Microstructure and biocompatibility of composite biomaterials fabricated from titanium and tricalcium phosphate by spark plasma sintering Dibakar Mondal, 1 Linh Nguyen, 1 Ik-Hyun Oh, 2 Byong-Taek Lee 1 1 Department of Biomedical Engineering and Materials, College of Medicine, Soonchunhyang University, Cheonan 330-090, Korea 2 Korea Institute of Industrial Technology, Gwangju Research Center, Korea Received 31 May 2012; revised 11 September 2012; accepted 14 September 2012 Published online 7 November 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.a.34455 Abstract: Important issues in developing hydroxyapatite (HAp)- and titanium (Ti)-based composite biomaterials for orthopedic or dental devices include the dissociation of HAp during fabrication and its influences in the microstructure and biocompatibility of the final composite. During the densifica- tion by sintering of HAp/Ti composites, Ti reacts with AOH freed from HAp to form TiO 2 thus dissociated HAp into Ca 3 (PO 4 ) 2 , CaO, CaTiO 3 , TiP, and so forth. To inhibit this reac- tion, composites were fabricated with Ti and 30, 50, and 70 vol % b-tricalcium phosphate (b-TCP) instead of HAp by spark plasma sintering at 1200 C. It has been observed that after sin- tering at 1200 C, Ti also reacted with TCP, but unlike HAp/Ti composites, the final TCP/Ti composites contained significant amounts of unreacted TCP and Ti phases. The initial 70 vol % TCP/Ti composite showed compressive strength of 388.5 MPa, Young’s modulus of 3.23 GPa, and Vickers hardness of 361.9 HV after sintering. The in vitro cytotoxicity and proliferation of osteoblast cells on the composites surfaces showed that the addition of a higher amount of TCP with Ti was beneficial by increasing cell viability, cell–composite attachment and prolif- eration. Osteopontin and collagen type II protein expression from osteoblasts cultured onto the 70% TCP–Ti composite was also higher than other composites and pure Ti. In vivo study verified that within 3 months of implantation in an animal body, 70% TCP–Ti had an excellent bone–implant interface compared with a pure Ti metal implant. V C 2012 Wiley Periodicals, Inc. J Biomed Mater Res Part A: 101A: 1489–1501, 2013. Key Words: titanium, tricalcium phosphate, composites, bio- activity, hard tissue replacement How to cite this article: Mondal D, Nguyen L, Oh I-H, Lee B-T. 2013. Microstructure and biocompatibility of composite biomaterials fabricated from titanium and tricalcium phosphate by spark plasma sintering. J Biomed Mater Res Part A 2013:101A:1489–1501. INTRODUCTION Titanium (Ti) is widely used as an orthopedic implant because of its favorable mechanical properties, excellent corrosion resistance, and biocompatibility. 1 However, the long-term clinical performances are compromised by creat- ing a stress shielding effect at the bone/implant interfaces. The stress shielding effect, whereby the reabsorption of nat- ural bone and the loosening implant arise because of the difference in Young’s modulus between natural bone and a Ti implant, is one of the primary reasons for revision sur- gery. 2–4 From the viewpoint of biocompatibility, calcium phosphate (CaP) ceramics [mostly hydroxyapatite (HAp)] seem to be the most suitable ceramics for hard tissue replacement. However, HAp and other CaP ceramics usually have poor mechanical properties, which have limited their applications as load-bearing tissue implants. Many efforts have been made to improve the mechanical properties of HAp 5,6 and the biological properties of Ti and its alloys. 7,8 Achieving a good combination of the biocompat- ibility of CaP and the favorable mechanical properties of metals is considered a promising approach to fabricate more perfect biomedical devices for load-bearing applica- tions. Most popular approaches are to apply a coating of HAp to a Ti surface, especially by plasma spray coating. But plasma spray coating is a complex process and damages the microstructure of HAp. 8–12 Moreover, the bonding strength at the interface is weak enough to keep this layer-by-layer coating process still under optimization. 8,12,13 Another well- studied approach is to fabricate homogenous composites of HAp and Ti by powder metallurgy to optimize both mechan- ical properties and biocompatibility. But in high temperature sintering to make dense composite bodies, Ti reacts with HAp and decomposes it into Ca 3 (PO 4 ) 2 , CaO, CaTiO 3 , TiO 2 , and so forth. 10,14,15 As reported, although sintering at tem- peratures >1000 C, hydroxyl radical (OH ) from HAp reacts with Ti to form TiO 2 . Then this TiO 2 dissociates HAp into various reaction products. In the final product, usually no HAp or Ti phase remains. The reactions occurred as 13,15–17 : Correspondence to: B.-T. Lee; e-mail: [email protected]Contract grant sponsor: Mid-career Researcher Program through NRF grant funded by the MEST; contract grant number: 2009-0092808 V C 2012 WILEY PERIODICALS, INC. 1489
Transcript
Microstructure and biocompatibility of composite biomaterialsfabricated from titanium and tricalcium phosphate by spark plasmasintering
How to cite this article:Mondal D, Nguyen L, Oh I-H, Lee B-T. 2013. Microstructure and biocompatibility of composite biomaterialsfabricated from titanium and tricalcium phosphate by spark plasma sintering. J Biomed Mater Res Part A 2013:101A:1489–1501.
INTRODUCTION
Titanium (Ti) is widely used as an orthopedic implantbecause of its favorable mechanical properties, excellentcorrosion resistance, and biocompatibility.1 However, thelong-term clinical performances are compromised by creat-ing a stress shielding effect at the bone/implant interfaces.The stress shielding effect, whereby the reabsorption of nat-ural bone and the loosening implant arise because of thedifference in Young’s modulus between natural bone and aTi implant, is one of the primary reasons for revision sur-gery.2–4 From the viewpoint of biocompatibility, calciumphosphate (CaP) ceramics [mostly hydroxyapatite (HAp)]seem to be the most suitable ceramics for hard tissuereplacement. However, HAp and other CaP ceramics usuallyhave poor mechanical properties, which have limited theirapplications as load-bearing tissue implants.
