apatite composite beads prepared by in situ generation of apatite or Si-apatite nanocrystals

Acta Biomaterialia 6 (2010) 466–476

Contents lists available at ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat

Chitosan/apatite composite beads prepared by in situ generation of apatite or Si-apatite nanocrystals Natalia Davidenko a,c,*, Raúl G. Carrodeguas a,b, Carlos Peniche a, Yaimara Solís a, Ruth E. Cameron c a

Centro de Biomateriales, Universidad de La Habana, La Habana 10400, Cuba Instituto de Cerámica y Vidrio, Consejo Superior de Investigaciones Científicas (CSIC), Madrid 28049, Spain c Department of Materials Science and Metallurgy, University of Cambridge, Cambridge CB2 3QZ, UK b

a r t i c l e

i n f o

Article history: Received 8 May 2009 Received in revised form 8 July 2009 Accepted 15 July 2009 Available online 25 July 2009 Keywords: Chitosan Hydroxyapatite Nanocomposite Silicate

a b s t r a c t The objective of this work was to develop nanocrystalline apatite (Ap) dispersed in a chitosan (CHI) matrix as a material for applications in bone tissue engineering. CHI/Ap composites of different weight ratios (20/80, 50/50 and 80/20) and with CHI of different molecular weights were prepared by a biomimetic stepwise route. Firstly, CaHPO42H2O (DCPD) crystals were precipitated from Ca(CH3COO)2 and NaHPO4 in the bulk CHI solution, followed by the formation of CHI/DCPD beads by coacervation. The beads were treated with Na3PO4/Na5P3O10 solution (pH 12–13) to crosslink the CHI and to hydrolyse the DCPD to nanocrystalline Ap. This new experimental procedure ensured that complete conversion of DCPD into sodium-substituted apatite was achieved without appreciable increases in its crystallinity and particle size. In addition, composites with silicon-doped Ap were prepared by substituting Na3PO4 by Na2SiO3 in the crosslinking/hydrolysis step. Characterization of the resultant composites by scanning electron microscopy, X-ray powder diffraction (XRD), thermal analysis and Fourier transform infrared spectroscopy confirmed the formation, within the CHI matrix, of nanoparticles of sodium- and carbonate-substituted hydroxyapatite [Ca10–xNax(PO4)6–x(CO3)x(OH)2] with diameters less than 20 nm. Relatively good correspondence was shown between the experimentally determined inorganic content and that expected theoretically. Structural data obtained from its XRD patterns revealed a decrease in both crystal domain size and cell parameters of Ap formed in situ with increasing CHI content. It was found that the molecular weight of CHI and silicate doping both affected the nucleation and growth of apatite nanocrystallites. These effects are discussed in detail. Ó 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction Effective substitution of defective bone tissue is a major challenge in orthopaedic surgery. A great deal of attention has been focused on the development of suitable biocompatible and bioactive materials, which may be used successfully for long-term bone repair and regeneration. For this, one of the most widely used materials is hydroxyapatite (HA) [Ca10(PO4)6(OH)2] [1–8], which is known to be biocompatible, bioactive (i.e. having an ability to form a direct chemical bond with surrounding tissues), osteoconductive, non-toxic, non-inflammatory and non-immunogenic. Conventional micro-scale HA is employed in different forms (powder, dense or porous blocks, etc.) depending on the requirement of the particular application. HA and calcium-deficient HA in nano-particulate form are beginning to play a significant role in bone-related therapies owing to their unique functional properties of high surface area to volume * Corresponding author. Address: Centro de Biomateriales, Universidad de La Habana, La Habana 10400, Cuba. Tel.: +44 1223 248244; fax: +44 1223 334324. E-mail address: [email protected] (N. Davidenko).

ratio and their ultrafine structure, similar to that of biological apatite. This similarity is of importance in cell–biomaterial interactions [9,10]. Micro-particulate HA has mainly been used for the treatment of periodontal osseous defects and alveolar ridge augmentation, because of its ease of production and handling and close surface contact with the surrounding tissue [11]. However, particulate HA suffers from the disadvantage that it can migrate from the implanted site, potentially causing damage to healthy tissue [12]. To resolve this clinical problem, composite materials based on HA and natural [13–24] or synthetic [25–27] polymers have been developed. The advantages of these kinds of systems lie, in general, in combining the functional properties of single phase materials. Natural polymer-based composites obtained from collagen [13], alginate [15], gelatine [17] and chitosan (CHI) [14,16,18–24] are playing a significant role in many tissue engineering applications due to their tailor-made physical, chemical, mechanical and biological characteristics. CHI (the N-deacetylation product of chitin) is being increasingly used in the design of environmentally responsive biomaterials

1742-7061/$ - see front matter Ó 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.actbio.2009.07.029

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

[16,18–24] because of its biocompatibility, biodegradability and bioactivity in the human body. This biopolymer can be easily obtained from chitin, an abundant and renewable material in the biosphere. In addition, CHI is flexible, has good solubility in various organic acid solutions and, due to the intramolecular hydrogen bonds formed between hydroxyl and amino groups, is suitably resistant to heat and certain alkaline media [19]. It has also been reported that the mechanical reliability and in vivo biomineralization of CHI/HA composites are superior to those of either HA or CHI alone [19,28]. Various ways of preparing CHI/HA composites have been developed, such as mixing an HA powder in a CHI solution [28] and coating HA particles onto a CHI film [29]. However, the composites obtained by these means were microscopically non-homogeneous and often caused inflammation when implanted. To overcome these drawbacks, Yamaguchi et al. [19,23] proposed a one-step co-precipitation method in which a CHI solution containing phosphoric acid was dropped into a calcium hydroxide suspension. A similar procedure was also employed by Chang et al. for the preparation of HA/gelatine nanocomposites [17]. Other authors biomineralized CHI in a solid form (for example, as membranes) in simulated body fluid [30]. Another approach was used by Hu et al. [31] in which the CHI hydrogel was mineralized by controlled in situ hybridization by an ionic diffusion process. The above approaches led to the incorporation of inorganic fillers into the structure of composites in the form of either nano- or micro-sized particles. Recently, an interesting bio-inspired route for the design of composite materials containing CHI and nano-sized HA has been reported [18]. The authors suggested that a biomimetic approach should enable both the architecture (the structure) and the chemistry (the composition) of synthetic biomaterials to be controlled. The method consisted of the formation of dicalcium phosphate dihydrate (DCPD) (CaHPO42H2O) in bulk CHI acidic solution and their co-precipitation by gradually increasing the pH of the reaction medium by the dropwise addition of NaOH solution. It should be noted that all of the above mentioned approaches lead to a particular type of CHI/HA composite material each with a distinctive structure and properties. In a similar way Peña et al. [24] dispersed a commercial DCPD powder in a CHI solution in diluted citric acid. The resulting viscous slurry was transformed at room temperature into a CHI/Ca-deficient apatite/composite on being poured into NaOH solution. DCPD is a well-known precursor of apatite (Ap) and will hydrolyse into either octacalcium phosphate or Ap in various aqueous solutions. The exact composition of the resulting hydrolysis product depends upon the concentration, composition and pH of the hydrolysing solution, as well as the time and temperature of hydrolysis and other factors [32– 37]. In reports by Rusu et al. [18] and Peña et al. [24] NaOH solution was used at a physiological temperature during hydrolysis of DCPD in the bulk of a CHI matrix. However, such hydrolysis conditions did not lead to complete conversion of DCPD to HA [18]. In the current work the use of a solution of sodium phosphate is proposed as an alternative hydrolysis medium for DCPD conversion into sodium-substituted Ap, on the assumption that the buffer capacity of a warmed solution of sodium phosphate would enable the conversion of DCPD to proceed to completion [32,37]. Thus, as described in the current paper, a new three step route was developed in which composites, formed from nano-Ap crystals dispersed in a CHI matrix, could be obtained. The selection of salts (calcium acetate and dihydrogen sodium phosphate) as DCPD precursors, rather than strong acids or bases, and the substitution of sodium hydroxide by tri-sodium phosphate in the hydrolysis medium represent the new aspects of this bio-inspired procedure, aimed at both avoiding any further destructive impact of the inor-

