Mechanical properties of hydroxyapatite ceramics sintered from powders with different morphologies
Jiangfeng Song,Yong Liu,Ying Zhang,Lei Jiao
DOI: https://doi.org/10.1016/j.msea.2011.03.078
2011-01-01
Abstract:In the present work, hydroxyapatite (HA) nanorods and submicron prisms were synthesized by using PVP (polyvinylpyrrolidone) and CTAB (cetyltrimethylammonium) as templates respectively, and then sintered at 1150–1250 °C. The mechanical properties of as-sintered ceramics were studied. The results show that the HA ceramics sintered from prism-like powders have a Vickers hardness of ∼5.11 GPa, a transverse rupture strength of ∼106.26 MPa, and a fracture toughness of ∼1.24 MPa m 1/2 , which are close to those properties of natural teeth. Moreover, it reveals that HA ceramics sintered from prism-like powders exhibit a higher density and mechanical properties than that sintered from HA nanorods, owing to the less agglomeration of raw powders and lower mass loss during the sintering process. Keywords Mechanical properties Hydroxyapatite ceramics Morphology Pressureless sintering 1 Introduction Hydroxyapatite (HA, Ca 10 (PO 4 ) 6 (OH) 2 ) is the main inorganic phase of human bone and teeth [1,2] . Synthetic HA has a wide range of biomedical applications because it is bioactive and biocompatible [3] . However, it differs significantly in terms of its micro- and macrostructure from its biological equivalent – cortical bone [4] . The difference of structures leads to the poor mechanical properties of synthetic HA, such as the brittleness and low fracture toughness, which can restrict its applications in non-load bearing or metallic implant surfaces coating [5,6] . Different approaches have been extensively explored to improve the mechanical properties of HA ceramics, such as adding dopants, making composites, and controlling microstructures via novel sintering techniques or utilizing nanostructured powders [1] . Organic surface can control the nucleation of inorganic materials by geometric, electrostatic, and stereochemical complementarity between the incipient nuclei and functionalized substrates [7] . For example, cetyltrimethylammonium bromide (CTAB) is widely used to control the morphology in many aqueous synthetic methods [8] . In an aqueous system, CTAB can ionize completely, resulting in a cation with a tetrahedral structure. Meanwhile, the phosphate anion also has a tetrahedral structure. The charge and structural complementarity endow CTAB with the capability to control the crystallization process [7] . Additionally, polyvinylpyrrolidone (PVP) has also been widely used as a surface-regulating polymer for synthesis of inorganic nanomaterials, especially for the architecture of one dimensional (1D) nanostructure. In the present work, PVP is introduced because of the imide (N–C O) group in the polymer structure, which also exists in collagen molecular structures [9] . The mechanical properties and microstructures of the resulting HA ceramics highly depend on the characteristics of the original HA powder, including crystallinity, agglomeration, stoichiometry, and substitutions [10] . However, so far, the effect of morphology on the sinterbility and mechanical properties of HA ceramics has not been fully investigated yet. Pressureless sintered HA ceramics often decompose into tricalcium phosphate and tetracalcium phosphate at high temperatures during the long sintering period. And also, pressure-assisted sintering of HA ceramics will also result in the decomposition of HA [11] , and introducing pressure extrinsically will bring about many limitations, from technical and economical considerations [12] . However, changing the microstructure of HA intrinsically can prevent the decomposition of HA ceramics during the pressureless sintering procedure [1,6,12] . In this study, HA powders with different morphologies were used to prepare dense and pure HA ceramics under pressureless sintering, and the mechanical properties were evaluated. 