Plasma-sprayed calcium phosphate particles with high bioactivity and their use in bioactive scaffolds

Transcription

1 Biomaterials 23 (2002) Plasma-sprayed calcium phosphate particles with high bioactivity and their use in bioactive scaffolds Jie Weng a,b, Min Wang a, *, Jiyong Chen b a School of Mechanical and Production Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore , Singapore b Engineering Research Center in Biomaterials, Sichuan University, Chengdu , China Received 21 May 2001; accepted 20 November 2001 Abstract Highly crystalline feedstock hydroxyapatite (HA) particles with irregular shapes were spheroidized by plasma spraying them onto the surface of ice blocks or into water. The spherical Ca P particles thus produced contained various amounts of the amorphous phase which were controlled by the stand-off distance between the spray nozzle and the surface of ice blocks or water. The smooth surface morphology without cracks of spherical Ca P particles indicated that there were very low thermal stresses in these particles. Plasma-sprayed Ca P particles were highly bioactive due to their amorphous component and hence quickly induced the formation of bone-like apatite on their surfaces after they were immersed in an acellular simulated body fluid at 36.51C. Bone-like apatite nucleated on dissolved surface (due to the amorphous phase) of individual Ca P particles and grew to coalesce between neighboring Ca P particles thus forming an integrated apatite plate. Bioactive and biodegradable composite scaffolds were produced by incorporating plasma-sprayed Ca P particles into a degradable polymer. In vitro experiments showed that plasma-sprayed Ca P particles enhanced the formation of bone-like apatite on the pore surface of Ca P/PLLA composite scaffolds. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Hydroxyapatite; Amorphous phase; Plasma spray; Composite scaffold; Bone-like apatite 1. Introduction Human hard tissues contain mineral crystallites that are substituted hydroxyapatite (HA) and the crystallites are distributed in the organic matrix [1]. Synthetic hydroxyapatite, Ca 10 (PO 4 ) 6 (OH) 2, is one ofthe calcium phosphates and can be anchored to bone by establishing a physicochemical bond with living tissue [2]. It has been under intensive research over the past 20 years as a leading bioceramic and has been used clinically in various forms due to its osteoconductive property [1]. However, brittleness and poor mechanical properties do not allow HA to be used in load-bearing parts ofthe body. Therefore, various methods have been investigated in order to realize the potential ofha as a bioactive ceramic. Among these methods, developing an HA coating on metal implants and using particulate HA *Corresponding author. Tel.: ; fax: address: mmwang@ntu.edu.sg (M. Wang). as the bioactive phase in composites have been extensively studied [3,4]. Among the coating techniques, plasma spraying HA on metal implants has been well established, which produces implants that combine excellent mechanical properties ofthe metal substrate with osteoconductivity ofha [1]. Coating deposition by plasma spraying is a high-temperature process and is thermodynamically non-equilibrium in nature. It induces unavoidable changes offeedstock HA powders, which makes plasma-sprayed HA coatings distinctively different from HA monoliths processed through the conventional compaction-sintering route. It was found that plasmasprayed HA coatings were more active to induce the formation of an apatite layer on their surfaces than the well-crystallized HA ceramic when they were immersed in a simulated body fluid (SBF) [5]. Therefore, plasma spraying may provide a new means to produce calcium phosphate particles with desired characteristics. Synthetic crystalline HA particles have been used in various composites [4,6,7]. Besides improvements in /02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S (01)

2 2624 J. Weng et al. / Biomaterials 23 (2002) mechanical properties, using a more bioactive particulate material than crystalline HA particles should significantly increase the bioactivity ofthe composites, which could lead to the faster establishment of a secure bond between the composite implant and tissue. Furthermore, biodegradable scaffolds containing a highly bioactive phase may have great potential for tissue-engineering applications. 2. Experimental procedures The hydroxyapatite powder for plasma spraying in the current investigation was synthesized using the wet method according to the following reaction: 10CaðNO 3 Þ 2 þ6ðnh 4 Þ 3 PO 4 þ2nh 4 OH -Ca 10 ðpo 4 Þ 6 ðohþ 2 þ20nh 4 NO 3 : ð1þ Calcined powders ofwell-crystallized HA without thermal decomposition were produced. Plasma spraying was conducted in normal atmosphere using a METCO MN Plasma System and an AR Thermal Spray Robot under a condition of A arc current and V arc voltage with nitrogen as the plasma gas. Calcium phosphate (Ca P) particles were produced by plasma-spraying feedstock HA powders into distilled water or onto the surface of ice blocks which were positioned 12 or 100 cm away from the spray nozzle. The sprayed particles were collected, dried in an oven and stored in an electrical