Many efforts have been made to improve the mechanicalproperties of HAp5,6 and the biological properties of Ti andits alloys.7,8 Achieving a good combination of the biocompat-
ibility of CaP and the favorable mechanical properties ofmetals is considered a promising approach to fabricatemore perfect biomedical devices for load-bearing applica-tions. Most popular approaches are to apply a coating ofHAp to a Ti surface, especially by plasma spray coating. Butplasma spray coating is a complex process and damages themicrostructure of HAp.8–12 Moreover, the bonding strengthat the interface is weak enough to keep this layer-by-layercoating process still under optimization.8,12,13 Another well-studied approach is to fabricate homogenous composites ofHAp and Ti by powder metallurgy to optimize both mechan-ical properties and biocompatibility. But in high temperaturesintering to make dense composite bodies, Ti reacts withHAp and decomposes it into Ca3(PO4)2, CaO, CaTiO3, TiO2,and so forth.10,14,15 As reported, although sintering at tem-peratures >1000�C, hydroxyl radical (OH�) from HAp reactswith Ti to form TiO2. Then this TiO2 dissociates HAp intovarious reaction products. In the final product, usually noHAp or Ti phase remains. The reactions occurred as13,15–17:
In a previous study by Ning and Zhou14 on fabricatingcomposites of Ti/HAp/bioglass at 1200�C, even 30 vol % Tiwith 60 vol % HAp and 10 vol % bioglass did not containany CaP phases in the final product. They tried to preparecomposites of 30, 50, and 70 vol % Ti with 10 vol % bio-glass in each composite, and the rest of the compositeswere composed of HAp. They showed that the final compo-sites had no CaP phases remained, and in 70 vol % Ti pre-cursor-based composite, no pure Ti phase remained due tovarious complex reactions that occurred between Ti andHAp while sintering. Yang et al.10 and Chu et al.18 alsoshowed that at elevated temperature >800�C, Ti reactedwith HAp as shown in Eqs. (1)–(3) and dissociated intoCaTiO3, TiO2, tricalcium phosphate (TCP), and so forth. Inanother study performed by Nath et al.,15 it was revealedthat with increasing the initial amount of HAp, the reactionrate also increased, and in final composites, the amount ofCaP decreased remarkably. Another approach was per-formed by Marcelo et al.19 in which they tried to fabricateTi–HAp composite by preparing a green body with TiH2 andHAp raw powder mixture and vacuum sintering (VS) themixture at 1150�C. TiH2 was used instead of Ti to reducethe reactivity of Ti, but they also ended up with dissociationof HAp and various alien products. Miura-Fujiwara et al.20
fabricated a composite of 50 vol % TCP–Ti through sparkplasma sintering (SPS) method and they characterized thehardness and microstructure behavior for different holdingtime of sintering. However, their study did not depict anydetail evaluation of the final composites including composi-tional optimization or biocompatibility.
In reviewing available literature, it was also criticallynoted that no systematic investigation has been carried outto fabricate and optimize the composition of homogenouscomposites of Ti and TCP. TCP has no Hydroxyl radical(OH�), which will inhibit the formation of TiO2 initially.Thus, dissociation of Ti metal or CaP may not occur, and theoptimizing of their mechanical properties and biocompatibil-ity will be possible.
a- and b-TCP are both currently used in various clinicalapplications in dentistry, maxillofacial surgery and orthope-dics. b-TCP is a component of several commercial mono- orbiphasic bioceramics and composites, and a-TCP is a majorconstituent of the powder component of various hydraulicbone cements.21,22 From the biological point of view, a-TCPis nontoxic, osteoconductive, and bioactive, both in vitro andin vivo.23 The main reason for the growing interest in a-TCPas a bone implant material is its biodegradability. It is morebioreabsorbable than HAp, which makes a-TCP an ideal
implant material that can be replaced by new bone fasterthan the other CaP-based materials currently available inthe market.23
In an earlier study, it was shown that for fabricating Tiand biphasic calcium phosphate composites, SPS in inertatmosphere is more convenient than conventional VS.24 Inbrief, VS process takes a longer time than SPS process tosinter the composites properly, and prolonged sintering du-ration helps the Ti/HAp system to be dissociated. In con-trast, SPS takes a very short time (<15 min) to sinter thecomposite properly, and existing external pressure helps thecomposite body to be denser and mechanically stronger.25
In this study, homogenous composites of 30, 50, and 70vol % commercially pure titanium (cpTi) with TCP werefabricated successfully by SPS method at 1200�C. X-ray dif-fraction (XRD), scanning electron microscopy (SEM), andenergy dispersive spectra (EDS) were used to investigateand compare the phase behaviors and microstructures ofthe Ti–TCP composites. In vitro studies with osteoblast-likeMG63 cells were investigated to confirm that the newly fab-ricated TCP–Ti composites are viable and biocompatiblewith bone cells by using MTT assay and confocal laser scan-ning microscopy (immunostaining). Collagen and osteopon-tin (OPN) expression were evaluated by western blot analy-sis. In addition, an initial in vivo study was carried out onthe femur of a mature male rabbit for 3 months to investi-gate the initial feasibility of clinical application of Ti–TCPcomposite.