467

ganic component on the polymeric chains of CHI and guaranteeing complete conversion of DCPD into Ap. This experimental procedure was also applied to introduce silicon (Si) into the crystalline network of the Ap formed in the CHI matrix. The use of Si [38–47] is currently receiving attention for biomedical applications as a substitute for Ca in the crystal network of HA and tri-calcium phosphate. Silicon plays an essential role in the metabolic events that lead to endochondral and intramembranous bone formation [48,49], and together with Ca, Na and P acts on the expression of certain genes responsible for controlling the cell cycle of animal and human osteoblasts and stimulates osteoproduction [50–52]. Many attempts to introduce Si into an HA network as an effective way to improve the bioactivity level of the material have been carried out [38,39,41–45,53–61]. With the above criteria in mind, and as a distinctive aspect of the current study (compared to the above cited works by Rusu et al. [18] and Peña et al. [24]) silicon (Si) was introduced as an ionic doping agent during the in situ generation of the Ap phase dispersed in the bulk CHI matrix. In this way it could be expected that, as a result of the partial or total substitution of the tri-sodium phosphate hydrolysing solution by sodium silicate solution, CHI/ silicon-doped apatite (CHI/Si-Ap) composites with improved osteogenic properties would be obtained. Studies were also made on the influence that the molecular weight and degree of deacetylation of CHI and the CHI/Ap weight ratio in the composites would have on the average crystallite size of the inorganic component, since it was considered that the polymer matrix would, in all likelihood, affect the nucleation and growth of Ap/Si-Ap nanocrystallites [62]. In summary, the goal of this research has been to introduce new experimental procedures to the biomimetic approach, to incorporate Si into the Ap structure and to determine how the different structural properties of the biopolymer influence the dimensional characteristics of Ap crystals, all with a view to establishing a suitable, bio-inspired route for producing novel materials under controlled conditions at the nano level. Scanning electron microscopy (SEM), powder X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, thermogravimetry (TG) and differential thermal analysis (DTA) were employed in the characterization of the composites produced. Structural data about the crystallite size of Ap and Si-Ap in the composites were obtained from XRD patterns.

2. Materials and methods 2.1. Materials Three different CHIs were employed for the preparation of composites. They are designated MD-CHI [deacetylation degree (DD) = 89.3%, Mv = 1.4  105] (Primex, Iceland), M-CHI (DD = 78.7%, Mv = 1.14  105) and L-CHI (DD = 79.4%, Mv = 5.89  104), both supplied by Aldrich, UK. Samples M-CHI (medium molecular weight) and L-CHI (low molecular weight) had similar degrees of deacetylation but differed in their molecular weights. Sample MD-CHI has a similar molecular weight to sample M-CHI, but a higher degree of deacetylation. Purification was as follows. CHI was dissolved in aqueous acetic acid (2%) until a homogeneous 1% CHI solution was attained. The pH of the CHI solution was increased to 8.0 by dropwise addition of 10% NaOH solution. The precipitated CHI gel obtained was washed several times with deionised distilled water and vacuum dried in a desiccator. Glacial acetic acid (CH3COOH) was obtained from PANREAC (Barcelona, Spain), calcium acetate [Ca(CH3COO)2] from BDH (Poole, UK), sodium phosphate (Na3PO4) from Riedel-de Haën (See-

468

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

ize, Germany), sodium tripolyphosphate (Na5P3O10) from Acros Organics (Geel, Belgium) and sodium dihydrogen phosphate (NaH2PO4H2O), sodium hydroxide (NaOH) and methanol (CH3OH) from Sigma Chemical Co. (Steinheim, Germany). All solutions were prepared with redistilled water. 2.2. Methods 2.2.1. Preparation of composites In order to obtain size-controlled nano-apatite particles dispersed in the CHI matrix, a stepwise co-precipitation route was followed. Dispersed CaHPO42H2O (DCPD) crystals were first precipitated in the bulk of a CHI solution. This was followed by the formation of CHI/DCPD beads, by dropping the CHI solution into a NaOH–CH3OH solution. After washing, the CHI beads containing dispersed DCPD crystals were treated with Na3PO4/ Na5P3O10 solution to crosslink the CHI and to hydrolyse DCPD into nanocrystalline apatite (Ap). To control the HA/Ap ratio in the final composite, different proportions of the soluble salt precursors were employed depending upon the CHI content. The synthesis route may be summarized in two fundamental steps: (a) in situ precipitation of DCPD; (b) further hydrolysis of DCPD to apatite. These steps are represented by the following equations: CHI

CaðCH3 COOÞ2ðaq:Þ þ NaH2 PO4ðaq:Þ þ 2H2 O ! CaHPO4  2H2 OðsÞ þ CH3 COOHðaq:Þ þ CH3 COONaðaq:Þ

ð1Þ

x ð10  xÞCaHPO4  2H2 OðsÞ þ Na3 PO4ðaq:Þ 3 CHI

þ xH2 CO3ðaq:Þ ! Ca10x Nax ðPO4 Þ6x ðCO3 Þx ðOHÞ2ðsÞ   12 þ x þ H3 PO4ðaq:Þ þ ð18  2xÞH2 O 3

ð2Þ

The experimental procedure to prepare the CHI/Ap composites of different weight ratios (20/80, 50/50 and 80/20) was as follows. To a determined volume of 2 wt.% CHI aqueous solution in 1% acetic acid were added different volumes of 5 wt.% NaH2PO4H2O and 5 wt.% Ca(CH3COO)2 aqueous solutions, at molar ratio Ca/P = 1.5 (Table 1). The mixture was then aged and vigorously stirred overnight to ensure homogeneous dispersion of the fine precipitate of DCPD in the viscous CHI solution. The resulting CHI/DCPD suspensions were dropped into an excess of 10 wt.% NaOH solution in CH3OH. The beads thus obtained were left to stand at room temperature for 2 h and subsequently washed four times with cold and warm water, successively. The CHI/DCPD beads were hydrolysed in a saturated solution of Na3PO4 containing 1 wt.% Na5P3O10 as crosslinking agent. The hydrolysis proceeded at 60 °C for 72 h with occasional stirring. Finally, the CHI/Ap composite beads were filtered, washed free of alkali and freeze dried (110 °C, 24 h).