desiccator. The ultrasonic treatment ofplasma-sprayed Ca P particles was performed at room temperature in a Cole- Parmer s 8890E-DTH ultrasonic cleaner at a frequency of42 khz. Ca P particles were immersed in distilled water during the ultrasonic treatment. They were collected and dried prior to analysis. The poly(l-lactic acid) (PLLA) polymer used in the current investigation had an average molecular weight of300,000 and was provided by Chengdu Institute of Organic Chemistry, Academia Sinica, China. To produce porous Ca P/PLLA composites, plasma-sprayed Ca P particles were firstly dispersed in acetone before PLLA chips were dissolved in the solvent. The composite solution was then mixed with sugar particles ofa particular size and subsequently cast into a metal mould. The Ca P/sugar/PLLA composite was subjected to extraction ofsugar particles in distilled water which was a non-solvent for PLLA but was miscible with acetone. In this way, a highly porous PLLA-matrix scaffold containing 20 wt% of plasma-sprayed Ca P particles was produced. The manufacture of porous PLLA followed the same procedure without adding Ca P particles at the beginning. Porous PLLA and Ca P/PLLA samples were cut into rectangular blocks of dimensions of8 mm 6mm 6 mm and stored in an electrical desiccator until utilization. An acellular SBF was used for in vitro experiments as it had ion concentrations that were nearly the same as those ofhuman blood plasma [8]. SBF was prepared by dissolving reagent-grade chemicals ofnacl, NaHCO 3, KCl, K 2 HPO 4 3H 2 O, MgCl 2 6H 2 O, CaCl 2 2H 2 O, Na 2 SO 4, and (CH 2 OH) 3 CNH 2 into distilled and deionized water and buffered to ph 7.4 at 36.51C with HCl. During in vitro experiments, as-sprayed Ca P particles and porous PLLA and Ca P/PLLA specimens were immersed in sealed polyethylene bottles filled with SBF, which were maintained at 36.51C in a PolyScience s 20L-M water bath and subjected to a gentle reciprocal motion controlled by a PolyScience s SH20L dual action shaker. A series ofevacuation pressurization cycles were carried out in a SHEL LAB 1410D vacuum oven to force SBF into the interconnecting pores in porous specimens before they were immersed in SBF in polyethylene bottles. The cycling was continued until no air bubbles emerged from the specimens. The volume of SBF in each polyethylene bottle was large enough as compared with the volume ofspecimens immersed. After different periods of time, specimens were removed from SBF, rinsed with distilled water and dried prior to analysis. After their respective treatments, specimens were coated with platinum and examined using a JEOL JSM5600LV scanning electron microscope (SEM). A Philips PW 1830/00 X-ray diffractometer (XRD) was used to investigate phase changes between feedstock HA powders and plasma-sprayed Ca P particles. Mineral crystals formed in vitro on surfaces of Ca P particles and their composites with PLLA were analyzed using a Rigaku DMAX thin-film X-ray diffractometer (TF-XRD). Fourier transform infrared (FTIR) spectra ofthe mineral crystals were obtained using a Perkin Elmer System 2000 FTIR spectrometer. A small amount ofthe mineral crystals was gathered from an immersed specimen, milled with KBr and pressed into a transparent film for FTIR analysis. Care was taken to avoid the incorporation ofmoisture in KBr plates during FTIR analysis. 3. Results and discussion The feedstock HA particles were irregular in shape and had a wide particle size range (Fig. 1a). They were spheroidized by plasma spraying no matter how they were collected, on the ice surface or in distilled water (Fig. 1b). The surface of plasma-sprayed spherical particles was very smooth. Almost all HA particles melted completely in the plasma flame under the current plasma-spraying condition. The few partially melted HA particles oflarge particle sizes had irregular shapes but with a smooth surface morphology, indicating the complete melting ofthe outer layer ofthe particle and

3 J. Weng et al. / Biomaterials 23 (2002) Fig. 2. XRD patterns of plasma-sprayed Ca P particles with different stand-off distances: (a) 12 cm and (b) 100 cm. Fig. 1. SEM micrographs of (a) feedstock HA powder for plasma spraying and (b) plasma-sprayed Ca P particles. an unmelted core. These particles were different from the partially molten HA particles which showed only partial melting ofthe particle surface [9]. It was shown in the current investigation that even with rapid solidification on ice, there were no cracks on the smooth surface of plasma-sprayed Ca P particles, indicating that the surface tension of the particles was high enough to resist thermal stresses. This was also different from the surface of partially molten HA particles which were split by internal stresses after being sprayed into water [9]. It seems that the splitting may have resulted from fast cooling of nearly unmelted HA particles from the plasma-spraying temperature. The spheroidized Ca P particles contained certain amounts ofthe amorphous phase. With the increase in stand-off distance between the spray nozzle and the surface of the ice block, the proportion of the amorphous component decreased significantly with a simultaneous increase