MATERIALS AND METHODS
MaterialscpTi (Ti powder, �325 mesh, 99.5%) was purchased fromAlfa Aesar. Tri-CaP (b-phase basis, <2lm particle size,>98% pure) was purchased from Sigma Aldrich. The MG-63cell line of human osteoblast-like cells, which were derivedfrom human osteosarcoma, was obtained from the KoreanCell Bank. Dulbecco’s Modified Eagle’s Medium (DMEM,HyClone, Logan, UT), fetal bovine serum (FBS, Grand Island,NY), penicillin/streptomycin (antibiotics), and trypsin–ethyl-enediaminetetraacetic acid (EDTA) were purchased fromGIBCO (Carlsbad, CA). Hexamethyldisilazane (HDMS, =99%,Sigma) and dimethylsulfoxide (DMSO 99.0%, Samchun PureChemical Co., Korea) were used as received. MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) pow-der was purchased from Sigma.
Ti–TCP composite preparationA 30 vol % TCP and 70 vol % Ti raw powders were mixedby ball milling in an agate milling pot using agate balls for24 h. Similarly, powders of pure TCP, 50 vol % TCP with 50vol % Ti, 70 vol % TCP with 30 vol % Ti raw powder mix-ture and pure Ti powder were prepared. The mixed powderof Ti and TCP was consolidated by using SCM DR. SinterLabTM (Model SPS-515S). During consolidation, approxi-mately 1 g of the powder mixture was charged into a graph-ite mold without prepressing and then heated to the desiredsintering temperature of 1200�C at a heating rate of approx-imately 192�C/min. A constant uniaxial pressure of 50 MPa
1490 MONDAL ET AL. MICROSTRUCTURE AND BIOCOMPATIBILITY OF COMPOSITE BIOMATERIALS
was applied to the powder. An external power source pro-vided a direct current (DC) on–off pulsed electrical dis-charge to activate the particle surface throughout the sinter-ing process. The on–off pulse ratio was 12:2. The sparkplasma discharges were generated by an instantaneouspulsed DC applied through electrodes at the top and bottompunches of graphite. Due to these discharges, the particlesurface was instantaneously activated and purified, and con-currently self-heating phenomena were generated amongthese particles, leading to heat transfer that was completedin an extremely short time. After the sintering was com-pleted, the sample was cooled to a temperature below100�C within 12 min. A PR-type thermocouple was insertedinto the graphite mold to measure the sintering tempera-ture. The total time for SPS was <10 min, and all the sam-ples were consolidated to a pellet shape, 10 mm in diame-ter, and 2 mm in height.
Phase and mechanical propertiesThe density of the composites was measured by Archimedesmethod. The Vickers’ hardness was measured using a hard-ness testing machine (Microvicker, Akashi, Japan) withapplying 9.8 N force and indentation duration of 16 s andthe compressive strength was measured using a universaltesting machine (Unitech TM, R&B, Korea). Young’s moduluswas evaluated using the three point bending method at aHounsfield series S testing machine with a crosshead speedof 1 mm s�1. The microstructure of the composites wasimaged with a Scanning Electron Microscope (JSM-6701F,JEOL, Japan) at 10.0 kV acceleration voltages after coatingwith Pt. EDS were analyzed by using INCAEnergy (OxfordInstruments Analytical). The EDS was operated in areamode with an accumulated live time of 50 s at 15 kV, pro-ducing a penetration depth of 300 nm. In addition, thephase and crystal structure were determined by XRD (Mini-flex II, Rigaku, Japan) with Cu Ka radiation (30 kV, 15 mA,and k ¼ 1.5406 Å). The XRD patterns were recorded in a2y range of 20–60� in continuous scan mode with 1�/minscan speed and step size of 0.02�. JCPDS-ICDD was used asreference data by PDXL-XRD data analysis software for theinterpretation of X-ray patterns obtained in this work.
In vitro cytocompatibilityMTT assay. MG-63 cells were cultured in the surface ofcomposites and suspended in a humidified incubator at37�C in a 5% CO2 atmosphere (incubator, ASTEC, Japan) forcertain time using DMEM supplemented with 10% FBS and1% penicillin–streptomycin as antibiotics (Bio-Whittaker).After collecting the cells in walls, 100 lL MTT (5 mg/mL inphosphate buffer saline) solutions were added to the wells.Metabolically active and viable cells produce mitochondrialdehydrogenase enzymes during incubation in media, whichis usually catalyzed by MTT salt. This reaction between cel-lular dehydrogenase and MTT salt produces purple-coloredformazan in the media. Cell proliferation is quantified bymeasuring the change in color intensity using a spectropho-tometer. The optical density (OD) is directly proportional tothe viable cell number. Therefore, in this study, cell prolifer-
ation after 1, 3, and 7 days were quantified after finishingthe successive incubation periods by adding the MTT solu-tion to each of the wells. After 4 h of incubation and solubi-lizing the formazan in 100 lL DMSO, the optical densities ofthe solution were measured using an enzyme-linked immu-nosorbent assay reader (EL, 312, Biokinetics reader, Bio-Tekinstruments) at a wavelength of 595 nm.
Cell morphology analyses. Ti–30% TCP, Ti–50% TCP, andTi–70% TCP scaffolds were initially sterilized under UV lightfor 6 h per surface, sterilized further with 70% ethanol (30min), washed with phosphate-buffered saline (PBS), andfinally soaked in DMEM for 1 h. MG-63 cells were seeded onthe scaffolds at a density of 104 cells/cm2 in a 24-well platewith 10% FBS and 1% PS culture medium. The attachmentbehaviors of MG-63 were analyzed after 30 and 60 min seed-ing and observed by SEM. For SEM observation, the cellswere fixed in 2% glutaraldehyde (v/v) and dehydrated inethanol solutions (50%, 70%, 90%, and 100%). The dehy-drated cells were dried in HDMS and coated by Pt sputtering.