2.2.2. Optical microscopy (OM) and scanning electron microscopy (SEM) Images of the ‘as obtained’ and lyophilized samples of the composites beads were obtained using a Zoom Stereomicroscope SMZ1000 (Nikon Instrument, USA) coupled to a CCD-IRIS video camera (Sony, USA). Freeze dried composite samples were glued onto a conductor sample holder, vacuum coated with gold and examined with a S470 cold field emission scanning electron microscope (Hitachi, Japan) operating at 20 kV. 2.2.3. FTIR analysis FTIR spectra were obtained with a Tensor 27 FTIR spectrometer provided with the OPUS Data Collection Program (Bruker Optics, USA). Samples were prepared in the form of KBr discs. Spectra were recorded over the range 4000–400 cm1 at 2 cm1 resolution, with the accumulation of 32 repeated scans. 2.2.4. Thermal analysis Thermal analyses were carried out with a Q500 V6.5 Build 196 thermogravimetric analyser (TA Instruments, USA). Measurements by thermogravimetry (TG), differential thermogravimetry (DTG) and differential thermal analysis (DTA) were recorded at a heating rate of 10 °C min1 under a flow of air in the range 25 up to 600– 1000 °C, as indicated in the corresponding figures. Aluminium crucibles were employed for the samples and reference (a-Al2O3). Sample sizes ranged from 5 to 9 mg. 2.2.5. X-ray diffraction X-ray diffraction patterns of the different samples studied in this work were obtained with a Bruker Focus diffractometer equipped with a D8 (h/h) goniometer and a LynxEyeTM position-sensitive detector with a Cu anode (Ia1/Ia2 ratio = 0.5), operating at 40 kV and 40 mA. Diffraction patterns were recorded over a 2h range of 5–80° in continuous mode. The step size was 0.02° and 0.5 s per step. The EVA program and PDF-2 JCPDS files data bases were used to perform the automated phase analysis. JCPDS files Nos. 46-0905 and 39-1894, corresponding to calcium-deficient hydroxyapatite (CDHA) [Ca9HPO4(PO4)5OH] and CHI, respectively, were used for identification. 2.2.5.1. Determination of crystal domain size and cell parameters of apatite formed in situ. X-ray diffraction was used to estimate crystallite size and cell parameters of the apatite phase formed inside the composites. The experimental set-up and conditions described above were employed for data collection, but the 2h range of 5–120° was scanned in continuous mode with a step size of 0.01° and 0.5 s per step.

Table 1 Solutions of reagents and their volumes employed for samples preparation.a Designation

CHI20 CHI50 CHI50-Si50 CHI50-Si100 CHI80 CHI80-Si100 Ap Ap-Si100 a

Volume (ml) CHI (2 wt.%)

Ca(CH3COO)2 (5 wt.%)

NaH2PO4 (5 wt.%)

NaOH/MeOH (10 wt.%)

Na3PO4/Na5P3O10 (sat./1 wt.%)

Na3SiO4 (8.8 wt.%)

25 62.5 62.5 62.5 100 100 0 0

60 37.5 37.5 37.5 15 15 75 75

45.4 28.5 28.5 28.5 11.4 11.4 57 57

150 150 150 150 150 150 0 0

70 70 35 0 70 0 70 0

0 0 35 70 0 70 0 70

Theoretical yield: 2.5 g of sample.

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

469

The Engel et al. [63] variant of the Pawley method [64] in the module Reflex Powder Refinement of Materials Studio Modelling v4.0.0.0 (Accelrys Software Inc.) was employed to estimate crystal domain size and cell parameters. This variant of the Pawley procedure uses the Bragg intensities as freely varying refinement parameters and disregards the underlying atomic structure. In refining those parameters affecting peak shapes and positions, such as profile parameters, peak broadening, asymmetry parameters, line shift corrections and cell parameters, the Pawley method has a distinct advantage, in comparison with the conventional Rietveld refinement method, in that it does not require any prior knowledge of the unit cell contents. It is, therefore, often easier to refine such parameters using the Pawley method. The algorithm employed for calculating crystal domain size from peak broadening was based on the Scherrer equation (Eq. (3)):

L¼K

k Dð2hhkl Þcos hhkl

ð3Þ

where L is the average crystal domain size, k is the wavelength of the X-ray radiation, D(2hhkl) is the broadening to full width at half maximum (FWHM) of the peak, h is the Bragg angle and K is a correction factor close to 1 which, among others, accounts for the difference between ‘‘integral breadth” and FWHM. Similarly, the estimation of cell parameters from peak position was supported by Eq. (4) [65]. 2

sen2 hpeak ¼

2

k2 h þ hk þ k 3 a2

!

2

þ

l c2

ð4Þ

where hpeak is the Bragg angle of the peak, k is the wavelength of the X-ray radiation, h, k, and l are the Miller indices of the reflection and a and c are the hexagonal cell parameters. Samples were mixed with 10 wt.% pure polycrystalline silicon as an internal standard and pressed into discs for examination. The experimental X-ray diffraction pattern from the silicon standard was refined to obtain peak profile and instrument variables, assuming an absence of peak broadening caused by strains and crystal domain size and a pseudo-Voigt peak profile. The Profile and instrumental variables thus obtained were used in the refinement of crystal domain size and cell parameters for the apatite in the samples. International Crystallographic Standards Database structure files Nos. 29287 and 152190, corresponding to silicon and CDHA, respectively, were used for refinement.

Fig. 1. Optical micrographs of beads of the composite M-CHI80-Si50. (a) As obtained after water washing (hydrogel); (b) cross-section of a bead; (c) freeze dried (Xerogel).

3. Results and discussion 3.1. Preparation and morphology of composites In order to vary the CHI/Ap ratio in the final composite a novel procedure was established whereby Ap was co-precipitated in a stepwise fashion from soluble precursor salts using different proportions of starting ingredients. Basically, by adding a solution of NaH2PO4 (pH 5) to a solution of CHI in dilute CH3COOH followed by a solution of Ca(CH3COO)2 (pH 5), all under stirring, a dispersion of fine DCPD crystals in acidic CHI solution was obtained. The resulting suspension was then added dropwise into a H2O–CH3OH solution of NaOH to precipitate CHI as droplets in which DCPD crystals were embedded. Finally, the DCPD crystals embedded in the CHI matrix were hydrolysed into Ap in a solution of Na3PO4 (pH 12). After washing (to pH 7) the resultant composite beads were roughly spherical, with average diameters between 300 and 250 lm, as shown for sample M-CHI80-Si50 in Fig. 1a. The crosssection of a bead of M-CHI80-Si50, obtained by cutting with a razor blade, is shown in Fig. 1b. The composite beads had a skin–core

structure which reflected the kinetics of gel drying with a shell thickness of 10 lm. The composite beads shrank and their surfaces became rough after freeze drying (Fig. 1c). The scanning electron micrographs of freeze dried composites shown in Fig. 2 reveal that they were made of aggregates of individual particles each of 50 nm size. No segregation of Ap crystals was observed, suggesting that they are embedded in the CHI matrix and both together constitute the individual particles that form, in turn, the above mentioned aggregates. 3.2. Qualitative analysis of crystalline phases The only crystalline phases detected in the samples of composites were those of Ap and CHI (see Fig. 3). The broad peak between 15° and 30° (2h) corresponds to the main diffraction peaks of CHI as reported in JCPDS file No. 39-1894. The overlapping of the characteristic peaks that coalesce into a broad band is characteristic of the low crystallinity polymeric network of CHI, where much of the material is amorphous.

