Fabrication, chemical composition change and phase evolution of biomorphic hydroxyapatite
- First Online:
- Cite this article as:
- Qian, J., Kang, Y., Zhang, W. et al. J Mater Sci: Mater Med (2008) 19: 3373. doi:10.1007/s10856-008-3475-5
- 180 Views
Biomorphous, highly porous hydroxyapatite (HA) ceramics have been prepared by a combination of a novel biotemplating process and a sol–gel method, using natural plants like cane and pine as biotemplates. A HA sol was first synthesized from triethylphosphate and calcium nitrate used as the phosphorus and calcium precursors, respectively, and infiltrated into the biotemplates, and subsequently they were sintered at elevated temperatures to obtain porous HA ceramics. The microstructural changes, phase and chemical composition evolutions during the biotemplate-to-HA conversion were investigated by scanning electron microscopy (SEM), X-ray powder diffraction (XRD), and Fourier-transform infrared (FT-IR) spectroscopy. The XRD and FT-IR analysis revealed that the dominant phase of the product was HA, which contained a small amount of mixed A/B-type carbonated HA, closely resembling that of human bone apatite. Moreover, the appearance of a small amount of secondary phase CaCO3 seemed unavoidable. The HA was not transformed to the other calcium phosphate phases up to 1400°C, but it contained a trace amount of CaO when sintered at above 1100°C. The possible transformation mechanism was proposed. The SEM observation and mechanical property test showed that as-produced HA ceramics retained the macro-/micro-porous structures of the biotemplates with high precision, and possessed acceptable mechanical strength, which is suggested to be potential scaffolds for bone tissue engineering.
Bone tissue engineering, involving the fabrication of a porous scaffold, has become an alternative approach to promote the repair and regeneration of diseased or damaged bone tissue . Ideal porous scaffolds for bone tissue engineering should exhibit some necessary properties, including excellent osteoconductivity and bioactivity, biocompatibility, controllable biodegradability and good absorbability, three dimensionally interconnected highly porous structure as a template favorable for new bone ingrowth, and irregular shape fabrication ability [2, 3], etc. Many studies reported that porous hydroxyapatite (Ca10(PO4)6(OH)2, HA) ceramics met most of the above criteria. Particularly, HA can induce the formation of bone-like apatite layer in vitro when exposed to SBF, which can strongly bond to living bone , stimulate new bone formation and growth  and exert a positive effect on the expression of genes regulating osteogenesis . So, synthetic HA similar to bone mineral in chemical composition, biology and crystallography [7, 8], has attracted special interest in the field of bone tissue engineering over past two decades. Compared with other techniques such as wet chemical precipitation technique , hydrothermal method , solid-state reaction , mechanochemical synthesis  and combustion preparation , sol–gel derived HA ceramics exhibit higher bioactivity, biocompatibility, osteoconductivity and better cell responses because of their poorly crystalline and presence of carbonate ions in the crystal lattice [14, 15].
It has been recognized that the pore structure is one of the decisive factors affecting the biological function of scaffolds [16, 17]. To imitate the porous structure of spongy bone, many techniques have been developed, such as polymeric sponge replication , rapid prototyping techniques , electrospinning , phase-separation , particulate leaching , sacrificial filler , freeze drying , solvent casting  and gel-casting techniques . These techniques endued scaffolds with a variety of porous microstructures to satisfy different applications. Recently, special attention is paid to some biomorphic bone substitutes and scaffolds for bone regeneration, originating from biological tissues and natural materials like cuttlefish , seastar , bovine bone , coral , seashell , and red algae , because their unique morphology and inter-connective highly porous structure similar to human bone may improve the solubility of the implant and facilitate cellular activity and faster bone in-growth [32, 33].
In recent years, a novel replication method, i.e. biotemplating technology based on wood, has been applied to manufacture biomorphic porous ceramics with woodlike structures via substitution or transformation processing, including carbide (SiC, TiC) , oxide (Al2O3, Cr2O3)  and nitride (TiN) , etc. Wood, a natural biopolymeric composite mainly comprised of hemicellulose, cellulose and lignin, exhibits a complex micro-/macro-structure and hierarchical microcellular architecture featuring honeycomb-like microchannels, which is very similar to that of cancellous bone. Many natural plants such as cane and wood have highly open trabecular structures, and have been used as sacrificial templates for porous ceramics. Some benefits of innovative wood-based biomorphic materials are the versatility for the fabrication of complex shapes, sufficient biomechanical properties , and intrinsic three-dimensional interconnected porous structure . These merits allow this kind of biomorphic porous materials for biomedical applications like bone implants and scaffolds in bone reconstitution. Coating biomorphic ceramics with bioactive materials can confer them excellent bone-bonding ability . P González et al.  proposed wood-based biomorphic SiC ceramics coated with bioactive glass by Pulsed Laser Deposition as a very promising device for dental and orthopaedic applications. Recently, they  tested the in vitro cytotoxicity of the biomorphic SiC ceramics uncoated or coated with bioactive glass, using MG-63 human osteoblast-like cells. Their results revealed that the biological response of the cells on the biomorphic SiC ceramics was similar to the one exhibited by well-known implant materials like Ti6Al4 V and bulk bioactive glass, which makes it possible to be applied in bone implantology like load bearing protheses.
