [8]

[24]

[25]

[9]

[26]

[10] [11] [12]

[27]

[28]

[13]

[29]

[14]

[30] [31]

[15]

[32]

[16]

[17] [18]

[33] [34]

[19]

[35]

[20]

glycolic acid) without the use of organic solvents. Biomaterials 1996;17:141722. Whang K, Thomas CH, Healy KE, Nuber G. A novel method to fabricate bioasorbable scaffolds. Polymer 1995;36:83742. Cornejo IA, McNulty TF, Lee S, Bianchi E, Danforth SC, Safari A. Proceedings of the Bioceramics: Materials and Applications Symposium, held at the 101st Annual Meeting of the American Ceramic Society, Indianapolis, Indiana, April 2528, 1999. p. 18395. Zein I, Hutmacher DW, Tan KC, Teoh SW. Fused deposition modelling of novel scaffold architectures for tissue engineering applications. Biomaterials 2002;23:116985. Hollister SJ, Levy RA, Chu TM, Halloran JW, Feinberg SE. An image-based approach for designing and manufacturing craniofacial scaffolds. Int J Oral Maxillofacial Surg 2000;29:6771. Fisher JP, Vehof JWM, Dean D, Waerden JPC, Holland TA, Mikos AG, Jansen JA. Soft and hard tissue response to photocrosslinked poly (propylene fumarate) scaffolds in a rabbit model. J Biomed Mater Res 2002;59:54756. Giordano RA, Wu BM, Borland SW, Cima LG, Sachs EM, Cima MJ. Mechanical properties of dense polylactic acid structures fabricated by three dimensional printing. J Bio Sci Polym 1996; 8:6375. Curodeau A, Sachs E, Caldarise S. Design and fabrication of cast orthopedic implants with freeform surface textures from 3-D printed ceramic shell. J Biomed Mater Res 2000;53:52535. Lee G, Barlow JW, Fox WC, Aufdermorte TB. Biocompatibility of SLS-formed calcium phosphate implants. Proceedings of Solid Freeform Fabrication Symposium, Austin, TX, August 1214, 1996. p. 1522. Lee G, Barlow JW. Selective laser sintering of bioceramic materials for implants. Proceedings of Solid Freeform Fabrication Symposium, Austin, TX, August 911, 1993. p. 376380. Charrier JM, Polymeric materials and processing. Munich: Hanser Publishers, 1990. p. 1446. Katzer A, Marquardt H, Westendorf J, Wening JV, Foerster G von. Polyetheretherketone-cytotoxicity and mutagenicity in vitro. Biomaterials 2002;23:174959. Ha SW, Kirch M, Birchler F, Eckert KL, Mayer J, Wintermantel E, Sittig C, Pfund-Klingenfuss I, Textor M, Spencer ND, Guecheva M, Vonmont H. Surface activation of polyetheretherketone (PEEK) and formation of calcium phosphate coatings by precipitation. J Mater Sci: Mater Med 1997;8: 68390. 3D Systems Inc.: http://www.3dsystems.com. Leong KF, Phua KKS, Chua CK, Du ZH, Teo KOM. Fabrication of porous polymeric matrix drug delivery devices using the selective laser sintering technique. Proceedings of The Institution of Mechanical Engineers Part H. J Eng Med 2001;215: 191201. Nelson JC, Selective laser sintering: a denition of the process and an empirical sintering model. PhD thesis, The University of Texas, Austin, USA, 1993.