470

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

The ashes obtained after calcination of CHI/Ap composites in air at 1000 °C for 15 h were also submitted to X-ray qualitative phase analysis. Strong characteristic diffraction peaks of HA (JCPDF file No. 9-432) and weaker peaks corresponding to b-CaNaPO4 (JCPDF file No. 29-1193) were detected. This result suggests that partial substitution of Na by Ca took place in the apatite produced during the in situ hydrolysis of DCPD in the presence of a high concentration of Na ions. 3.3. Infrared spectroscopy

Fig. 2. Scanning electron micrographs of composites (a) M-CHI20 and (b) M-CHI50.

Fig. 4 shows the FTIR spectra of Ap, M-CHI and their respective composites. The infrared spectrum of Ap (Fig. 4a) exhibited the characteristic absorption bands of HA at 3568 cm1 (OH stretch) and 631 cm1 (OH librational). The broad bands peaking around 3430 and 1645 cm1 are due to adsorbed water (HOH stretch and bending, respectively) [66]. It also displayed the typical absorption bands of the PO4 group in HA at 1000–1100 cm1 (m3) and 962 cm1 (m1) [67]. The absorption bands observed at 604 and 565 cm1 correspond to the m4 mode of the PO4 group [67]. Bands at 3568 and 631 cm1 were hardly visible, indicating the low crystallinity of the sample, which was also confirmed by the presence of only two peaks around 600 cm1 instead of the three described in reports on more crystalline apatites, such as stoichiometric HA [16,68]. The occurrence of bands at 1456, 1418 and 874 cm1 indicates the presence of type B carbonate ions in the apatite [66]. As bone mineral contains a substantial amount of carbonate (4–8 wt.%) in the human body, the presence of carbonate in the Ap composites is advantageous, as it increases the mechanical consistency and bioactivity of the apatite [69,70] and, hence, more osteoconduction and tissue in-growth is expected upon implantation. The infrared spectrum of M-CHI (Fig. 4e) displayed the distinctive absorption bands at 2942–2784 cm1 (aliphatic C–H stretching band), 1657 cm1 (amide I) and 1597 cm1 (NH2 bending) and 1321 cm1 (amide III). The absorption bands at 1154 cm1 (anti-symmetric stretching of the C–O–C bridge), 1082 and 1032 cm1 (skeletal vibrations involving C–O stretching) are characteristic of its saccharide structure [71]. The spectra of the CHI/ CDHA composites (Fig. 4b–d) exhibited the absorption bands of both components, while the characteristic peaks of M-CHI became stronger and more defined as the CHI content in the composites increased. This confirmed that the composites were richer in the polysaccharide in the order expected.

Fig. 3. X-ray powder diffraction patterns for Ap and M-CHI and composites M-CHI20, M-CHI50 and M-CHI80. Ap, JCPDS file No. 46-0905; CHI, JCPDS file No. 39-1894.

Peaks characteristic of Ap matched JCPDS file No. 46-0905, reported for CDHA (Ca9HPO4(PO4)5OH), well (Fig. 3). These peaks were also quite broad, which suggests that the apatite crystals were of small size.

Fig. 4. Infrared spectra of samples (a) Ap, (b) M-CHI20, (c) M-CHI50, (d) M-CHI80 and (e) M-CHI.

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

Fig. 5 shows the infrared spectra of those materials prepared in the presence of silicate ions during the hydrolysis step in comparison to their Si-free analogues. The spectra of Ap (Fig. 5a) and ApSi100 (Fig. 5b) are very similar. According to previous reports no differences exist between the infrared spectra of ‘‘as-precipitated” HA and Si-doped HA. Small differences are only noticeable after heat treatment of the material above 1000 °C [61,72,73]. The same is observed for spectra of composites containing 50 wt.% CHI obtained by hydrolysis in silicate (Fig. 5c), silicate–phosphate (Fig. 5d) and phosphate (Fig. 5e) solutions. 3.4. Thermal analysis In the traces from the thermogravimetric analysis of the MDCHI/CDHA composites (Fig. 6) the three main decomposition steps of chitosan are observed. In the first step an endothermic effect (see DTA curve in Fig. 6), associated with the loss of water in the temperature interval 30–150 °C, takes place. The second decomposition step corresponds to a complex exothermic process in which dehydration of the saccharide rings of chitosan occurs, together with decomposition of the acetylated and deacetylated units of the polymer [74]. The final step of pyrolysis corresponds to the oxidative decomposition of chitosan residues. The curves are parameterised in Table 2. It can be seen from Fig. 6 that the shapes of the weight loss vs. temperature curves and the DTA profiles differ considerably from one sample composition to another. This is indicative of the existence of an interaction between the chitosan matrix and the inorganic phase [68]. Moreover, mixing chitosan with HA may cause a decrease in its crystallinity and the chain order in the composite. These factors may explain why, in Table 2, one can observe that as the chitosan content of the composites increased they became more thermally stable and the decomposition of chitosan started at higher temperatures. The same pattern of behaviour was observed for the composites of the three chitosan samples employed (M-CHI, L-CHI and MD-CHI). The curve of Ap (Fig. 7d) does not show any transition in the temperature interval studied, indicating its high stability. The small weight loss observed below 300 °C has been attributed to dehydroxylation of nanocrystalline apatite [75]. The total weight loss due to the thermal decomposition of chitosan decreased as the inorganic content in the composites increased (Fig. 7b–d). Thermogravimetric analysis is a useful tool for determining the quantity of inorganic content in CHI/Ap com-

Fig. 5. Infrared spectra of samples (a) Ap, (b) Ap-Si100, (c) M-CHI50-Si100, (d) MCHI50-Si50 and (e) M-CHI50.