In fact, there have been some observations on bone ingrowth into Clematis alba—derived charcoal  and birch wood  implants since the 1980 s. In another paper, bone ingrowth and appositional growth in small prosthesis implants of juniper, pretreated by a boiling procedure, was reported . In a recent paper, heat-treated deciduous wood (birch) was directly used as replacement material for osteochondral bone defects in the knee joint of rabbit . As a result, the natural porous channel structure of wood made it serve as a porous bioactive scaffold, which allowed ongrowth of bone and cartilage differentiation on its surface and ingrowth the porous structure (with a diameter of 100–200 μm) of the wood, demonstrating osteoconductive and chondroconductive contact. Then, wood and wood-derived materials might become a promising tissue engineering scaffold strategy.
Much interest within tissue engineering is paid to the design of the scaffolds to ensure good tissue growth. However, a common problem encountered when using scaffolds for tissue engineering is the rapid formation of tissue on the outer edge, leading to the development of a necrotic core, due to the limitations of cell penetration, and oxygen and nutrient exchanges . To overcome this issue, several different strategies have been developed, such as adopting dynamic culture systems , incorporating chemotactic growth factors  and choosing an appropriate composition [49, 50]. It is suggested that the issue may also be solved by an alternative approach, i.e., tailoring the porous structure and morphological properties of scaffolds. For example, the incorporation of macro- and/or micro-channels within scaffolds can not only enhance cell penetration and tissue infiltration but also maintain suitable living cell, nutrient and oxygen concentrations throughout the scaffolds in vitro and in vivo . Moreover, such scaffolds possess higher strength than the scaffolds with the same porosity . Some methods have been developed to produce microchannels in scaffolds, including novel rapid-prototyping techniques , coextrusion process , freeze-dry processing , fiber templating technique , and molding process using acupuncture needles as mandrels , etc. However, it is difficult for them to obtain microchannels with less than 100 μm.
To the best of our knowledge, there has been no report on the preparation of biomorphic HA from natural plant templates. In the present study, we attempt to utilize the combination of this biotemplating method and sol–gel processing, for the first time, to develop novel biomorphic HA ceramic scaffolds from natural plants like wood and cane. The main objective is to probe the influence of the thermal treatment on the morphological, structural and phase changes during the biotemplate-to-scaffold conversion. The final goal is to create an ideal scaffold for bone tissue engineering application.
2.1 HA sol synthesis
The aim of this work is to fabricate biomorphic HA ceramics by a biotemplating process, involving the infiltration of a HA sol into biotempates, e.g. cane and pine (Local plants, Purchased from Xi’an Wood Company, Xi’an, China), and subsequent sintering at elevated temperatures. The HA sol was prepared by a sol–gel route using calcium nitrate tetrahydrate (Ca(NO3)2 · 4H2O, AR, purity > 99.9%, Chongqing Rearents Company, Chongqing, China) and triethyl phosphate (OP(OC2H5)3, TEP, CP, purity > 99.5%, Kunshan Kunhua Group Company, Kunshan, China) as calcium and phosphorous sources, respectively. Other reagents are analytical reagents and purchased from Xi’an Chemical Plant (Xi’an, China). Briefly, TEP was first mixed with distilled water and ethanol, and hydrolyzed for 6 h under vigorous stirring using a magnetic stirrer to form a 0.6 M solution. Then, a 1 M aqueous solution of Ca(NO3)2 · 4H2O was slowly added dropwise into the above-mentioned hydrolyzed TEP solution, according to a stoichiometric molar ratio for the formation of HA. The pH value of the mixed solution is adjusted to 8 using ammonium hydroxide [NH4OH, 17% in H2O]. After the reactants were continuously stirred for additional 12 h at room temperature, a clear liquid was obtained, namely, HA sol. Finally, the sol was kept in a sealed container to allow to age for 5 days.
2.2 Fabrication of biomorphic HA ceramics
2.3 Materials characterization
The morphologies of the fracture surfaces of original templates and template-derived scaffolds were observed by S-2700 scanning electron microscopy (SEM, Hitachi, Japan). The samples were mounted on copper stubs, coated in vacuum with gold using a sputter coater, and observed by SEM at an accelerating voltage of 20 kV.
