Abstract
In the present work, alumina and diopside were introduced in hydroxyapatite matrix, and hydroxyapatite/Al2O3/diopside ceramic composites with good mechanical properties were fabricated by uniaxial hot pressing. The behaviours of hydroxyapatite/Al2O3/diopside ceramic composites in simulated body fluid were studied by SEM, Fourier transform infrared spectroscopy and electron probe microanalyser. Scanning electron microscopy images showed an obvious mineral layer formed on the soaked composite surface, which indicated that the introduction of Al2O3 and diopside in hydroxyapatite matrix could not only improve the strength and toughness of the composites but also maintain its ability to precipitate an apatite layer.
Introduction
Hydroxyapatite (HA) has been successfully applied in medicine for several years due to its excellent biocompatibility. However, the application of HA materials are limited to small unloaded implants, powders, coatings and low loaded porous implants because their lower fracture toughness does not exceed the value of ∼1 MPa m1/2.1 Nowadays, there have been a number of studies reporting on additives to improve the fracture toughness of pure HA.2–6 Improved mechanical properties are obtained owing to the incorporation of second phases such as ZrO2,2 SiCw (Ref. 5) and Al2O3 (Ref. 6) into HA; the biological activities of these fabricated HA matrix ceramic composites, however, decrease mainly due to the decomposition of HA during the sintering process and subsequent formation of tricalcium phosphate or CaO, which may increase the biodegradability of HA matrix ceramic composites.1
The introduction of alumina and diopside in HA matrix ceramic composites can improve their bending strength and fracture toughness, and composites with high ratio of performances versus cost were obtained for advanced biomaterials.7 Our previous work has discussed the behaviours of HA/diopside ceramic composites in simulated body fluid (SBF), and results showed that HA/diopside ceramic composites have the ability of inducing HA to deposit on the surfaces of the composites.8 Few works about the behaviours of HA/Al2O3/diopside ceramic composites have been carried out. In the present paper, the behaviours of HA/Al2O3/diopside ceramic composites in SBF are discussed to assess whether the introduction of diopside and alumina in HA matrix can maintain its ability to precipitate an apatite layer.
Experimental
Preparation and characterisation of HA/Al2O3/diopside ceramic composites
Hydroxyapatite of high purity (99%) and small grain size (1–3 μm), produced by the Department of Materials Science and Engineering in Shandong University, was used as the starting material. Commercial small grain size (0·5–1 μm) 99·9% pure Al2O3 powder and diopside [MgCa(SiO3)2], composed of 55 wt-%SiO2, 24 wt-%CaO and18 wt-%MgO, were used as additives. The raw materials were blended with each other according to their proportions listed in Table 1. Milling was carried out for 100 h in alcohol using a vibratory ball mill with cemented carbide balls, and then metal mill media impurities were removed by washing in 10 mol.-% hydrochloric acid. After drying, densification of the powder was achieved in a graphite die by uniaxial hot pressing at 1320°C (heating rate, 20°C min−1) at a pressure of 20 MPa in an N2 atmosphere for 60 min.
Starting compositions and mechanical properties of HA/Al2O3/diopside ceramic composites
The sintered bodies were cut into bars, and then standard testpieces (3×4×36 mm) were ground and polished with diamond paste. Three-point bending mode was used to measure the bending strength using an electronic universal experimental instrument (Jinan Test Co., Ltd) with a span of 20 mm at a crosshead speed of 0·5 mm min−1. At least 12 specimens were tested for each series of composition in air at room temperature. Hardness was measured on the polished surfaces with a load of 9·8 N for 5 s using a microhardness tester (Shanghai Hengyi Electronic Testing Instrument Corporation). Fracture toughness measurements were performed using the indentation method. The indentations on polished surfaces were generated by a Vickers microhardness tester with a diamond pyramid indenter at a load of 196 N and a loading time of 30 s. The formula proposed by Cook and Lawn9 was used to calculate the final fracture toughness. Data of hardness and fracture toughness were determined using at least 10 indentations on polished surfaces with an Ra of 0·1 μm for each specimen.
Ability of HA/Al2O3/diopside ceramic composites to form apatite layer
Simulated body fluid containing ion concentrations similar to those in human blood plasma was prepared according to the method described by Kokubo.10 Briefly, reagent grade CaCl2, K2HPO4.3H2O, NaCl, KCl, MgCl2.6H2O, NaHCO3 and Na2SO4 were dissolved in distilled water and adjusted to pH 7·25. Pure HA and HA/Al2O3/diopside ceramic bars were soaked in SBF at 37°C for 9 days. After soaking, the bars were dried at 100°C for half day. Microstructures of the specimens were studied on the original polished surfaces and on the soaked bar surfaces by scanning electron microscopy (Hitachi S-570). Fourier transform infrared spectroscopy (FTIR; Vector) analysis was adopted to identify the appearance of OH– and
. The element analysis in the microzone on the surfaces of the soaked pure HA and HA/Al2O3/diopside ceramic composites was detected by an electron probe microanalyser (EPMA; JXA-8800R).
Results and discussion
Microstructures of HA/Al2O3/diopside ceramic composites
Scanning electron microscopy images of the H0 specimen before and after soaking in SBF for 9 days are shown in Figs. 1 and 2. Those of the H1 specimen before and after soaking in SBF for 9 days are shown in Figs. 3–5 respectively. The polished surfaces of H0 and H1 specimens before soaking in SBF were smooth (Figs. 1 and 3). Compared with the polished surfaces of H0 and H1 specimens before soaking in SBF, there were obvious bright mineral layers precipitated on the polished surfaces of H0 and H1 specimens after soaking in SBF for 9 days (Figs. 2, 4 and 5). The mineral layers that precipitated on the polished surfaces of H0 specimen showed a typical ‘cauliflower’ morphology of HA crystals (Fig. 2). There were some crystals with cuboidal shape precipitated on the polished surfaces of H1 specimen (Figs. 4 and 5), which suggested that they may be sodium chloride. The precipitation of some cuboidal shaped crystals may be mainly because SBF was not rinsed off. Some crystals with cauliflower shape were also found in Fig. 5, which indicated that the introduction of Al2O3 and diopside in the HA matrix could not only improve the strength and toughness of the composites but also maintain its ability to form an apatite layer. In order to detect the phases of the bright mineral layers formed on the composite surfaces, the specimens were characterised by FTIR and EPMA respectively.