in the crystalline HA phase, as indicated by XRD patterns shown in Fig. 2. As-sprayed particles had experienced the same heating process in the plasma flame but different cooling processes as the stand-off distance changed. The crystalline HA phase in Ca P particles was a result ofrecrystallization ofapatite from the fused state and its amount increased with the stand-off distance. Although almost all particles experienced the same melting process in the plasma flame, the cooling rate was higher for particles solidified in a short stand-off distance than for those in a longer distance. A high cooling rate had led to a large percentage ofthe amorphous component (the frozen liquid) in the solidified particle. The spherical as-sprayed Ca P particles were ultrasonically treated in distilled water in order to investigate their morphological changes after enhanced dissolution. It was observed that the dissolution was not homogeneous in a layer-by-layer fashion but was preferential at different locations of the particle surface. Ultrasonically treated Ca P particles also showed the grain structure inside the particle (Fig. 3), which is similar to that of sintered HA ceramic after surface preferential dissolution [10]. The direct observation ofha grains in plasma-sprayed Ca P particles indicated that recrystallization ofapatite occurred in molten Ca P particles during their rapid solidification. There were no cracks appearing on the surface of plasma-sprayed Ca P particles upon ultrasonic treatment in distilled water. The spherical morphology with a

4 2626 J. Weng et al. / Biomaterials 23 (2002) Fig. 3. SEM micrograph ofplasma-sprayed Ca P particles after ultrasonic treatment in distilled water for 12 h. smooth surface suggested that the thermal contraction in as-sprayed Ca P particles was homogeneous, resulting in very small and homogeneous thermal stresses in particles. Therefore, the occurrence of cracks due to water-enhanced stress relaxation during ultrasonic treatment in distilled water was not observed on Ca P particles. In the case ofplasma-sprayed HA coatings which were subjected to the same ultrasonic treatment in distilled water, cracks appeared on the coatings even for very short periods oftreatment time [10]. In the plasmaspray process, thermal stresses built up in the coatings due to the immobilization ofsolidified particles on the substrate surface during flattening of molten particles upon impinging the substrate. In addition, thermal stresses varied from layer to layer away from the coating substrate interface because of the different thermal experience ofcoating layers during continuous spraying and flattening. A non-uniform state of tensile strain was found to exist in plasma-sprayed HA coatings [11]. Therefore, water-enhanced relaxation of thermal stresses induced cracks and the subsequent preferential dissolution destroyed the integrity ofplasma-sprayed HA coatings [10]. Plasma-sprayed Ca P particles were highly bioactive and hence quickly induced the formation of a calcium phosphate mineral on their surfaces after they were immersed in SBF at 36.51C. The TF-XRD pattern ofthe mineral formed on Ca P particles immersed in SBF for 5 days is shown in Fig. 4a. The broad peak between 261 and 321 (2y) belonged to apatite ofpoor crystallinity. Fig. 4b is the FTIR spectrum ofthe mineral formed on Ca P particles. The broad peaks at around 1060 and 600 cm 1 were due to phosphate groups in apatite. The hydroxyl peak for stretching vibration at 3571 cm 1 was not observed, but the peaks at and 875 cm 1 due to carbonate groups were present. FTIR result indicated that carbonate groups were incorporated and Fig. 4. Analysis ofmineral crystals formed on plasma-sprayed Ca P particles immersed in SBF at 36.51C for 5 days: (a) TF-XRD pattern and (b) FTIR spectrum. had substituted for hydroxyls in the apatite lattice. The apatite formed in vitro was similar to those found on other bioactive materials [12 14] and was termed bonelike apatite. Fig. 5 shows the nucleation and growth ofbone-like apatite on Ca P particles immersed in SBF. During the first few hours of immersion, there was only dissolution ofthe amorphous component in the surface layer ofassprayed Ca P particles (Fig. 5a). This dissolution process was the same as the one that occurred on plasma-sprayed HA coatings immersed in SBF [5]. The nucleation ofbone-like apatite took place subsequently on the dissolved surface of individual Ca P particles as shown in Fig. 5b. The bone-like apatite on individual Ca P particles grew with time and coalesced between adjacent Ca P particles to form an integrated apatite plate which had a smooth, sanddune-like morphology (Fig. 5c). The growth and coalescence ofapatite layers on neighboring Ca P particles also filled the interparticle space. This observation may explain why the bone-like apatite layer always exhibited a similar morphology on plasma-sprayed HA coatings after their immersion in SBF regardless oftheir original surface roughness [5]. Results from in vitro experiments showed that plasma-sprayed Ca P particles enhanced the formation ofbone-like apatite on the surface ofca P/PLLA composite scaffolds when they were immersed in SBF.