Cell morphology behavior was visualized using a confo-cal microscope. Osteoblast-like MG-63 cells (104 cells/mL)were seeded onto the scaffolds and immunofluorescencedwith a fluorescence-conjugated antibody. In short, after 1, 3,and 7 days of culture (in a humidified CO2 incubator provi-sioned with 5% CO2 level, 37�C), the scaffolds were rinsedwith PBS twice and fixed in 4% paraformaldehyde (Sigma–Aldrich) for 15 min at room temperature. The cells werethen permeabilized with 0.25% Triton X-100 (Sigma–Aldrich) for 10 min. Treatment with 2.5% BSA for 30 minwas used as blocking reagent. Cells were immunostained byusing fluorescein isothiocyanate conjugated Phalloidin (25lg/mL-Sigma) for 2 h at room temperature. Nuclei werecounterstained with 30 lg/mL of DAPI (40,6-diamidino-2-phenylindole, Invitrogen). The scaffolds were visualizedunder a confocal fluorescent microscope (Olympus, FV10i-W) and images were analyzed using accompanying FV10i-ASW 3.0 Viewer software. Cell nuclei were colored orangeby viewer software during processing the images for bettercontrast and visualization.
Immunoblotting. The cells were rinsed and harvested usinga lysis buffer (Tris 50 mM, pH 7.4, NaCl 40 mM, EDTA 1 mM,Triton X-100 0.5%, Na3VO4 1.5 mM, NaF 50 mM, sodiumpyrophosphate 10 mM, glycerolphosphate 10 mM, PMSF 1mM, and protease inhibitor cocktail 10 mM), vortexed, andcentrifuged for 10 min at 13,000 rpm at 4�C. Aliquots (30lg) of the proteins were analyzed via Western blotting usinga 1:200 dilution of anticollagen type I antibody and anti-OPNantibody. An anti-b actin antibody was used as the loadingcontrol. The immune-complexes were visualized usingenhanced chemiluminescence reagent (Amersham) in ac-cordance with the manufacturer’s instructions.
In vivo study70% TCP–Ti composites and pure Ti sintered at 1200�Cwere used as specimens to be implanted in male NewZealand rabbits for 3 months. Implants were prepared in
ORIGINAL ARTICLE
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MAY 2013 VOL 101A, ISSUE 5 1491
rectangular block shapes with 3 mm � 2 mm � 0.5 mmdimensions. The implant surfaces were polished with 600#grit papers and washed ultrasonically for 10 min each in ac-etone, ethanol and deionized water, followed by a 2 himmersion in 75% ethanol. Subsequently, the implants weresterilized by autoclave at 121�C for 30 min. Adult white rab-bits weighing 3 kg were used as the animal model in thisstudy. Animals were treated in accordance with the Guide-lines of the local Ethical Committee. Under general anesthe-sia, the implants were inserted with a press-fit techniqueinto holes that had been drilled to the same diameter as theimplants at the metaphases of the rabbit femur after shav-ing. During the drilling, sufficiently sterilized physiologicalsaline solution was irrigated for cooling and cleaning. Thewound was closed with conventional suturing. The implan-tation was for 3 months, and each month, an X-ray photo-graph was taken to investigate the composite degradation.The animals were sacrificed by intravenous injections of airunder general anesthesia 3 months after surgery. Theimplant and surrounding tissue were removed en bloc bysawing and investigated by micro-computed tomography(CT). The specimens for Villanueva bone stain were fixed in70% ethanol in which Villanueva bone stain powder wasdissolved (Maruto Instrument Co., Tokyo, Japan), accordingto the method of Villanueva.26 Subsequently, the specimenswere dehydrated through a series of graded ethanol concen-trations, and embedded in methyl methacrylate (MMA,Sigma). The embedded specimens for Villanueva stain weresliced using a diamond saw blade into sections with a thick-ness of 500 lm, after which the sections were thinned to athickness of <30 lm by polishing with 9 lm, 3 lm, and 1
lm diamond particle bed subsequently. The thin specimenswere then observed under an optical microscope (Olympus,BX53, Japan).
Statistical analysisThe mechanical properties of the TCP–Ti composites weremeasured by using 16 pieces of identically shaped speci-mens from each composite. All data are expressed as mean6 standard deviation. Statistical analysis was performed inMicrosoft Office Excel, and the differences among groupswere analyzed by ANOVA followed by student’s t-test. pvalue <0.05 was considered significant.
RESULTS
Microstructure of the sintered compositesRepresentative FE-SEM images of spark plasma sintered Ti–TCP composites are shown in Figure 1. This figure showsthat with increasing amount of TCP, the interconnected po-rosity decreases. In 30% TCP–Ti composites [Fig. 1(a)], theTi particles and TCP particles are easily visible and differen-tiable. The 50% TCP–Ti composites [Fig. 1(b)] have similarmicrostructure to the 30% TCP–Ti composites, but TCP par-ticles are more distributed and cover the Ti particles, whichis characteristic of proper sintering. In 70% TCP–Ti compo-sites there is a drastic change in microstructure. The poros-ity is decreasing, and the particle interfaces are barely visi-ble. For better realization, backscattered SEM images arepresented in Figure 2. From Figure 2, in 30% TCP–Ti com-posites [Fig. 1(d)], various particles are visible and separa-ble, and TCP phases are clearly detectable. In 50% TCP–Ti[Fig. 2(b,e)], it is clear that large Ti particles are fully
FIGURE 1. FE-SEM images of (a and d) 30% TCP–Ti, (b and e) 50% TCP–Ti, and (c and f) 70% TCP–Ti composites. Black and white arrows are
indicating possible TCP and Ti phases, respectively.
1492 MONDAL ET AL. MICROSTRUCTURE AND BIOCOMPATIBILITY OF COMPOSITE BIOMATERIALS
layered by thick TCP layers degraded onto the particlesurfaces, but the interconnected porosity still remained [Fig.1(b,e)]. However, in 70% TCP–Ti [Fig. 2(c,f)] phases, Ti par-ticle interfaces are not visible, and it is difficult to separatethe different phases.