471

posites [68,76]. Table 3 shows the following: (a) the composition, as calculated from their corresponding TG curves, of all materials prepared; (b) weight losses resulting from the first degradation stage (loss of water), and from the second and third steps (burning out of chitosan at 600 °C); (c) the amount of residue. Table 3 shows reasonably good correspondence between the calculated inorganic content of the composites and the theoretically expected value for the samples prepared with chitosans of medium molecular weight. The composites obtained with CHI of low molecular weight contained a higher proportion of the organic phase for the three compositions studied. This Ap loss could be the result of a higher porosity of the L-CHI beads. 3.5. Average crystal domain size of Ap formed in situ The results obtained for average crystal domain size are displayed in Table 4. From Fig. 8a and b it is evident that the crystallite size along the c and [0 k 0] directions decreased as the chitosan content in composites prepared with L-CHI (Fig. 8a) and M-CHI (Fig. 8b) increased, in agreement with previous reports [18,75]. Crystal domain size along [0 0 l] exhibited the same trend for high chitosan content. However, from a comparison of Fig. 8a and b it seems that the molecular weight of CHI does not affect the average crystal size of Ap in the composites. The use of silicate solution in the hydrolysis step also caused a reduction in the crystallite size along all studied directions, as previously observed for silicon-substituted apatites [77]. Fig. 9 shows the effect of adding silicate ions to the reaction solution in the DCPD ? Ap hydrolysis step. In the absence of CHI average crystal domain size clearly decreased from sample Ap to Ap-Si100, which was prepared in the same way as Ap except that the Na3(PO4)/Na5P3O10 solution in the hydrolysis step was replaced by a saturated solution of Na2SiO3. In composites M-CHI50, MCHI50-Si50 and M-CHI50-Si100 silicate addition during the hydrolysis step also reduced the average crystal domain size, although the decrease was less pronounced than in the absence of CHI. 3.6. Cell parameters of the Ap formed in situ The values obtained for cell parameters, a = b and c of the different samples are listed in Table 4. Fig. 10 shows the variation with CHI content of the cell parameters of Ap formed in situ for composites with low and medium molecular weight CHI. Parameter a = b (Fig. 10a) decreased with CHI content for L-CHI and M-CHI, but the decrease was more intense for composites prepared from L-CHI. A similar picture is observed in Fig. 10b for the parameter c, where the parameter value decreased drastically with the increase of CHI content in the composites for L-CHI. However, for those composites made of M-CHI the decrease in c with CHI content is considerably less significant. The presence of silicate ions during the hydrolysis step increased the parameter a = b and decreased the parameter c, as can be seen in Fig. 11 by comparing sample Ap with Ap-Si100. Composites prepared with 50 wt.% M-CHI (samples M-CHI50 and M-CHI50-Si50 in Fig. 11) showed the same effect. However, MCHI50-Si100, the composite prepared by hydrolysis in phosphate-free silicate solution, deviated from this behaviour, as seen in Fig. 11. Cell parameters reported for Ap in the International Crystallographic Standards Database structure file No. 152190 are a = b = 0.94396 nm and c = 0.68792 nm. The experimental value obtained in this work for the parameter a = b for Ap formed in the absence of CHI is smaller than the reported value, and this effect may be caused by the partial substitution of PO4 by CO3. This indicates that the experimentally obtained Ap was a partially car-

472

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

Fig. 6. Thermal analysis curves for composites (a) MD-CHI20, (b) MD-CHI50 and (c) MD-CHI80. ––, TG;   , DTG; – –, DTA.

bonate-substituted type B apatite [78], as already shown by the FTIR spectroscopy results.

Both cell parameters decreased as the CHI content in the experimental composites increased, more markedly for those prepared

473

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476 Table 2 Observed effects for the three stages of thermal decomposition of M-CHI composites. Sample

DTG maximum (°C)

Temperature range (°C)

Mass loss (%)

First stage M-CHI20 M-CHI50 M-CHI80

41 60 59

31–150 33–150 28–150

4.0 6.8 4.8

Second stage M-CHI20 M-CHI50 M-CHI80

270 294 297

200–295 210–338 227–350

14.0 26.2 31.3

Third stage M-CHI20 M-CHI50 M-CHI80

462 483 561

338–600 340–600 350–676

8.8 21.4 40.3

from L-CHI than from M-CHI. As previously observed by Murugan et al. [75] this effect cannot be attributed to the inclusion of CHI in the crystal lattice, due to the relatively large size of CHI molecules. Instead, vacancies and imperfections caused by the decrease in crystallinity should account for the decrease in these parameters. Experiments carried out in silicate-containing hydrolysing solutions led to variations in the parameters a = b and c, suggesting that incorporation of silicate species into the lattice had taken place to some degree. The variation observed in the magnitude of both parameters could not be related to the silicate concentration in the hydrolysing solution. Other authors have described small parameter variations in silicon-substituted HA, attributable to the replacement of PO4 tetrahedra by SiO4 in the crystal lattice. The SiO4 group has a slightly higher volume than PO4 and may produce lattice expansion [79]. On the other hand, some OH groups must be lost in order to maintain the electrical balance of the lattice, causing contraction. The global balance and the effects on the parameters a = b and c will depend upon the degree of incorporation of SiO4 and loss of OH. 3.7. Crystallite size and cell parameters of Ap formed in situ Bone implant materials made of sintered apatite are bioactive but are only slowly reabsorbed in vivo. The relative physiological stability of sintered apatite implants is, amongst other factors, strongly related to its high crystallinity and large crystallite size. Apatites produced by wet methods are more soluble and reabsorbable because of their smaller crystallinity and crystallite size [78]. The in situ procedure employed in this work to produce Ap in the CHI matrix led to crystallite (crystal domain) sizes in the range

Fig. 7. Thermal analysis curves of composites (a) M-CHI20, (b) M-CHI50, (c) M-CHI80 and (d) Ap. ––, TG;   , DTG.

Table 3 Weight losses of composites and experimental compositions determined by thermal analysis. Theoretical composition (%)

Weight loss at degradation stage (%)

Residue at 600 °C (%)

Experimental composition (%)

1st

2nd + 3rd

CHI/Ap 20/80 MD-CHI20 M-CHI20 L-CHI20

6.0 4.0 4.0

17.1 19.6 20.7

76.9 73.1 73.6

18.2/81.8 23.2/76.8 22.0/78.0

CHI/Ap 50/50 MD-CHI50 M-CHI50 M-CHI50-Si100 L-CHI50 L-CHI50-Si100

6.5 6.8 5.4 7.1 6.9

45.3 47.6 51.4 55.6 54.1

48.2 45.0 44.9 37.2 37.9

48.5/51.5 51.4/48.6 53.4/46.6 59.7/40.3 58.8/41.2

CHI/Ap 80/20 MD-CHI80 M-CHI80 L-CHI80

4.9 4.8 5.8

71.0 71.4 83.4

24.1 23.0 10.8

74.7/25.3 75.7/24.3 88.5/11.5

12.9–6.2 nm for the [h 0 0] and [0 k 0] directions and between 19.3 and 11.3 nm in the [0 0 l] direction. This last size is very similar to the value of 18.6–16.4 nm recently reported for natural apatite from mature cow bone [80]. It should be borne in mind that crystal domain or crystallite size is not necessarily related to particle size, hence particles may be composed of several crystallites or crystal domains. According to the a/c ratio of crystallites in the composites obtained individual crystallites are rod-shaped with the z-axis along the rod axis. The study of composite series with varying CHI levels showed that the higher the CHI content in a composite the smaller the average crystallite size of the resultant Ap. These results agree with data reported for CHI– and gelatine–apatite nanocomposites [17,18,75]. Other authors who have observed this effect have not provided any explanation as to its mechanistic causes. A possible explanation could be that the increase in viscosity of the medium leads to a decrease in the diffusion rate of ionic species involved in the nucleation and growth of calcium phosphate phases in the CHI matrix. In this work some experiments were carried out where the Na3PO4/Na5P3O10 hydrolysing solution was replaced by a 1:1 vol.% mixture of Na3PO4 and Na2SiO3 saturated solutions (MCHI50-Si50) or completely by a saturated Na2SiO3 solution (MCHI50-Si100). The aim was to introduce anionic silicate species into the Ap lattice as a route to improving its bioactivity [US Patent No. US6312468], with silicate anions entering the Ap lattice by replacing some of the phosphate ions [79]. It was observed that for those composites hydrolysed in silicate solutions the crystallite size of the resulting Ap decreased with regard to Ap prepared in a CHI-free medium, and the magnitude of decrease was directly related to the silicate content of the hydrolysing solution. Silicon-substituted HA prepared by wet co-precipitation and sintering exhibits a similar effect. A decrease in apparent crystallite size in the [h k 0] direction that was not attributable to microstrain was found by Zou et al. They proposed that as well as an effective decrease in crystallite size, crystal imperfections and defects could also contribute to peak broadening [77]. Bearing in mind that silicate ions replaced phosphate in the apatite lattice and that some hydroxyl groups were lost to minimize the resulting electrical imbalance [79], the presence of defects and some disorder are not unexpected. Thus, in our work the Ap particles contained in the composites prepared in the presence of silicate solutions (M-CHI50-Si50 and M-CHI50-Si100) are expected to be Si-substituted Ap with a nano-