The effects of sintering temperature on the crystalline phase and chemical structure evolutions were investigated by powder X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FT-IR). The scaffolds were first ground into a powder and 1 g of the powder was collected for XRD analysis. The XRD experiments were performed on a Philips X’Pert MPD diffractometer in the 10° ≤ 2θ ≤ 80° range with a step size of 0.033° for 2θ and a scanning speed of 0.14°/min, employing Cu Kα (λ = 1.54056 Å) radiation generated at 40 kV and 40 mA. The crystalline phases were determined from a comparison of the registered standard JCPDF cards (HA: 09-0432, CaCO3: 05-0586, CaO: 37-1497, Ca(NO3)2: 07-0204, β-TCP: 09-0169) available in the system software with the obtained powder diffraction files. To characterize the chemical composition and molecular structure, FT-IR spectra were recorded on a Shimadzu IR Prestige 21 spectrometer in the range 4,000–400 cm−1 in transmission mode with 1 cm−1 resolution using a typical KBr pellet technique. In order to allow comparisons, the spectra were normalized from the ν4 band of the phosphate group at 602 cm−1 according to a classical procedure .
The samples’ porosity was measured by water displacement method based on the Archimedean principle. For the compressive strength test, cylindrical samples with dimensions of Φ6 mm × height 8 mm were loaded with an Instron 1195 universal testing machine (Instron Corp., USA) at a crosshead speed of 0.5 mm/min. Loading direction was parallel to that of the honeycomb-like pores. More than five samples were tested to obtain the average value along with its standard deviation.
3 Results and discussion
3.1 Morphological characteristics of biomorphic HA ceramics
It can be seen from Fig. 2c, d that, after the removal of organic biotemplates and 3 h of sintering at 1,200°C in air, the hierarchical pore structure of the initial template was preserved intact to cane-derived HA ceramics. The resulting HA is of a rougher wall surface morphology, and many micropores with irregular shape probably originating from the pits of cane and/or deficient sintering clearly exist in walls among the elongated channel-like pores, which makes the resultant HA ceramics exhibit an interconnecting porous network. In addition, it is also noted that the overall pore size observed in HA ceramics decreased lightly in comparison with that of original biotemplates.
The above-mentioned results suggest a possible route for producing novel porous HA ceramics using natural organic templates. According to the existing reports [31, 61], such porous HA ceramics containing interconnected microchannels will permit the circulation of the physiological fluid throughout them, induce protein adhesion, stimulate significantly cell growth and promote bone ingrowth, and modify their resorption behavior, when they are used as bone implants and/or scaffolds for bone tissue engineering.
3.2 Chemical composition changes
The main characteristic bands of PO43− tetrahedral apatite’s structure [59, 64] are clearly observed, given as follows. The peak at 472 cm−1 is attributed to v2 O–P–O bending. The double peak at 571 cm−1 and 602 cm−1 with a high resolution belongs to v4 O–P–O antisymmetric and symmetric bending modes [62, 65], and 962 cm−1 for v1 P–O symmetric stretching. The peaks at 1,040 and 1,088 cm−1 are due to v3 P–O antisymmetric stretching modes . Moreover, the v1 and v3 peaks overlap to give a broad band in the wavenumber interval of 900–1,200 cm−1 . The group of weak intensity bands in the 1,950–2,200 cm−1 region is a reflection of overtones and combinations of the v3 and v1 PO43− modes [33, 68]. These results clearly indicate the formation of a typical HA structure. The double peaks at 2,365 and 2,344 cm−1 are due to CO2 in air . In the present study, however, the characteristic bands of β-TCP at 1,118 cm−1 and 945 cm−1  were not observed for all the sintered samples.
It should be noted that the obvious absorption bands at 1,600–1,400 and 875 cm−1 confirm the presence of carbonate group . The bands centered at 1,420 cm−1 and 1,500 cm−1 correspond, respectively, to v3 asymmetric and symmetric C–O stretching modes , which is different from a single characteristic peak (1,415 cm−1) of free CO32− or carbonates, indicating the presence of at least two carbonated phases [72, 73]. The peaks at 875 cm−1 and 714 cm−1 correspond to the v2 out-of-plane deformation and v4 in-plane deformation O–C–O bending modes in CO3 group [15, 74], respectively. The FT-IR band at 1,420 cm−1 together with the XRD peak at 2θ = 29.42° supports the existence of a CaCO3 phase. The peaks at 2,515 cm−1 and 1,794 cm−1 are attributed to v1 + v3 and v1 + v4 combination bands of CaCO3, respectively. The weak bands at 1,560, 1,510 and 1,460 cm−1 correspond to the incorporation of CO32− groups at the OH position (A-type) , and those at 1,460 and 1,420 cm−1 are ascribed to the substitution of CO32− ions for PO43− sites in the apatite structure (B-type) . The band at 875 cm−1 is assigned to three types of CO32− groups.