Phase characteristicsTo study the phase behavior and characteristic properties ofspark plasma sintered TCP–Ti composites, EDS, and XRDanalysis were carried out. SEM images and the EDS profileof specific zones are shown in Figure 3. From Figure3(a,d,e), 30% TCP–Ti composites showed pure Ti particlesand Ca/P phases on final composite. The 50% TCP–Ti com-posite [Fig. 3(b,f,g)] showed evidence of some dissociationoccurred because few amount of atomic P, Ca, and O on Tiparticle zones and some atomic Ti on the nanogranularzone were detected. However, the nanogranule coveredzones are still dominated by atomic Ca and atomic P. In con-trast, in 70% TCP–Ti composites [Fig. 3(c,h,i)], the particlesare mixed up, and the composite surfaces are dominated byatomic Ca. From XRD analysis (Fig. 4), the XRD profiles ofsintered TCP–Ti composites [Fig. 3(c,d,e)] in comparisonwith sintered pure TCP and Ti [Fig. 3(a,b)], 30% TCP–Ti hasa large amount of remaining Ti (PDF No. 44-1294) and pureTCP (PDF No. 55-0898) phase. However, a small amount ofCaTiO3 (PDF No. 1070-8503) formed here. In 50% TCP–Ti,the amount of CaTiO3 is increasing. While, in 70% TCP–Ticomposites, the dominating phase is CaTiO3, but the amountof remaining Ti and TCP phases are also significant. No cal-cium oxide or notable Ti oxide phases were detected amongall TCP–Ti composites. Little amount of TiP is detected infinal composite of 70% and 50% TCP–Ti.
Mechanical propertiesThe characteristic mechanical properties are represented inTable I. The relative density of sintered TCP–Ti compositesdecreases with increasing amount of TCP. Relative density ofTCP is quite lower (3.14 g cm�3) than cpTi (4.51 g cm�3). Soit is obvious that with increasing the amount of TCP, relativedensity of final composite will be decreased. However, it iscarefully noted that density value drastically dropped offwhile amount of TCP increased from 50 vol % to 70% com-pare to 30 vol % to 50%. The hardness value and compressivestrength also decreased in similar way for introducing higheramount of TCP. But, the Young’s modulus value increased withincreasing the TCP amount. The 70% TCP–Ti compositesshowed the highest Young’s modulus value of 3.23 GPa.
In vitro analysisTo evaluate the cytotoxicity, cell viability and cell attach-ment on the composites were determined through in vitroexperiments with MTT assay and by examining the morphol-ogy of surface-grown individual cells. Figure 5 representsthe MTT assay results performed with MG63 cell lines. Theabsorbance detected from cells cultured onto the surface ofspark plasma sintered TCP–Ti composites showed that withincreasing TCP percentage, the composites provided moreviable environment. Initially, after 1 day of cell seeding,30% TCP–Ti composite showed the best cell response. How-ever, after 3 and 7 days of cell culturing, the OD for 70%TCP–Ti composite increased very rapidly, which representsa high rate of cell growth on the composite surface and bet-ter cell viability compared with the other compositions.
MG63 osteosarcoma cell line attachment behavior for 30min of cell seeding on the TCP–Ti composites surfaces are
FIGURE 2. Backscattered SEM images of (a and d) 30% TCP–Ti, (b and e) 50% TCP–Ti, and (c and f) 70% TCP–Ti composites.
ORIGINAL ARTICLE
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MAY 2013 VOL 101A, ISSUE 5 1493
represented in Figure 6. All composites showed good cellattachment, but comparatively, the 70% TCP–Ti compositesurfaces provide the best viable environment, as shown in
MTT assay analysis, having cells with clearly visible filopo-dia attached to the specimen surface. With increasing cellseeding time, cells became more flattened, and clearly visi-ble filopodia were attached and proliferated on the compos-ite surface. In Figure 7, the results of cell attachment after60 min of cell culturing on the surfaces of TCP–Ti compositeare shown. The results are similar to cell seeding for 30min. Yet the difference is here, the cells are more regularlyshaped, well attached, and flattened on the surface of 70%TCP–Ti composite.
Cell proliferation behavior under immunofluorescencestaining is depicted in Figure 8. After 1 day of cell seeding,the cells proliferated similarly to the surface of all TCP–Ticomposites [Fig. 8(a–d)] under no significant effect by con-taining higher amount of TCP. But after 3 and 7 days ofcell seeding, cells proliferated in 70% TCP–Ti compositesurfaces were well structured with long filopodia stronglyattached to the surface, compared with those cells attachedonto the surface of 30% TCP–Ti, thus confirming the previ-ous MTT assay and cell attachment evaluation. Higher mag-nification images of cell proliferated on the surface of Tiand TCP–Ti composites after 7 days of cell seeding are
FIGURE 3. SEM images of (a) 30% TCP–Ti, (b) 50% TCP–Ti, and (c) 70% TCP–Ti composites with EDS profiles of area marked by squares at (d
and e) of 30% TCP–Ti, (f and g) of 50% TCP–Ti, and (h and i) of 70% TCP–Ti composites surfaces. Tables attached in the bottoms of EDS profiles
are showing atomic percentage of elements present at those zones. The interaction depth is 100 nm. EDS was performed and analyzed with
INCAEnergy, Oxford.
FIGURE 4. XRD profiles of TCP–Ti composites with pure Ti and TCP
powders. (a) cp Ti, (b) pure TCP, (c) 30% TCP–Ti, (d) 50% TCP–Ti, and
(e) 70% TCP–Ti. Peaks are marked by (~) TCP, (n) Ti, (^) CaTiO3, and
unmarked peaks are for TiaPb. Peaks are detected by (PDXL software,
JCPDS-ICDD).
TABLE I. Mechanical Properties of TCP–Ti Composites
1494 MONDAL ET AL. MICROSTRUCTURE AND BIOCOMPATIBILITY OF COMPOSITE BIOMATERIALS
shown in Figure 9. All TCP–Ti composites showed excellentextended F-actin expression. While cells seeded on 50%[Fig. 9(c)] and 70% [Fig. 9(d)] TCP–Ti composites arestrongly attached and having higher amount of clearly dis-tinguishable bundles of F-actin. The expression of OPN andcollagen type I protein from osteoblast cells cultured onthe surface of the composites were evaluated by westernblot. Osteoblast MG63 cell line was cultured for 3 and 7days on the surfaces of all TCP–Ti composites along withpure Ti. Figure 10 represents the protein expression fromthe cultured cells after 3 and 7 days. Cells seeded on thesurface of pure Ti for 3 days showed higher expression ofboth OPN and collagen type I proteins than other TCP–Ticomposites. However, after 7 days of cell culturing, theTCP–Ti composites showed remarkably high expression of
OPN and collagen, whereas for the pure Ti, the proteinexpression did not change much.
In vivo analysisAll rabbits tolerated the operation well and stayed aliveuntil the sample harvest. No inflammation or purulence wasobserved. X-ray observation after 1, 2, and 3 months of im-plantation showed no dislocation of the implants (Fig. 11).The micro-CT images (Fig. 12) of the implantation showedthat the growth of new bone occurred preferentially at thesurface of the composite. Compared with pure Ti implant,the 70% TCP–Ti composites showed better degradation andbone formation. The histological analysis of pure Ti and70% TCP–Ti implants in vivo after staining with VillanuevaOsteochrome Bone Stain and embedded into MMA areshown in Figure 13. From Figure 13(a,c), no immunologicalreaction occurred at the bone implant interface of pure Ti,but an insignificant amount of bone formation occurredthere. In contrast, 70% TCP–Ti in vivo [Fig. 13(b,d)] showsexcellent bone–implant interaction and formation of newbone which can be detected by light pink color zone. Afterimplantation, surface voids are filled up with new bone. Adark zone in Figure 13(d) surrounded by light pink zoneindicates degradation of the composite and formation ofbone–implant matrix.
DISCUSSION
The above results have shown that 70% TCP–Ti compositecould degrade both in vitro and in vivo with having excellentbiocompatibility. It may fulfill the demand for sufficient
FIGURE 5. MTT assay analysis of TCP–Ti composites by using MG63
cell cultured on the composites surfaces.
FIGURE 6. SEM images of MG63 cell attachment on the surfaces of (a and d) 30% TCP–Ti, (b and e) 50% TCP–Ti, and (c and f) 70% TCP–Ti com-
posites after 30 min of cell seeding. (d), (e), and (f) are higher magnified images from circularly marked zones of (a), (b), and (c), respectively.
Arrows are showing attached filopodia on specimen surfaces.
ORIGINAL ARTICLE
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MAY 2013 VOL 101A, ISSUE 5 1495
FIGURE 7. SEM images of MG63 cell attachment on the surfaces of (a and d) 30% TCP–Ti, (b and e) 50% TCP–Ti, and (c and f) 70% TCP–Ti com-
posites after 60 min of cell seeding. (d), (e), and (f) are higher magnified images from circularly marked zones of (a), (b), and (c), respectively.
Arrows are showing attached filopodia on specimen surfaces.
FIGURE 8. Confocal micrographs of MG63 cell attachment and proliferation on the surfaces of (a, e, and i) 0% TCP–Ti (pure cpTi), (b, f, and j)
30% TCP–Ti, (c, g, and k) 50% TCP–Ti, and (d, h, and l) 70% TCP–Ti composites after (a–d) 1 day, (e–h) 3 days, and (i–l) 7 days of MG63cell seed-
ing. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
mechanical strength for reducing stress shielding effect andsupporting the system during formation of bone–implantmatrix. Although the system here also goes through reac-tion, compare with HAp/Ti composites, having such amountof CaP and Ti phases remaining here is a development forTi/CaP-based hard tissue implant. However, further modifi-
cations should be needed to improve the composites by in-hibiting complex reactions. The SEM images and EDS pro-files of the composites (Figs. 1–3) are clearly evident oflarge particles of Ti that are surrounded by TCP particles inthe final composite bodies. With increasing the amount ofTCP, Ti particles are gradually covered by TCP layers. In the
FIGURE 9. Confocal micrographs of MG63 cell attachment on the surfaces of (a) Ti, (b) 30% TCP–Ti, (c) 50% TCP–Ti, and (d) 70% TCP–Ti compo-
sites after 7 days of cell seeding. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
FIGURE 10. The expression of osteopontin and collagen type 1 protein of MG63 cells for control and different compositions of TCP–Ti compo-
sites after (a) 3 days and (b) 7 days of cell culture. b-Actin levels are used as internal control.
FIGURE 11. X-ray photographs of implanted 70% TCP–Ti composites on rabbit femur after (a) just implanted, (b) 1 month, (c) 2 months, and (d)
3 months. Position of the implant is marked by white circles.
ORIGINAL ARTICLE
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MAY 2013 VOL 101A, ISSUE 5 1497
30% TCP–Ti composite, the particles are easily separable[Figs. 1(d) and 2(d)]. No surface degradation occurred inthis composition after sintering [Fig. 3(a,d)]. But in 70%TCP–Ti composites, the particle interfaces are diminishedand the phases are very difficult to recognize. It is clearfrom XRD profiles (Fig. 4) that 50% TCP–Ti composites con-tain a large amount of TCP phases which are deposited onthe surfaces of Ti particles. But in the 70% TCP–Ti compos-ite, there is no such TCP layer or extremely distinguished Tiand TCP phases (Fig. 2) as it is also clear from atomic per-centage in EDS analysis at various zone [Fig. 3(h,i)]. BothEDS profiles and XRD analysis showed that the 30% TCP–Ticomposite contained a high amount of TCP and pure Tiphase inside. Earlier studies with Ti/HAp composites ended
up with decision that even 30% HAp–Ti composite was notstable and HAp dissociated into various reaction prod-ucts.12–14 Without a hydroxyl radical, TCP does not enhanceTi to form TiO2, and thus, TCP is not vigorously dissociatedinto other products. The possible explanation would be withhigh temperature sintering, while Ti becomes very reactiveand some amount of TCP breaks into CaO and P2O5 in pres-ence of Ti. Some P2O5 then reacts with Ti to form titaniumphosphides (TiaPb) and TiO2, which later react with CaO toform CaTiO3 and rest of the P2O5 might be evaporated. Thepresence of a few amount of TiP [Fig. 4(e)] and loweratomic percentage of P [Fig. 3(i)] in final products are sup-porting this point. With increasing amount of TCP, the reac-tion rate also increases, and in the 70% TCP–Ti composite,
FIGURE 12. 2D micro-CT images of bone segments after 3 months of implantation of (a) pure Ti and (b) 70% TCP–Ti composite. (c) is 3D micro-
CT image of bone segments for pure Ti in vivo and (e) is the enlarged view of implantation zone. (d) and (f) are similar images for 70% TCP–Ti
composite in vivo. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
1498 MONDAL ET AL. MICROSTRUCTURE AND BIOCOMPATIBILITY OF COMPOSITE BIOMATERIALS
the dominating phase becomes CaTiO3. Still, in 70% TCP–Ticomposite, a significant amount of TCP and pure Ti phaseremain, and that is a significant improvement for Ti/CaP-based composites compared with Ti/HAp composites fabri-cated earlier.
The mechanical properties of the final composites areshown in Table I. For sintered composites, the relative den-sity of sintered TCP–Ti composites is decreasing withincreasing the amount of TCP. This is because the TCP isreacted with Ti, and the major reaction product CaTiO3 haslower relative density (3.35 g cm�3) than Ti. With theincreased amount of TCP, the regular porous structure isgone, and the amount of CaTiO3 also increased which helpsto decrease the hardness and compressive strength values.20
In contrast, for 50% TCP–Ti composite, Young’s modulus is>30%, because a higher amount of remaining Ti and TCPimproved the composite stiffness. For similar reason, 70%TCP–Ti expressed the highest modulus value. Still, the com-pressive strength and hardness values for 70% TCP–Ti arelaid down in the range of cortical bone mechanical proper-ties. The Young’s modulus and compressive strength valuefor cortical bone27 are in the range of 3–30 GPa and130–180 MPa, respectively, whereas 70% TCP–Ti showed amodulus and compressive strength value of 3.23 GPa and
�388.5 MPa, respectively. Based on the mechanical proper-ties of commercial pure Ti or Ti alloy-based dentalimplants28,29 and the stem part of artificial hip-joints,30,31
the spark plasma sintered 70% TCP–Ti composite would bean excellent candidate for use as orthopedic implants.
For a biomaterial, good cell viability is one of the mostimportant criteria. MTT assay for the fabricated materialsshown in Figure 5 revealed that all the compositions exhib-ited good cell viability after 1, 3, and 7 days of cell culturing.From MTT assay analysis with MG63 cell line, 70% TCP–Ti-based composite showed the highest cell viability comparedwith other compositions. From the XRD data, it could be seenthat with increasing TCP, the amount of CaTiO3 alsoincreased, but a significant amount of TCP and Ti stillremained. From the MTT assay graph, the cell compatibilityalso increased with increasing initial TCP concentration.Research has shown that CaTiO3 is a bioactive materialbecause of its dielectric properties, and it enhances osteo-blast adhesion and growth.32,33 This infers that even afterreaction occurred, the final composites remained biocompati-ble. The interface between the implant and the host bone isone of the most important issues for biomaterials used inhard tissue repair applications. The bonding type is consid-ered to be a vital criterion for evaluating the biocompatibility
FIGURE 13. Optical micrographs of bone–implant interfaces of (a) pure Ti and (b) 70% TCP–Ti after 3 months in vivo and stained with Villanueva
Stain. (c) and (d) are enlarged images from (a) and (b), respectively. [Color figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
ORIGINAL ARTICLE
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MAY 2013 VOL 101A, ISSUE 5 1499
and bioactivity of biomaterials. When the morphology of indi-vidual cells that had been incubated with the composites for30 min was analyzed, all composites displayed good cellattachment behavior (Fig. 5). On the pure Ti surface, cellscannot proliferate in such a short time.34 In this study, theMG63 cell line was used for cell attachment analysis. Thecells were bound to the composite surfaces and proliferatedwell. Cell attachment depends on the topography of the spec-imen surface. Rough and irregular surfaces are better for cellattachment. After 60 min of cell seeding, cell morphologychanged for 70% TCP–Ti remarkably, compared with 30%and 50% TCP–Ti. Cells on 70% TCP–Ti showed a flattenedand large shape may be for the presence of biocompatibleand bioactive TCP, Ti, and CaTiO3. The filopodial activities forTCP–Ti composites were the same (Fig. 6). But in the case of70% TCP–Ti composites [Fig. 7(c)], cells started to divideinto new daughter cells and proliferated. The filopodia ofcells in 70% TCP–Ti-based composite surface were mostelongated compared with the other TCP–Ti composites. Cellproliferation, evaluation by immunofluorescence staining(Fig. 8), showed that osteosarcoma MG63 cells attached andproliferated on the surfaces of 70% TCP–Ti composites betterthan any other TCP–Ti composites and Ti. Cell attachmentdepends on surface morphology and ion content of the speci-men surface. TCP biodegrades similarly at the rate of bonemodeling.35,36 Due to its high water solubility which enhancedissolution of Ca and P in tissue fluid and absorption byosteoclasts in vivo.21–23,35,36 Thus, the bioactivity also becamehigher with increasing TCP content, even though someamount of TCP was going through complex reactions. Forma-tion of CaTiO3 is also responsible for this high bioaffinity. Af-ter 7 days of cell culturing on the 70% TCP–Ti composite sur-face, the surface was almost covered with cells. Cell filopodiaare longer and more strongly attached to the 70% TCP–Ticomposite surface [Fig. 8(d,h,l)] compared with other compo-sites. The assessment of the F-actin cytoskeleton of MG63cultured on the 70% TCP–Ti composite surface showed ahigh level of actin organization [Fig. 9(d)]. The remaining Tiand TCP phases and the formation of CaTiO3 help the 70%TCP–Ti composite to become more viable and compatiblewith the osteosarcoma cell line.