474

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

Table 4 Values of average crystal domain size and cell parameters obtained for in situ formed Ap. Sample

Crystal domain size (nm)a [h 0 0] = [0 k 0]

[0 0 l]

a=b

c

Rwp

Rp

Ap Ap-Si100 L-CHI20 L-CHI50 L-CHI80 M-CHI20 M-CHI50 M-CHI80 M-CHI50-Si50 M-CHI50-Si100

13.4(1) 5.7(1) 11.9(1) 11.0(1) 6.2(1) 12.9(1) 12.9(1) 9.4(1) 10.3(1) 8.6(1)

21.2(1) 11.8(1) 16.8(1) 17.0(1) 11.3(1) 19.3(1) 19.3(1) 15.0(1) 15.1(1) 12.4(1)

0.9430(1) 0.9434(1) 0.9434(1) 0.9421(1) 0.9380(1) 0.9428(1) 0.9425(1) 0.9414(1) 0.9427(1) 0.9418(1)

0.6885(1) 0.6874(1) 0.6892(1) 0.6882(1) 0.6861(1) 0.6888(1) 0.6889(1) 0.6883(1) 0.6886(1) 0.6887(1)

2.89 2.46 2.30 2.01 1.66 2.30 2.22 1.84 2.13 2.18

2.18 1.94 1.80 1.58 1.30 1.80 1.76 1.46 1.67 1.73

a

Cell parameters (nm)a

Fitness parameters (%)

Uncertainty in the last figure is in parentheses.

Fig. 9. Effect of silicate ion addition to the hydrolysis solution on the average crystal domain size of Ap.

Fig. 8. Variation in average crystal domain size of in situ formed Ap with CHI content for composites with (a) low molecular weight CHI and (b) medium molecular weight CHI.

metric crystal size and imperfections that should cause their bioactivity and reactivity to increase. 4. Conclusions A novel procedure was established to obtain size-controlled apatite/CHI nanocomposites. This involved the stepwise co-precipitation of Ap from soluble precursor salts using different proportions of starting ingredients in order to vary the CHI/Ap ratio in the final composite. The lack of any trace of residual DCPD in any sample after hydrolysis confirmed the effectiveness of the proposed synthetic route. The buffer capacity of the Na3PO4 solution employed ensured an almost constant pH of about 12 during the hydrolysis and, in addition, the temperature of 60 °C during this step speeded up the hydrolysis reaction to completion without an appreciable

increase in the crystallinity and particle size of the Ap obtained. The proposed route seems to be highly effective for introducing Si substitution into the crystalline network of the apatite formed in the CHI matrix of the resulting composites. The difference in TG and DTA profiles with change in composition clearly showed that interactions took place between the CHI matrix and the inorganic phase. The increase in CHI content enhanced the thermal stability of composites for all CHI molecular weights studied and all degrees of deacetylation. Structural data obtained from XRD studies of the Ap phase formed in situ showed crystallite (crystal domain) sizes smaller than 20 nm. Furthermore, the procedure employed to produce Ap in the CHI matrix led to a crystal domain in the range 12.9– 6.2 nm for the [h 0 0] and [0 k 0] directions and between 19.3 and 11.3 nm in the [0 0 l] direction, this last size being very similar to the value of 18.6–16.4 nm recently reported for natural apatite from mature cow bone. It was established that the CHI content influences this parameter: the higher the CHI fraction in the composite, the smaller the average crystallite size of the resultant Ap. However, the molecular weight of CHI seems not to affect the average crystal size. Both cell parameters (a = b and c) decreased as the CHI content in the experimental composites increased, more markedly so for those prepared from L-CHI than from M-CHI. The presence of carbonate in CHI/Ap, as detected by FTIR spectroscopy, was a benefit in that it increased the mechanical reliability and bioactivity of the apatite phase and, in this way,

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

475

Acknowledgements We thank the Royal Society UK for funding the International Joint Project ‘‘Development of chitosan/calcium phosphate-based bone implants and scaffolds”. The participation of R.G. Carrodeguas was funded by Project CICYT Spain MAT2006-12749-C02-01. References

Fig. 10. Variation of cell parameters of in situ formed Ap with CHI content for composites with low and medium molecular weight CHI: (a) a = b, (b) c. Values in nm.

Fig. 11. Effect of silicate ion addition to the hydrolysis solution on the cell parameters of Ap.

contributed to an improvement in the osteoconductivity of the material. The use of silicate solution in the hydrolysis step caused a reduction in the crystallite size along all directions studied. It was observed that the decrease was directly related to the silicate content of the hydrolysing solution and dependent upon the molecular weight of CHI (more intense for L-CHI). Ap particles contained in the composites prepared by this means in the presence of silicate solutions are expected to contain ionic substitutions of Si and CO3 for PO4 and Na for Ca and have a nanometric crystal size and imperfections, which should increase their bioactivity and reactivity.

[1] Kokubo T, Kim HM, Kawashita M. Novel bioactive materials with different mechanical properties. Biomaterials 2003;24:2161–75. [2] Uchida A, Araki N, Shinto Y, Yoshikawa H, Ono K, Kurisaki E. The use of calcium hydroxyapatite ceramic in bone tumor surgery. J Bone Joint Surg 1990;72B:298–302. [3] Cooke FW. Ceramics in orthopedic surgery. Clin Orthop Relat Res 1992; 276:135–46. [4] Ogiso M. Reassessment of long-term use of dense HA as dental implant: case report. J Biomed Mater Res B 1998;43:318–20. [5] Hench LL. Bioceramics: from concept to clinic. J Am Ceram Soc 1991; 74:1487–510. [6] Bajpai PK. Ceramics: a novel device for sustained long term delivery of drugs. In: Hulbert JA, Hulbert SF, editors. Bioceramics. Terra Haute, IN: Rose-Hulman Institute of Technology; 1992. p. 87. [7] Hench LL. Bioceramics. J Am Ceram Soc 1998;81:1705–28. [8] Murugan R, Panduranga RK. Controlled release of antibiotic from surface modified coralline hydroxyapatite. Trends Biomater Art Org 2002;16:43–5. [9] Webster TJ, Siegel RW, Bizios R. Enhanced functions of osteoblasts on nanophase ceramics. Biomaterials 2000;21:1803–10. [10] Elliot JC. Structure and chemistry of the apatites and other calcium orthophosphates. Amsterdam: Elsevier Science B.V; 1994. [11] Doi Y, Horiguchi T. Formation of apatite–collagen complexes. J Biomed Mater Res 1996;31:43–9. [12] Miyamato Y, Shikawa KI. Basic properties of calcium phosphate cement containing atelocollagen in its liquid or powder phases. Biomaterials 1998;19:707–15. [13] Itoh S, Kikuchi M, Takakuda K, Nagaoka K, Koyama Y, Tanaka J, et al. Implantation study of a novel hydroxyapatite/collagen (HAp/col) composite into weight-bearing sites of dogs. J Biomed Mater Res 2002; 63:507–15. [14] Zhao F, Yin Y-J, Yao K-D, Guo G, Wang B-L, Zhang J-Y, et al. Tissue engineering study on chitosan–gelatin/hydroxyapatite composite scaffolds – osteoblasts culture. Zhongguo Xiu Fu Chong Jian Wai Ke Za Zhi 2002;16(2):130–3. [15] Murugan R, Panduranga RK. Biodegradable coralline hydroxyapatite composite gel using natural alginate. Key Eng Mater 2002;240–242:407–10. [16] Murugan R, Kumar TSS, Yang F, Ramakrishna S. Hydroxyl carbonate apatite hybrid bone composites using carbohydrate polymer. J Compos Mater 2005;39(13):1159–67. [17] Chang MC, Ko CC, Douglas WH. Preparation of hydroxyapatite–gelatine nanocomposites. Biomaterials 2003;24(17):2853–62. [18] Rusu VM, Ng CH, Wilke M, Tiersch B, Fratzl P, Peter MG. Size-controlled hydroxyapatite nanoparticles as self-organized organic–inorganic composite materials. Biomaterials 2005;26(26):5414–26. [19] Yamaguchi I, Tokuchi K, Fukuzaki H, Koyama Y, Takakuda K, Monma H, et al. Preparation and microstructure analysis of chitosan/hydroxyapatite nanocomposites. J Biomed Mater Res 2001;55(1):20–7. [20] Grodzinski JJ. Biomedical applications of functional polymers. React Funct Polym 1999;39:99–138. [21] Burg KJL, Porter S, Kellam JF. Biomaterial developments for bone tissue engineering. Biomaterials 2000;21:2347–59. [22] Hutmacher DW. Scaffolds in tissue engineering bone and cartilage. Biomaterials 2000;21:2529–43. [23] Yamaguchi I, Tokuchi K, Fukuzaki H, Koyama Y, Takakuda K, Momma H, et al. Preparation and mechanical properties of chitosan/hydroxyapatite nanocomposites. Key Eng Mater 2001;192–195:673–6. [24] Peña J, Izquierdo-Barba I, Martínez A, Vallet-Regí M. New method to obtain chitosan/apatite materials at room temperature. Solid State Sci 2006;8(5):513–9. [25] Wang X, Li Y, Wei J, De Groot K. Development of biomimetic nanohydroxyapatite/poly(hexamethylene adipamide) composites. Biomaterials 2002; 23:4787–91. [26] Murugan R, Panduranga RK. Graft polymerization of glycidylmethacrylate onto coralline hydroxyapatite. J Biomater Sci Polym E 2003;14:457–68. [27] Murugan R, Panduranga RK. Grafting of glycidylmethacrylate upon coralline hydroxyapatite in conjugation with demineralized bone matrix using redox initiating system. Macromol Res 2003;11:14–8. [28] Ito M, Niiro T, Mori K, Yokoyama K, Nakayama Y, Yamagishi T. Relation between mechanical properties of chitosan film and content of hydroxyapatite. Jpn Soc Dent Mater Dev 1994;3:351–7. [29] Varma HK, Yokogawa Y, Espinosa FF, Kawamoto Y, Nishizawa K, Nagata F, et al. Porous calcium phosphate coating over phosphorylated chitosan film by a biomimetic method. Biomaterials 1999;20:879–84. [30] Beppu M, Santana CC. In vitro mineralization of chitosan. Key Eng Mater 2001;192–195:31–4.

476

N. Davidenko et al. / Acta Biomaterialia 6 (2010) 466–476

[31] Hu QL, Li BQ, Wang M, Shen JC. Preparation and characterization of biodegradable chitosan/hydroxyapatite nanocomposite rods via in situ hybridization: a potential material as internal fixation of bone fracture. Biomaterials 2004;25(5):779–85. [32] Fulmer MT, Brown PW. Hydrolysis of dicalcium phosphate dihydrate to hydroxyapatite. J Mater Sci Mater Med 1998;9(4):197–202. [33] Kumar M, Dasarathy H, Riley C. Electrodeposition of brushite coatings and their transformation to hydroxyapatite in aqueous solutions. J Biomed Mater Res 1999;45(4):302–10. [34] Kumar M, Xie J, Chittur K, Riley C. Transformation of modified brushite to hydroxyapatite in aqueous solution: effects of potassium substitution. Biomaterials 1999;20(15):1389–99. [35] Martin RI, Brown PW. Hydrolysis of CaHPO4 in sodium fluoride solutions at 37.4 °C. Caries Res 1998;32(5):365–77. [36] Martin RI, Brown PW. Effects of sodium fluoride, potassium fluoride and ammonium fluoride solutions on the hydrolysis of CaHPO4 at 37.4 °C. J Cryst Growth 1998;183(3):417–26. [37] García Carrodeguas R, Morejón-Alonso L, Delgado García-Menocal JA, Morejón Alonso L, Ginebra Molins MP, Martínez Manent S, et al. Hydrothermal method for preparing calcium phosphate monoliths. Mater Res 2003;6(3):395–401. [38] Best SM, Zou S, Brooks R, Huang J, Rushton N, Bonfield W. The osteogenic behaviour of silicon substituted hydroxyapatite. Key Eng Mater 2008;361– 363:985–8. [39] Thian ES, Huang J, Best SM, Barber ZH, Bonfield W. Silicon-substituted hydroxyapatite: the next generation of bioactive coatings. Mat Sci Eng C 2007;27(2):251–6. [40] Dorozhkin SV. In vitro mineralization of silicon containing calcium phosphate bioceramics. J Am Ceram Soc 2007;90(1):244–9. [41] Botelho CM, Brooks RA, Best SM, Lopes MA, Santos JD, Rushton N, et al. Human osteoblast response to silicon-substituted hydroxyapatite. J Biomed Mater Res A 2006;79A(3):723–30. [42] Thian ES, Huang J, Vickers ME, Best SM, Barber ZH, Bonfield W. Siliconsubstituted hydroxyapatite (SiHA): a novel calcium phosphate coating for biomedical applications. J Mater Sci 2006;41(3):709–17. [43] Porter AE, Patel N, Skepper JN, Best SM, Bonfield W. Effect of sintered silicatesubstituted hydroxyapatite on remodelling processes at the bone–implant interface. Biomaterials 2004;25(16):3303–14. [44] Porter AE, Patel N, Skepper JN, Best SM, Bonfield W. Comparison of in vivo dissolution processes in hydroxyapatite and silicon-substituted hydroxyapatite bioceramics. Biomaterials 2003;24(25):4609–20. [45] Gibson IR, Hing KA, Revell PA, Santos JD, Best SM, Bonfield W. Enhanced in vivo response to silicate-substituted hydroxyapatite. Key Eng Mater 2002;218– 2:203–6. [46] Patel N, Best SM, Bonfield W, Gibson IR, Hing KA, Damien E, et al. A comparative study on the in vivo behavior of hydroxyapatite and silicon substituted hydroxyapatite granules. J Mater Sci Mater Med 2002; 13(12):1199–206. [47] Langstaff S, Sayer M, Smith TJN, Pugh SM. Resorbable bioceramics based on stabilized calcium phosphates. Part II: evaluation of biological response. Biomaterials 2001;22(2):135–50. [48] Carlisle EM. Biochemical and morphological changes associated with long bone abnormalities in silicon deficiency. J Nutr 1980;110:1046–55. [49] Carlisle EM. A silicon requirement for normal skull formation. J Nutr 1980;110:352–9. [50] Hench LL. Glass and genes: The 2001 W.E.S. Turner Memorial Lecture. Glass Technol 2003;44(1):1–10. [51] Bielby RC, Pryce RS, Hench LL, Polak JM. Enhanced derivation of osteogenic cells from murine embryonic stem cells after treatment with ionic dissolution products of 58S bioactive sol–gel glass. Tissue Eng 2005;11(3–4):479–88. [52] Xynos ID, Edgar AJ, Buttery LDK, Hench LL, Polak JM. Gene-expression profiling of human osteoblasts following treatment with the ionic products of Bioglass 45S5 dissolution. J Biomed Mater Res 2001;55:151–7. [53] Botelho CM, Brooks RA, Best SM, Lopes MA, Santos JD, Rushton N, et al. Biological and physical–chemical characterization of phase pure HA and SIsubstituted hydroxyapatite by different microscopy techniques. Key Eng Mater 2004;254–2:845–8. [54] Gibson IR, Best SM, Bonfield W. Effect of silicon substitution on the sintering and microstructure of hydroxyapatite. J Am Ceram Soc 2002;85(11):2771–7. [55] Porter AE, Best SM, Bonfield W. Ultrastructural characterisation of hydroxyapatite and silicon-substituted hydroxyapatite. Key Eng Mater 2003; 240–2:505–8.

[56] Porter AE, Best SM, Bonfield W. Ultrastructural comparison of hydroxyapatite and silicon-substituted hydroxyapatite for biomedical applications. J Biomed Mater Res A 2004;68A(1):133–41. [57] Porter AE, Botelho CM, Lopes MA, Santos JD, Best SM, Bonfield W. Ultrastructural comparison of dissolution and apatite precipitation on hydroxyapatite and silicon-substituted hydroxyapatite in vitro and in vivo. J Biomed Mater Res A 2004;69A(4):670–9. [58] Thian ES, Huang J, Best SM, Barber ZH, Bonfield W. Silicon-substituted hydroxyapatite thin films: effect of annealing temperature on coating stability and bioactivity. J Biomed Mater Res A 2006;78A(1):121–8. [59] Thian ES, Huang J, Best SM, Barber ZH, Bonfield W. Novel silicon-doped hydroxyapatite (Si-HA) for biomedical coatings: an in vitro study using acellular simulated body fluid. J Biomed Mater Res B 2006;76B(2):326–33. [60] Thian ES, Huang J, Best SM, Barber ZH, Brooks RA, Rushton N, et al. The response of osteoblasts to nanocrystalline silicon-substituted hydroxyapatite thin films. Biomaterials 2006;27(13):2692–8. [61] Vandiver J, Dean D, Patel N, Botelho C, Best S, Santos JD, et al. Silicon addition to hydroxyapatite increases nanoscale electrostatic, van der Waals, and adhesive interactions. J Biomed Mater Res A 2006;78A(2):352–63. [62] Chang MC, Tanaka J. XPS study for the microstructure development of hydroxyapatite–collagen nanocomposites crosslinked using glutaraldehyde. Biomaterials 2002;23:3879–85. [63] Engel GE, Wilke S, Konig O, Harris KDM, Leusen FJJ. PowderSolve – a complete package for crystal structure solution from powder diffraction patterns. J Appl Crystalogr 1999;32:1169–79. [64] Pawley GS. Unit-cell refinement from powder diffraction scans. J Appl Crystalogr 1981;14:357–61. [65] Klug HP, Alexander LE. X-ray diffraction procedures. 2nd ed. New York: John Wiley & Sons; 1974. [66] Wilson RM, Elliott JC, Dowker SEP, Rodriguez-Lorenzo LM. Rietveld refinements and spectroscopic studies of the structure of Ca-deficient apatite. Biomaterials 2005;26:1317–27. [67] Fowler BO. Infrared studies of apatites. I. Vibrational assignments for calcium, strontium, and barium hydroxyapatite utilizing isotopic substitution. Inorg Chem 1974;13:194–207. [68] Pena J, Izquierdo-Barba I, Martinez A, Vallet-Regi M. New method to obtain chitosan/apatite materials at room temperature. Solid State Sci 2006;8(5):513–9. [69] Mery JC, Gibson IR, Best S, Bonfield W. Synthesis and characterization of carbonate hydroxyapatite. J Mater Sci Mater Med 1998;9:779. [70] Walsh WR, Labrador DP, Kim HD, Guzelsu N. The effect of in vitro fluoride ion treatment on the ultrasonic properties of cortical bone. Ann Biomed Eng 1994;22(4):404–15. [71] Argüelles-Monal W, Peniche-Covas C. Study of interpolyelectrolyte reaction between chitosan and carboxymethyl cellulose. Makromol Chem Rapid Commun 1988;9:693–7. [72] Botelho CM, Lopes MA, Gibson IR, Best SM, Santos JD. Structural analysis of Sisubstituted hydroxyapatite: zeta potential and X-ray photoelectron spectroscopy. J Mater Sci Mater Med 2002;13(12):1123–7. [73] Leventouri T, Bunaciu CE, Perdikatsis V. Neutron powder diffraction studies of silicon-substituted hydroxyapatite. Biomaterials 2003;24(23):4205–11. [74] García I, Peniche C, Nieto JM. Determination of the degree of acetylation of chitin and chitosan by thermal analysis. J Thermal Anal 1983;28:189–93. [75] Murugan R, Ramakrishna S. Bioresorbable composite bone paste using polysaccharide based nano hydroxyapatite. Biomaterials 2004;25(17): 3829–35. [76] Redepenning J, Venkataraman G, Chen J, Stafford N. Electrochemical preparation of chitosan/hydroxyapatite composite coatings on titanium substrates. J Biomed Mater Res A 2003;66A(2):411–6. [77] Zou S, Huang J, Best S, Bonfield W. Crystal imperfection studies of pure and silicon substituted hydroxyapatite using Raman and XRD. J Mater Sci Mater Med 2005;16(12):1143–8. [78] LeGeros RZ. Calcium phosphates in oral biology and medicine. Basel: Kärger; 1991. [79] Gibson IR, Best SM, Bonfield W. Chemical characterization of siliconsubstituted hydroxyapatite. J Biomed Mater Res 1999;44:422–8. [80] Danilchenko SN, Kukharenko OG, Moseke C, Protsenko IY, Sukhodub LF, Sulkio-Cleff B. Determination of the bone mineral crystallite size and lattice strain from diffraction line broadening. Cryst Res Technol 2002;37(11): 1234–40.