2 Materials and methods 2.1 Synthesis of HA powders HA powders were synthesized by using calcium nitrate (Ca(NO 3 ) 2 ·4H 2 O) and ammonium dihydric phosphate (NH 4 H 2 PO 4 ) as reagents. The PVP and CTAB were used as templates. All the chemicals were of analytical grade and purchased from Alfa Asear. Firstly, the calcium solution and phosphate solution were prepared by dissolving 1.0 mol Ca(NO 3 ) 2 ·4H 2 O and 0.6 mol NH 4 H 2 PO 4 in 2000 ml distilled water, respectively. Then, 58 g of PVP or 3.64 g of CTAB was added into the calcium solution and the pH was adjusted to ∼14 by the addition of concentrated ammonia solution after the reagents was dissolved completely. Then the phosphate solution was added dropwise to the above-mentioned solution under vigorously stirring at room temperature. The obtained suspension was aged for 72 h and then filtrated, washed with distilled water and ethyl alcohol alternately several times. The white precipitate was subsequently dried in vacuum at 80 °C for 12 h to obtain HA powder. For comparison, the PVP and CTAB templated powders were marked as HA-PVP and HA-CTAB, correspondingly. 2.2 Sintering of hydroxyapatite ceramics HA ceramics were prepared by traditional pressureless sintering method. Firstly, 1 wt.% polyvinyl alcohol (PVA) solution were mixed thoroughly with the HA powder. Then, the mixed powders through 100 mesh sieve were biaxially compressed to discs with a diameter of 12 mm under the pressure of 100 MPa for Vickers hardness and fracture toughness testing, or cubic shapes with a size of 50 mm × 10 mm × 7 mm for transverse rupture strength testing. In order to investigate the effect of the grain size on mechanical prosperities of HA ceramics, HA ceramics were sintered at 1150–1250 °C for 2 h under ambient atmosphere, with a heating rate of ∼10 °C/min, and then furnace cooled to minimize internal cracks [13] . 2.3 Microstructure characterization The as-synthesized HA powders were characterized by X-ray diffraction (XRD; D/ruax 2550PC), thermalanalyzer (TGA and DTA; SETSYS EVO18/24) and field emission transmission electron microscope (FETEM; JEOL JEM-2100F). XRD was also performed to investigate the phase change and decomposition of HA in sintered bodies synthesized at different temperatures. The microstructure was observed using the field emission scanning electron microscope (FESEM; NOVA NANOSEM 230). The samples were polished to a 1-μm finish, and then chemically etched for 3 min using 10 vol.% hydrochloric acid. The average grain size was determined from FESEM images using the quantitative metallographic method. The density was measured by Archimedes’ method. 2.4 Characterization of mechanical properties Sintered HA discs were polished to a 1-μm finish before the measurement of Vickers hardness and indentation toughness. The Vickers hardness measurement was performed using a hardness tester with 200 g load. The Vickers indentation was used to estimate the toughness of sintered HA ceramics via generating cracks at the four corners of the Vickers indenter. The toughness value was calculated by the following equation [14–16] : (1) K C = ξ E H 1 / 2 P C 3 / 2 where H is the Vickers hardness; E is the elastic modulus; P is the indentation load and C is the indentation crack length. The value of the non-dimensional constant ξ is empirically determined to be 0.016 ± 0.004. The parameters C and H were measured for each sample condition, and E was assumed to be 115 GPa for all the samples [1] . The hardness and fracture toughness values were averaged of 5 valid measurements for each group. The transverse rupture strength was tested by the three-point bending method on a span of 25 mm at a crosshead speed of 0.5 mm/min. 3 Results 3.1 Microstructures Figs. 1 and 2 compare the XRD patterns of as-synthesized HA powders and as-sintered HA ceramics at different temperatures using HA-PVP and HA-CTAB powders, respectively. For both kinds of raw HA powders, all the peaks match the JCPDS pattern 09-0432, and show no trace of decomposition. Both the as-synthesized HA powders show low crystallinity, whereas the corresponding sintered ceramics have relatively sharp peaks. Fig. 3 corresponds to the thermogravimetric analysis (TGA) and differential thermal analysis (DTA) of the as-synthesized HA powders. The total mass losses of HA-PVP and HA-CTAB are 9.39% and 5.47%, respectively after heating up to 1200 °C. The first stages of the weight loss in both samples up to 100 °C associate with the removal of adsorbed water humidity. The other significant loss appeared at around 900 °C, may be due to the evaporation of carbon monoxide CO of the polymer [17] . According to the TG curves of both HA powders, the major difference of the mass loss appeared at the temperature up to 500 °C. The mass losses of HA-PVP and HA-CTAB are 6.89% and 3.42%, respectively. The as-synthesized raw HA-PVP powders have a morphology of irregular nanorods with an average length of ∼100 nm and a mean diameter of ∼20 nm, as shown in Fig. 4 (a) . Chemically etched cross-sections of the HA ceramics sintered at different temperatures are shown in Fig. 4 (b)–(f). For comparison, the morphology of HA-CTAB powders are shown in Fig. 5 (a) , and Fig. 5 (b)–(f) presents the evolution of the etched cross-sections of sintered HA ceramics. It was found that the samples consist of regular prism-like submicron powders with an average length of ∼200 nm and a mean width of ∼100 nm, have a high crystallinity. The grain sizes of all samples measured by the line intercept method are listed in Table 1 . For HA-PVP, the HA ceramics sintered at 1150 °C have an average grain size of 0.78 μm, and gradually increase to 2.16 μm when the sintering temperature rises to 1250 °C. While for HA-CTAB, it increases from 0.82 μm to 2.32 μm. Obviously, increasing sintering temperature leads to grain growth as well as the densification process [2,18,19] . 3.2 Vickers hardness The relationship between the Vickers hardness and the sintering temperature is presented in Fig. 6 . For both kinds of ceramics, the Vickers hardness increases at first, and then decreases with the sintering temperature. It is resulted from the difference in the density. Porosity is the most significant factor for determining the hardness of sintered ceramics [20,21] , and the hardness increases with the relative density. It is noted that the hardness in this study is comparable with those reported previously for pure HA ceramics (1.0–5.5 GPa) [22–24] . The highest Vickers hardness of HA-CTAB ceramics is 5.11 GPa at 1200 °C, while for HA-PVP, the peak value is 4.37 GPa at 1230 °C. In this paper, the hardness of HA ceramics decreases with the porosity, as presented in Fig. 7 . Additionally, it is shown that HA ceramics follow the model which proposed by Rice [25] with the following formula: (2) ln H = ln H 0 − b P where H is the measured Vickers hardness, H 0 is the zero porosity hardness, b is the decay constant, P is the porosity. The calculated zero porosity H 0 = 5.58 ± 1.74 GPa, b = 4.30. 3.3 Fracture toughness The fracture toughness of HA ceramics as a function of the grain size is shown in Fig. 8 . According to Table 1 and Fig. 8 for HA-CTAB, when the relative density is comparable (96.11–95.90%), the fracture toughness decreases with the grain size, showing the same trend as those reported in literatures [1,10,11] . Specifically, in this density range, the maximum value is 0.99 MPa m 1/2 at 1.02 μm sintered at 1180 °C. As the grain size decreases, the crack deflection increases, which leads to continuous increase in the toughness [1] . Additionally, the fracture toughness of as-sintered HA-PVP ceramics show no obvious change trend of grain size, maybe because of both the relative density and grain size change which have a complicated influence on the toughness. The highest fracture toughness achieved in this study is 1.24 MPa m 1/2 for HA-CTAB sintered at 1150 °C, 1.20 MPa m 1/2 for HA-PVP sintered at 1200 °C. The whole results of the present study are equivalent to those in Ref. [1] . 3.4 Transverse rupture strength (TRS) Fig. 9 displays the relationship between the transverse rupture strength and the sintering temperature. It reveals that the transverse rupture strength increases at first and then decreases with the sintering temperature, similar to the Vickers hardness. For the density is a key factor of the strength. The maximum transverse rupture strength is 106.26 MPa for CTAB-templated HA ceramics sintered at 1200 °C, similar to those reported in the literatures (30–101 MPa) [13,26,27] . 4 Discussion In a basic aqueous system, both PVP and CTAB will ionize. The Ca 2+ will selectively adhere to the imide (N–C O) group in PVP polymer by the charge complementarity, as shown in Fig. 10 stage I for PVP. And then, it will combine with (PO 4 ) 3− and OH − , initiating the nucleation of nanometer-sized HA particles on organic polymer [17] , as presented in Fig. 10 stage II for PVP. Meanwhile, for CTAB, as a cationic surfactant, it will ionize completely and result in the cation with tetrahedral structure. (PO 4 ) 3− also has a tetrahedral structure. Thus the ionized CTAB can be well incorporated to (PO 4 ) 3− by the charge and structure complementarity, as presented in Fig. 10 stage I for CTAB. Then at a later stage, (PO 4 ) 3− will adsorb Ca 2+ and OH − , and finally form original hydroxyapatite crystals on the CTAB molecular ( Fig. 10 stage II for CTAB). It is worth mentioning that there are a lot of imide (N–C O) groups in one PVP, meaning that the PVP molecular will provide many sites for adhesion and nucleation, while one CTAB molecular only can offer one nucleation site. It is obvious that PVP has much more nucleation sites than CTAB. Then, the original number of HA crystals is larger for PVP ( Fig. 10 stage III). For both templates, the quantity of reactants is equal. Consequently, the PVP templated HA particles show a smaller size and lower crystalline integrity ( Fig. 10 stage IV). It is interesting to find that the strength of as-sintered HA-CTAB ceramics is remarkably higher than HA-PVP, including Vickers hardness and transverse rupture strength, due to a higher density. This phenomenon is abnormal, since the HA-PVP powders have a higher specific surface area, which will increase the driving force for densification during sintering, and normally result in higher mechanical properties [6] . The unusual phenomenon may be due to the following two reasons. Firstly, it is easier for HA-PVP powders to agglomerate than HA-CTAB powders due to their higher specific surface area. Additionally, water vapor will more likely to condense between small size particles, which will lead to the formation of chemical bond and acceleration of agglomeration [28] . Normally, agglomerated ceramic powders exhibit a low sinterbility [29,30] , due to the formation of pores and inhomogeneous microstructures. Secondly, the mass loss of HA-PVP (9.39%) is higher than that of HA-CTAB (5.47%) during the sintering procedure, which may retard the densification behavior, as presented in Fig. 3 . This mass loss is attributed to the removel and/or dissociation of the components of the polymer [31] , including PVA and the residual templates. And the total mass of residual template PVP is higher than that of CTAB, because the addition of PVP is much more than CTAB. Thus, the low agglomeration and mass loss in the sintering of HA-CTAB powders induce a higher density and mechanical properties. Considering the mechanical properties of teeth enamel (transverse rupture strength: 76 MPa [32] , hardness: 3–5 GPa, fracture toughness: 0.6–1.5 MPa m 1/2 [33] ), the HA ceramics sintered from HA-CTAB powders have a potential application in artificial teeth. 5 Conclusions In this study, two kinds of HA powders with different morphologies were synthesized using PVP and CTAB as templates, respectively. The HA powders templated by PVP are irregular nanorods with an average length of ∼100 nm and a mean diameter of ∼20 nm. While the HA powders templated by CTAB are regular prism-like with an average length of ∼200 nm and a width of ∼100 nm, and a high crystallinity. All the pressureless sintered ceramics show no decomposition even at the sintering temperature as high as 1250 °C. The Vickers hardness of HA ceramics decreases with the porosity. The hardness and the transverse rupture strength of as-sintered HA-CTAB ceramics is remarkably higher than that of HA-PVP ones. Even though HA-PVP powders have a higher specific surface area, it exhibits a lower sinterbility due to the agglomeration and high mass loss during the sintering process. The ceramics sintered from CTAB-templated HA powders have a potential application in artificial teeth, with a Vickers hardness of ∼5.11 GPa, a transverse rupture strength of ∼106.26 MPa, and a fracture toughness of ∼1.24 MPa m 1/2 . Acknowledgements The authors are grateful for the financial support from National Natural Science Foundation of China (NO. 50823006) References [1] J. Wang L.L. Shaw Biomaterials 30 2009 6565 6572 [2] L.-H. He O.C. Standard T.T.Y. Huang B.A. Latella M.V. Swain Acta Biomater. 4 2008 577 586 [3] C.Y. Tang P.S. Uskokovic C.P. Tsui D. Veljovic R. Petrovic D. Janackovic Ceram. Int. 35 2009 2171 2178 [4] A.J. Ruys M. Wei C.C. Sorrell M.R. Dickson A. Brandwood B.K. Milthorpe Biomaterials 16 1995 409 415 [5] M. Mazaheri M. Haghighatzadeh A.M. Zahedi S.K. Sadrnezhaad J. Alloys Compd. 471 2009 180 184 [6] J. Wang L.L. Shaw Adv. Mater. 19 2007 2364 2369 [7] W.-J. Shih M.-C. Wang M.-H. Hon J. Cryst. Growth 275 2005 e2339 e2344 [8] M. Salarian M. Solati-Hashjin S.S. Shafiei R. Salarian Z.A. Nemati Ceram. Int. 35 2009 2563 2569 [9] Y. Zhang J. Lu Cryst. Growth Des. 8 2008 2101 2107 [10] D. Veljovic B. Jokic R. Petrovic E. Palcevskis A. Dindune I.N. Mihailescu D. Janackovic Ceram. Int. 35 2009 1407 1413 [11] Y.W. Gu N.H. Loh K.A. Khor S.B. Tor P. Cheang Biomaterials 23 2002 37 43 [12] H.-W. Kim Y.-M. Kong Y.-H. Koh H.-E. Kim H.-M. Kim J.S. Ko J. Am. Ceram. Soc. 86 2003 2019 2026 [13] S. Pramanik A.K. Agarwal K.N. Rai A. Garg Ceram. Int. 33 2007 419 426 [14] G.R. Anstis P. Chantikul B.R. Lawn D.B. Marshall J. Am. Ceram. Soc. 64 1981 533 538 [15] H. Miyazaki H. Hyuga Y.-i. Yoshizawa K. Hirao T. Ohji Ceram. Int. 35 2009 493 501 [16] S.W.K. Kweh K.A. Khor P. Cheang Biomaterials 21 2000 1223 1234 [17] A.A. Mostafa H. Oudadesse M.B. Mohamed E.S. Foad Y. Le Gal G. Cathelineau Chem. Eng. J. 153 2009 187 192 [18] S. Ramesh C.Y. Tan S.B. Bhaduri W.D. Teng I. Sopyan J. Mater. Proc. Technol. 206 2008 221 230 [19] J. Wang L.L. Shaw J. Am. Ceram. Soc. 93 2010 601 604 [20] S. Best W. Bonfield Mater. Med. 5 1994 516 521 [21] S.V. Dorozhkin Biomaterials 31 2010 1465 1485 [22] K.A. Gross K.A. Bhadang Biomaterials 25 2004 1395 1405 [23] S.-J. Kim H.-G. Bang J.-H. Song S.-Y. Park Ceram. Int. 35 2009 1647 1650 [24] S. Bose A. Banerjee S. Dasgupta A. Bandyopadhyay J. Am. Ceram. Soc. 92 2009 323 330 [25] R.W Rice J. Mater. Sci. 31 1996 102 118 [26] I. Mobasherpour M. Solati Hashjin S.S. Razavi Toosi R. Darvishi Kamachali Ceram. Int. 35 2009 1569 1574 [27] A. Rapacz-Kmita A. ́ Sĺosarczyk Z. Paszkiewicz J. Eur. Ceram. Soc. 26 2006 1481 1488 [28] L. Jiang C. Caifeng c. zhigang Ceram. Eng. 2 2001 45 47 [29] G. Xingyuan G. Julie E. X. Ping L. Jing S. Zhenjian J. Biomed. Res. Part A 21 2007 1022 1032 [30] J.L. SHI J.H. GAO Z.X. LIN D.S. YAN J. Mater. Sci. 28 1993 342 348 [31] C. Rey C. Combes C. Drouet A. Lebugle H. Sfihi A. Barroug Mat. Wiss. U Werkstofftech. 38 2007 996 1002 [32] E.S Ahn N.J. Gleason A. Nakahira J.Y. Ying Nano Lett. 1 2001 149 153 [33] D. Bajaj A. Nazari N. Eidelman D.D. Arola Biomaterials 29 2008 4847 4854