A crucial factor for orthopedic implants is the rapidbone ingrowth by quick cellular acceptance of the surround-ing tissue after implantation. High expression of ECM pro-teins (such as OPN and collagen type I) induces furtherosteoblastic cell adhesion and proliferation. The proteinexpression from western blot (Fig. 10) showed that a highlevel of OPN was expressed on the 70% TCP–Ti compositewithin 3–7 days of cell culturing on its surface. This fastexpression of ECM proteins collagen and OPN refers to thefast growing of osteoblast cells around the composite sur-face. The higher rate of protein expression for TCP–Ti com-posites compared with pure Ti indicates that the 70% TCP–Ti composite is more osteoinductive than pure Ti implant.
Having favorable strength, hardness, elasticity, and suchlevel of bioactivity with osteoblast cell line, 70% TCP–Ticomposite showed much potential as an orthopedic implant,which was encouraging for studying in vivo bone formation
in an animal body. After implantation in a rabbit femur for3 months, the 70% TCP–Ti composite and pure Ti implantsdid not show any detrimental effect. After 3 months of im-plantation, the 3D micro-CT images indicated that new bonewas formed around and inside the implants, which is con-sistent with the in vitro analysis. Histological observation(Fig. 13) showed that degradation of the TCP–Ti compositeand new bone formation occurred at the bone–implantinterface. Compared with the pure Ti implant, the TCP–Ticomposite showed better bone–implant bonding andingrowth of bone tissue inside the implant within 3 monthsof implantation, which would have resulted in a successfulbone–Ti matrix. The result clearly shows that the presenceof CaTiO3 and Ca3(PO4)2, bone mineralization and bone–implant bonding to the native bone tissue occurred veryfast in the animal body with the TCP–Ti composites. Addi-tionally, in the presence of Ti, the mechanical strength alsowould not be compromised.
CONCLUSIONS
A 30 vol %, 50 vol %, and 70 vol % TCP with Ti compositeswere prepared by SPS at 1200�C. The results showed thatcomplex reactions occurred between Ti and TCP during sin-tering due to highly reactive Ti at elevated temperature, anddissociated some amount of Ca3(PO4)2 into CaTiO3.Although CaTiO3 was formed after being sintered at 1200�C,significant amounts of Ti and TCP phases remained. No CaOphase was detected in the final composites. The mechanicalproperties were good enough to apply as human hard tissuereplacements, especially the 70% TCP–Ti composite, whichhad Young’s modulus value in between the modulus valueof natural cortical bone. The biocompatibility of the TCP–Ticomposites was enhanced with increasing TCP amount andpresence of CaTiO3, as evident by increased cell adhesion,proliferation rate and protein expression, and by in vivobone formation. The TCP–Ti composites listed in the orderof bioactivity: 70% TCP > 50% TCP > 30% TCP > pure Ti.Among the fabricate TCP–Ti composites, the 70% TCP–Tienhanced the expression of osteogenic proteins and pro-duced excellent improvement of osteogenesis comparedwith the pure Ti orthopedic implant. In vivo study alsoshowed that within 3 months after implantation in an ani-mal body, the 70% TCP–Ti composite can biodegrade toform new metal–bone matrix, which strongly suggests thatthe composite could be a perfect orthopedic implant mate-rial for guided bone tissue regeneration under load-bearingconditions.
ACKNOWLEDGMENT
This work supported by Mid-career Researcher Programthrough NRF grant funded by the MEST (NO 2009-0092808).
REFERENCES1. Geetha M, Singh AK, Asokamani R, Gogia AK. Ti based biomateri-
als, the ultimate choice for orthopaedic implants—A review. Prog
Mater Sci 2009;54:397–425.
2. Robertson DM, St. Pierre L, Chahal R. Preliminary observations of
bone ingrowth into porous materials. J Biomed Mater Res 1976;
10:335–344.
1500 MONDAL ET AL. MICROSTRUCTURE AND BIOCOMPATIBILITY OF COMPOSITE BIOMATERIALS
3. Head WC, Bauk DJ, Emerson RH Jr. Titanium as the material of
choice for cementless femoral components in total hip arthro-
plasty. Clin Orthop Relat Res 1995;311:85–90.
4. Krishna BV, Bose S, Bandyopadhyay A. Low stiffness porous Ti
structures for load-bearing implants. Acta Biomater 2007;3:
997–1006.
5. Kong Y, Kim S, Kim H, Lee I. Reinforcement of hydroxyapatite
bioceramics by addition of ZrO2 coated Al2O3. J Am Ceram Soc
1999;82:2963–2968.
6. Li J, Hermansson L, Soremark R. High strength biofunctional zir-
conia: Mechanical properties and static fatigue behaviour of zirco-
nia–apatite composite. J Mater Sci Mater Med 1993;4:50–54.
7. Kokubo T, Matsushita T, Takadama H. Titania-based bioactive
materials. J Eur Ceram Soc 2007;27:1553–1558.
8. Vehof JWM, Spauwen PHM, Jansen JA. Bone formation in cal-