5 J. Weng et al. / Biomaterials 23 (2002) Fig. 6. SEM micrographs of Ca P/PLLA composite scaffolds immersed in SBF at 36.51C for 7 days: (a) formation of bone-like apatite and (b) degradation ofsome struts. Fig. 5. SEM micrographs ofplasma-sprayed Ca P particles immersed in SBF at 36.51C for (a) 12 h, (b) 24 h and (c) 120 h. After 7 days ofimmersion, a large number of islands of bone-like apatite were formed on the pore surface of composite scaffolds (Fig. 6a), indicating high bioactivity ofthe composite. It was also observed that within the same immersion period, some struts ofthe composite scaffolds degraded (Fig. 6b), possibly due to the simultaneous degradation ofplla and dissolution of Ca P particles. In contrast, almost no nuclei ofbonelike apatite appeared on the pore surface of PLLA scaffolds (Fig. 7). The apatite formed on Ca P/PLLA composite scaffolds had the same characteristics as those of the apatite formed on plasma-sprayed Ca P particles (Fig. 4). They were all bone-like apatite. This result suggests that different substrates, as long as they are bioactive and they can provide necessary ions, may be used as templates for the formation and growth of bone-like apatite in SBF. This observation can have significant implications for developing new biomaterials using the biomimetic approach. It was reported that after 15 days of immersion in a physiological fluid which had ion concentrations 1.5 times those ofsbf, apatite microparticles were observed on the pore surface of a PLLA foam [15]. The PLLA surface may be suitable for apatite formation if ion

6 2628 J. Weng et al. / Biomaterials 23 (2002) Fig. 7. SEM micrograph of PLLA scaffold immersed in SBF at 36.51C for 7 days. concentrations in the physiological fluid are sufficiently high. Due to the low ion concentrations in SBF used in the current investigation, the nucleation ofapatite would take a long time to occur on the pore surface of pure PLLA as compared with the nucleation process when 1.5SBF was used as the physiological fluid for in vitro experiments. Therefore, it is feasible to enhance the bioactivity ofplla by incorporating resorbable bioceramics such as plasma-sprayed Ca P particles, which has been shown in the current investigation. Plasma-sprayed Ca P particles are ready to dissolve in SBF and can quickly induce the formation of bonelike apatite on their surfaces. The dissolution of the Ca P particles incorporated in the composite was believed to increase local concentrations ofcalcium and phosphate ions and hence promote the nucleation and growth ofapatite (Fig. 6a). The enhanced nucleation and growth in SBF by incorporating plasmasprayed Ca P particles in the PLLA matrix was evidently more effective than those by increasing ion concentrations in the physiological fluid. This implies that Ca P/PLLA scaffolds may be very bioactive in vivo and hence can induce the formation of bone-like apatite on their pore surface when they are implanted in the human body. Therefore, Ca P/PLLA composite scaffolds can be considered for tissue-engineering applications due to their high bioactivity and biodegradability. 