As sintering temperature was increased from 700°C to 1,200°C, the characteristic peaks (3,571 cm−1 and 632 cm−1) of OH groups became stronger in intensity and better in resolution. However, when sintering temperature was further increased from 1,200°C to 1,400°C, the intensity of the OH− peaks started to decrease. In contrast, the PO43− bands located at 602, 962, 900–1,300 and 1,950–2,200 cm−1 significantly increased in intensity and became broader with the increasing sintering temperature, indicating a deteriorating molecular arrangement. Such a behavior is in accordance with the XRD analysis which indicated partial dehydration and decomposition of HA phase. The fast reduction in the intensity of CO32− bands at 1,420 and 875 cm−1 with temperature up to 1,400°C was seen, which is strongly indicative of the decomposition of CaCO3. In general, the decomposition of CaCO3 occurs at 580°C or higher , so the stretching modes of CO32− groups should be absent in the spectra of the samples obtained at the present sintering temperatures. However, the remnant presence of the carbonate peaks are still evident in these spectra, even if the sample was heat-treated at 1,400°C, which further proves the inclusion of CO32− groups in the apatite structure . Surprisedly, it is found that there seems be more CO32− ions incorporated into HA crystal lattice structure at 1,300 and 1,400°C than those at 1,100 and 1,200°C. But, this is accordance with the fact that the bands of PO4 groups became broader with increasing heat-treatment temperature.
3.3 Phase evolution
It was previously reported that apatite could evolve from the reactions among the impurity phases such as Ca2P2O7, Ca3(PO4)2, CaCO3, and Ca(NO3)2 present in lower temperature calcined HA .
Comparison of XRD peak height ratios, HA(211) vs. CaCO3(104)
HA(211) vs. CaCO3(104)
Lattice parameters of HA heat-treated at 700–1,400°C and standard HA (JCPDF 09-0432)
In addition, the gases released during the removing biotemplates, such as water, CO, CO2, etc., may also become incorporated into the HA lattice to form carbono-apatite .
According to Hwang et al’s discussion on the changes in crystal lattice parameters caused by carbonate substitution in HA , it can be deduced from the results in Table 2 that a mixed B-type and A-type carbonate substitution occurred, because both the a axis and the c axis expanded except the sample sintered at 1,400°C. It was further confirmed by the phenomenon, in which the (hk0) peaks and (00l) peak shifted to higher angles and to a lower angle than those of standard HA , respectively (data not given here). The result is in agreement with the previous investigations [59, 72, 76], in which the formation of B- or A/B-type carbonated HA was often observed in sol–gel-derived HA ceramics, especially in the sol–gel process involving organic reagents . Then, β-TCP formed during the heat-treatment at 700°C may be due to the decomposition of carbonated HA, and other products should include HA and CaO .
Crystallinity and crystallite size of biomorphic HA ceramics sintered at various temperatures
Degree of crystallinity/%
Crystallite size along c-axis/nm
Crystallite size along a-axis/nm
3.4 Porosity and mechanical properties of biomorphic HA ceramics
Porosity and compressive strength of cane- and pine-derived HA ceramics obtained at 1,100, 1,200, 1,300 and 1,400°C for 3 h
0.19 ± 0.07
0.27 ± 0.03
0.29 ± 0.02
0.28 ± 0.05
0.29 ± 0.09
0.36 ± 0.06
0.37 ± 0.03
0.37 ± 0.04
It has been proven that the novel biotemplating method presented in this study, in combination with a sol–gel technique, may be employed to prepare biomorphic HA ceramic scaffolds from the biotemplates like wood and cane. The inherent pore structure of the templates, both multimodal and monomodal pore size distribution, can be transferred to the final HA scaffolds with high precision. The interconnectivity among channel-like macropores in HA ceramic scaffolds varies with the kind of biotemplates. The extensive densification of the macro-pore walls occurred in the pine-derived HA scaffolds. In contrast, cane-derived HA ceramics exhibited an interconnecting porous network. The FT-IR and XRD results show the dominant crystal phase is identified as HA with a small amount of a mixed A and B type carbonated HA, where hydroxyl or phosphate groups are partially replaced by carbonate ones, whilst the appearance of CaCO3 is unavoidable. Under the same experimental condition, cane-derived HA ceramics have higher porosity and lower compressive strength than pine-derived HA ceramics because of their inherent different macro- and micro-structures. The ideal sintering condition for HA scaffolds should be 1200°C/3 h in terms of the purity and appropriate strength of the products. This study reveals the potential of natural plants in fabricating biomorphic porous HA ceramic scaffolds suitable for bone tissue engineering application.
This work is financially supported by the National Natural Science Foundation of China (No. 50603020 and No. 50773062) and the Program for New Century Excellent Talents in University (NCET-07-0673).