Image (SEM) of H0 specimen before soaking in SBF (×12 000)

Image (SEM) of H0 specimen after soaking in SBF for 9 days (×500)

Image (SEM) of H1 specimen before soaking in SBF (×8000)

Image (SEM) of H1 specimen after soaking in SBF for 9 days (×4000)

Image (SEM) of H1 specimen after soaking in SBF for 9 days (×6000)
Fourier transform infrared spectroscopy analysis of HA/Al2O3/diopside ceramic composites
Fourier transform infrared spectra of the H0 and H1 specimens after soaking in SBF for 9 days were analysed by FTIR (Vector, Germany) (Figs. 6 and 7). The vibration peak of hydroxyl bonds located at 3575, 3444, 3440, 1646, 634, 603 and 601 cm−1, and that of
at 1644 and 1470 cm−1. The vibration peak of
ranged from 1124 to 946 cm−1. The analysis of FTIR spectra made it sure that the phase of bright mineral layers, formed on the surfaces of the H0 and H1 specimens after soaking in SBF for 9 days, were mainly HAP or hydroxycarbonate apatite (HCA), which had a similar mineral composition to human bones.11,12

Fourier transform infrared spectra of H0 specimen after soaking in SBF for 9 days

Fourier transform infrared spectra of H1 specimen after soaking in SBF for 9 days
Electron probe microanalyser analysis of HA/Al2O3/diopside ceramic composites
Energy dispersive spectroscopy spectra of the H0 and H1 specimens after soaking in SBF for 9 days are shown in Figures 8 and 9 respectively. Ca and P appeared on the surfaces of H0 and H1 specimens after soaking in SBF for 9 days, which indicated that the phase of bright mineral layers were mainly HAP or HCA.

Energy dispersive spectroscopy spectra of H0 specimen after soaking in SBF for 9 days

Energy dispersive spectroscopy spectra of H1 specimen after soaking in SBF for 9 days
It was reported that silicon played an important role for the nucleation and growth mechanism of the apatite-like layer on bioactive composites.13,14 In the present study, silicon was introduced by adding diopside in the composites, so an interchange was suggested to take place between the Ca2+ ions of the HA/Al2O3/diopside ceramic composites and the H3O+ of the SBF, which could promote the formation of Si–OH groups and induce apatite nucleation. The nuclei thus formed and grew at the expense of the ions in the solution that has been saturated with respect to apatite. Therefore, silicon seemed to be very important in maintaining the in vitro bioactive behaviour of HA/Al2O3/diopside ceramic composites.
Conclusions
Hydroxyapatite/Al2O3/diopside ceramic composites with good mechanical properties were fabricated by uniaxial hot pressing. In vitro test behaviour confirmed that a bright mineral layer precipitated on the surfaces of pure HA and HA/Al2O3/diopside ceramic composites soaked in SBF for 9 days. Scanning electron microscopy examination, EPMA and FTIR analysis showed that the bright mineral layer were mainly HAP or HCA. It was concluded that the fabricated HA/Al2O3/diopside ceramic composites not only had better mechanical properties than pure HA but also had the ability to form an apatite layer.
Footnotes
Acknowledgements
The work described in the present paper is supported by the Shandong Province Science and Technology Development Plan (grant no. 2011GGX10204), the Natural Science Foundation of Shandong Higher Education Institutions of China (grant no. ZR2011EL026) and the Research Award Fund for Outstanding Young Scientists of Shandong Province in China (grant no. BS2012ZZ002).