4. Conclusions Highly crystalline HA particles can be spheroidized into Ca P particles containing certain amounts ofthe amorphous phase by the plasma-spray technique. The proportion ofthe amorphous phase in Ca P particles is controlled by the stand-off distance between the spray nozzle and the collection surface. The crystalline phase in plasma-sprayed Ca P particles is a result of recrystallization of HA from its fused state upon impinging a cold surface. The smooth surface without cracks ofspherical Ca P particles indicates that there are very low thermal stresses in these particles. Plasma-sprayed Ca P particles are highly bioactive due to the amorphous component and can quickly induce the formation of bone-like apatite on their surfaces after they are immersed in SBF at 36.51C. Bone-like apatite nucleates on dissolved surface (due to the amorphous phase) ofindividual Ca P particles and grows to coalesce between neighboring Ca P particles thus forming an integrated apatite plate. Bioactive and biodegradable composite scaffolds can be produced by incorporating plasma-sprayed Ca P particles into degradable polymers. In vitro experiments show that plasma-sprayed Ca P particles enhance the formation ofbone-like apatite on the pore surface ofca P/PLLA composite scaffolds. Acknowledgements The authors would like to thank Nanyang Technological University (NTU), Singapore, Chinese National Natural Science Foundation, China, for funding the research. J. Weng thanks NTU for providing a research fellowship. Assistance provided by technical staff in the School ofmpe, NTU, is gratefully acknowledged. References [1] Hench LL, Wilson J, editors. An introduction to bioceramics. Singapore: World Scientific, [2] Jarcho M, Kay JF, Gummaer KI, Doremus RH, Drobeck HP. Tissue, cellular and subcellular events at a bone ceramic hydroxyapatite interface. J Bioeng 1977;1: [3] de Groot K, Geesink R, Klein CPAT, Serekian P. Plasma-sprayed coatings ofhydroxyapatite. J Biomed Mater Res 1987;21: [4] Bonfield W, Wang M, Tanner KE. Interfaces in analogue biomaterials. Acta Materialia 1998;46: [5] Weng J, Liu Q, Wolke JGC, Zhang X, de Groot K. Formation and characteristics ofapatite layer on plasma-sprayed hydroxyapatite coatings in simulated body fluid. Biomaterials 1997;18: [6] Ito M, Hidaka Y, Nakajima M, Yagasaki H, Kafrawy AH. Effect ofhydroxyapatite content on physical properties and connective tissue reactions to a chitosan hydroxyapatite composite membrane. J Biomed Mater Res 1999;45: [7] Cerrai P, Guerra GD, Tricoli M, Krajewski A, Guicciardi S, Ravaglioli A, Maltinti S, Masetti G. New composites of hydroxyapatite and bioresorbable macromolecular material. J Mater Sci Mater Med 1999;10: [8] Kokubo T, Kushitani H, Sakka S, Kitsugi T, Yamamuro T. Solution able to reproduce in vivo surface-structure changes in bioactive glass-ceramic A-W. J Biomed Mater Res 1990;24:

7 J. Weng et al. / Biomaterials 23 (2002) [9] Palka V, Postrkova E, Hoerten HK. Some characteristics of hydroxylapatite powder particles after plasma spraying. Biomaterials 1998;19: [10] Weng J, Wolke JGC, Zhang X, de Groot K. Environmentally induced cracks and detachment ofplasma-sprayed hydroxyapatite coatings under ultrasonic treatment. In: Wilson J, Hench LL, Greenspan D, editors. Bioceramics, vol. 8. Oxford: Pergamon, p [11] Tufekci E, Brantley WA, Mitchell JC, Foreman DW, Georgette FS. Crystallographic characteristics ofplasma-sprayed calcium phosphate coatings on Ti-6Al-4V. Int J Oral Maxillofac Implants 1999;14: [12] Kokubo T. Bioactive glass ceramics: properties and applications. Biomaterials 1991;12: [13] Ohtsuki C, Kokubo T, Yamamuro T. Mechanism ofapatite formation on CaO SiO 2 P 2 O 5 glasses in a simulated body fluid. J Non-Cryst Solids 1992;143: [14] Weng J, Wang M. In vitro formation of bone-like apatite on the surface of solution-cast poorly crystallized hydroxyapatite/chitin composite. Key Eng Mater 2001; : [15] Zhang R, Ma PX. Porous poly(l-lactic acid)/apatite composites created by biomimetic process. J Biomed Mater Res 1999;45: