Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Water-based synthesis and characterization of sol-gel hydroxyapatite ceramics Liu, Dean-Mo 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2004-902218.pdf [ 12.12MB ]
Metadata
JSON: 831-1.0078658.json
JSON-LD: 831-1.0078658-ld.json
RDF/XML (Pretty): 831-1.0078658-rdf.xml
RDF/JSON: 831-1.0078658-rdf.json
Turtle: 831-1.0078658-turtle.txt
N-Triples: 831-1.0078658-rdf-ntriples.txt
Original Record: 831-1.0078658-source.json
Full Text
831-1.0078658-fulltext.txt
Citation
831-1.0078658.ris

Full Text

Water-Based Synthesis and Characterization of Sol-Gel Hydroxyapatite Ceramics B y Dean-Mo L i u M . S., Department of Materials Science and Engineering, Virginia Polytech Institute and State University, 1991 M . S. Department of Chemical Engineering, Chung-Yuan Christian University, 1986 B . S., Department o f Chemical Engineering, Chung-Yuan Christian University, 1984 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T F O R T H E D E G R E E O F D O C T O R OF P H I L O S O P H Y in T H E F A C U L T Y OF G R A D U A T E S T U D I E S (The Department of Metals and Materials Engineering) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A May, 2004 © Dean-Mo L i u , copyright, 2004 A B S T R A C T Sol-gel technology for hydroxyapatite (HAp) synthesis has been developed more than a decade ago. The existing sol-gel HAp synthetic methods which used alkoxide-based precursors require solvent-based diluting media, a moisture-controlled atmosphere,! prolonged synthesis time generally greater than 24 hours, and a heat treatment at temperature > 500°C over a time period of several hours. Those factors limit practical applications df HAp, in particular as coatings. Therefore, in this work, HAp ceramics were synthesized using a novel water-based sol-gel route with triethyl phosphite and calcium nitrate as phosphorous and calcium precursors, respectively. The use of water as the only diluting medium and the alkoxide-metal salt combination as the precursors have never been reported in the literature. It was also expected that the novel sol-gel process developed in this work can be a feasible materials technology for biomedical applications, particularly coatings. Processing design in terms of time and temperature for both hydrolysis and ageing steps of the sol-gel synthesis was thoroughly investigated and successfully optimized for phase-pure HAp formation. This also allows further consolidation of a phase evolution map, which can instructively be used as a guidance for the synthesis of the sol-gel HAp at relatively low temperature. It is found that the novel sol-gel process currently developed for H A p provides advantages, included environmentally-friendly, cost-effective, and user-friendly over existing technologies. The process has been granted US patent number 6426114 B l , 2003 and US patent #2002155144 and is now under use in the biomedical industry. ii Table of Contents Abstract List of Figures List of Tables ix List of Symbols x List of Acronyms xi Acknowledgements x i i Chapter 1 Introduction 1 Chapter 2 Literature Review 5 Chapter 3 Objectives and Scope 29 Chapter 4 Experiments 32 Chapter 5 Results and Discussion 40 5-1 Process Development 40 5-2 Effect of Hydrolysis 54 5- 3 Effect of Ageing of Sol 70 Chapter 6 Sol-Gel Hydroxyapatite Coatings 89 6- 1 Characterization of the Coatings 89 6-2 In-Vitro Test 105 6-3 Hydroxyapatite Coatings for Drug Delivery 119 Chapter 7 Conclusions 127 Chapter 8 Future Work 132 References 133 Thesis-related Publications 147 iii List of Figures Figure 2-1 Hydroxyapatite structure projected down the c-axis. The a and b axes, intersecting at 120°, are perpendicular to the c axis. 6 Figure 2-2 X-ray diffraction patterns of apatite from different sources. 7 Figure 4-1 Flowchart of the water-based sol-gel H A p synthesis. 35 Figure 5-1-1 Thermogravimetric analysis of both N A and N E gels, showing a similar weight loss behavior. 44 Figure 5-1-2 The X R D patterns for (a) NA-gel and (b) NE-gel under different calcination temperatures. A predominant apatitic phase is obtained, with a minor amount of TCP phase which developed at higher temperature. A n X R D pattern (top curve in Fig. 5-l-2a) of commercial HAp is also provided for comparison purpose. 46 Figure 5-1-3 Differential thermal analysis of N A and N E gels. 48 Figure 5-1-4 Crystal size of the N A and N E gels, calculated based on the Scherrer equation, calcined at different temperatures. 52 Figure 5-1-5 Scanning electron micrograph of the microstructure for the NA-gel calcined at 800°C, an average size of approximately 200 nm, in agreement with calculation, was observed. Some sintering behavior is also seen (as arrows indicated). 53 Figure 5-2-1 X R D patterns for the gels calcined at 375°C for different hydrolysis time of the phosphite sol, without acid catalyst. 55 Figure 5-2-2 Thermal gravimetric analysis of the dried gels prepared under different conditions (with and without acid catalyst) of hydrolysis of the phosphite. —56 Figure 5-2-3 X R D patterns for the calcined gels derived from the phosphite sol hydrolyzed for lh , 4h, and 24 h. A small amount of T C P was also detected in all cases.-59 Figure 5-2-4 pH change in phosphite sol solution during hydrolysis. 60 iv Figure 5-2-5 Solution pH in a mixed sol containing both Ca and P precursors after 24 h of ageing. 62 Figure 5-2-6 X R D patterns of the gels hydrolyzed for 5 min. and calcined at 375°C for 2 hrs. The gels were prepared under the presence of nitric acid catalyst of different concentrations. Commercial H A is given for comparison purpose. 68 Figure 5-2-7 The morphology of the gel powders prepared (a) with and (b) without acid catalyst, after calcination at 375°C. — ~ ~ 69 Figure 5-3-1 X R D patterns of the 400°C/2 h calcined gels, prepared from different ageing time periods. Impure TCP phase appeared after aging for greater than 16 h.-73 Figure 5-3-2 X R D patterns of the gels, prepared from different ageing time periods, after 500°C calcinations, showing better crystallinity. 74 Figure 5-3-3 X R D patterns of the dried gels after (a) 4 h and (b) 16 h of ageing at ambient temperature. 75 Figure 5-3-4 Dried gels prepared after 4 h and 16 h ageing periods show different weight loss and thermal behaviors under (A) thermogravimetric and (B) differential thermal analysis. 77 Figure 5-3-5 X R D patterns of gels, derived from sols after different ageing periods at 45°C, after 400°C calcinations. A decrease in apatite associated with a corresponding increase in TCP was observed as ageing time increased. 79 Figure 5-3-6 X R D patterns of the gels, derived from 24 h-aged sol at 45°C, prepared after (A) 25°C (dried gel), (B) 400°C, and (C) 500°C calcinations. A considerable amount of the apatite appeared at 500°C. 80 Figure 5-3-7 Phase evolution map for apatite formation in terms of ageing time and ageing temperature. 83 Figure 5-3-8 Ageing kinetic shows an exponential correlation with time and turns to be more pronounced with ageing temperature. The solid curves are obtained through the used of least-squared fit of Eqn.5-3-6. The correlation coefficients indicated in parenthesis. 85 v Figure 5-3-9 Arrhenius plot of the rate constant vs. ageing temperature shows good correlation with an apparent activation energy calculated to be 10.35 kcal/mole. 86 Figure 5-3-10 Solution pH, which is most suitable for apatite formation under current aqueous sol-gel system, can be determined as highlighted between two dash lines, based upon the phase evolution map depicted in Fig. 5-3-7. 88 Figure 6-1-1 X-ray diffraction patterns of the sol-gel HAp coating annealed in air at various temperatures. 89 Figure 6-1-2 Fourier transform infrared spectra for the sol-gel HAp coatings prepared at different temperatures. 91 Figure 6-1-3 Surface microstructure of the coatings onto the sandblasted 316 stainless steel substrates annealed at (a) 375°C, (b) 400°C, and (c) 500°C. 94 Figure 6-1-4 Cross-section view of the coatings annealed at 375°C shows a uniform thickness of about 0.6 pm.. 95 Figure 6-1-5 Cross-section view of the coatings annealed at 375°C shows the ability of the sol-gel method to coat cavities/grooves of different geometry on the substrates. 96 Figure 6-1-6 Cross-section view of the 5 00°C-annealed sol-gel H A p coating.. 97 Figure 6-1-7 Some sparsely distributed nano-scale pores were found within the HAp coating annealed at 500°C. 97 Figure 6-1-8 The bonding strength of the sol-gel H A p coatings on 316 SS prepared at different annealing temperatures. Bars represent the standard deviation of the measurement. 99 Figure 6-1-9 The morphology of the porous surface dental implant (a) before and (b) after sol-gel H A p coating. A closer look of the coated surface (c) shows pores and cracks (arrows indicated) distributed along the neck and triple-junction areas. The surface feature of the underlying T i beads can be observed and reproduced, indicating a thin, dense, and adhesive H A p coating.. 102 vi Figure 6-2-1 Surface microstructural evolution of in-vitro calcium phosphate layer on the 400-ACS substrate for an immersion time period of (a) 3, (b) 7, and (c) 14 days. Bar: 5 pm.. 106 Figure 6-2-2 C P L deposit layer thickness as a function of time of immersion, showing a linear relationship for both substrates. 107 Figure 6-2-3 The variation of the Ca/P ratio along the C P L deposit layer. 108 Figure 6-2-4 The pH of the simulated body fluid increases with time of immersion and reaches plateau after about 10 days of test. 109 Figure 6-2-5 Surface morphology of the C P L deposit layer, 14-day immersion in SBF.--110 Figure 6-2-6 Surface microstructure of the deposit layer in-vitro on (a) 375-ACS and (b) 400 A C S substrates, after 14-day immersion in SBF. 111 Figure 6-2-7 Cross-sectional view of the deposit C P L layer on (a) 375-ACS and (b) 400-ACS substrates, after 14-day immersion in SBF. 111 Figure 6-2-8 FTIR spectra for the (a) 375-ACS and (b) 400-ACS substrates before and after the in-vitro immersion test for a time period of 14 days. 112 Figure 6-3-1 The CPC-HAp layer showed extensive debonding for the model substrate without pre-SG-HAp-coated treatment; however, for the substrate with the SG-HAp pre-treatment, a well attached CPC layer was occurred. Both CPC layers have the same thickness of about 20 um. 120 Figure 6-3-2 X R D pattern of the CPC layer (bottom curve), showing a poorly-crystalline apatitic structure. The X R D pattern of the commercial H A p (top curve) is used for comparison purpose. 121 Figure 6-3-3 Cross-section examination of the CPC-HAp coating confirms the development of a well-bonded SG-HAp - CPC-HAp, (as well as an adhesive SG-HAp -substrate interface), resulting in an interfacial strength of 6 MPa. Infrared analysis (b) shows that the CPC-HAp layer is essentially a calcium-deficient, carbonate apatite, closely resembling that of human bone mineral.. 123 vii Figure 6-3-4 A sustained release of the model drug (amethoperin, 5% by weight with respect to the weight of the CPC-HAp layer, 20 mg) from the C P C - H A p coating for a time period of 72 h. A rapid release (the "burst" effect), was detected for the first 8 h, following a t 1 / 2 - dependent diffusion behavior. With a post-coated PVA film on the CPC-HAp layer, the burst effect was reduced to a certain extent, indicating the release pattern can be adjusted for practical needs. 126 viii List o f Tables Table 2-1: Chemical precursors, mutual solvents, and synthesis parameters, i.e., temperature and time, required to form phase-pure crystalline H A p in various sol-gel H A P processes developed over the recent decade. 18 ix List of Symbols A(26) peak width at half maximum intensity of the reflection (002) of H A p X wavelength for C u K a , [nm] D crystal size, [nm] ki apparent rate constant, [mol/s] n' and m' reaction order t time of reaction, [second] E apparent activation energy, [kcal or kJ] R gas constant, [4.18 cal/mol] T absolute temperature, [K] HCPL deposit layer thickness, [pm] k' deposit rate constant, [um/h] R growth rate of deposition, [um/h] s parameter proportional to number of growth site a relative supersaturation ratio AV/At weight change of particulate substrate, [g/h] m0 initial weight of the particulate substrate, [g] As surface area of the substrate, [cm2] C effective concentration of the solution with respect to the deposit phase, [mole/L] x List of Acronyms HAp: Hydroxyapatite TCP: Tricalcium phosphate A C P : Amorphous Calcium Phosphate BCP: Biphasic Calcium Phosphate PP: Pyrophosphate C N : Calcium Nitrate SG: Sol-Gel Process SG-HAp: Sol-Gel Synthesized Hydroxyapatite CPC: Calcium Phosphate Cement CPC-HAp: Hydroxyapatite film derived from calcium phosphate salt precursors DTA: Differential Thermal Analysis T G A Thermogravitational Analysis N M R : Nuclear Magnetic Resonance SBF: Simulated body fluid; fluid with ion concentration mimics that of human blood plasma. CPL: Calcium phosphate layer, deposited from immersion of a substrate with SBF. xi ACKNOWLEDGEMENTS I am gratefully indebted to the supervision of Prof. Tom Troczynski, Department of Metals and Materials Engineering and valuable comments from Prof. Helen Burt, Department of Pharmaceutical Science, and Prof. Thomas Oxland, Department of Mechanical Engineering, during the research period. I also give my thanks to Dr. Quanzu Yang, Research Associate, UBCeram, for his assistance in some experiments during the period. xii Chapter 1: Introduction Calcium phosphate ceramics, particularly those with Ca/P ratios between 1.5 and 1.67, i.e., tricalcium phosphate (Ca3(P04)2) and stoichiometric calcium hydroxyapatite (Caio(P04)6(OH)2), respectively, have long been used as prime candidate biomaterials to restore, re-construct, and replace human bone tissues. This is because of their close chemical similarity to minerals found in calcified tissues, such as bone, tooth enamel and dentine, in humans and vertebrates. Hydroxyapatite (HAp) is one of the most important calcium phosphates and has received wide attention as bone substitutes over the past few decades [1-7]. One remarkable property of HAp is the close resemblance in chemical and crystallographic structure to that of the mineralized component in vertebrate teeth and bones [8-12]. HAp has been the most widely studied calcium phosphate since it demonstrated excellent biological affinity and activity to host tissue, e.g., by forming chemical bond at the implant-tissue interface. Natural bone mineral is non-stoichiometric and contains H P 0 4 2 " and CO3 2 " groups that potentially replace the PO4 3" and OH" groups in the apatite lattice, so its chemical composition could be formulated as Caio-x(P04)6-y(HP04)y(C03)(OH)2-x-y/3. Thus, naturally occurring apatite is essentially a carbonate-containing apatite and has a nanometered-size, poorly crystalline feature where metabolic activity with respect to the surrounding physiological environment is higher than for conventional, synthetic, well-crystallized HAp. Osteoconductivity, i.e., the ability to "guide" the growth of bone tissue, is one of the major biological functions of the HAp that has been observed in animal models. 1 Osteoinductivity (the ability to "induce" the formation of bone, even within soft tissues such as muscles) of H A p has also been mentioned in a number of reports [13-16]. A l l these provide an extremely strong rationale for the HAp as a prime candidate for biomedical applications. HAp is also a strong candidate for replacement, restoration, and regeneration of defective or disordered bones or hard tissues due to traumatic or nontraumatic events. H A p offers several advantages over other synthetic or naturally-occurring materials, such as autografts (i.e., hard tissue from humans), xenograft (i.e. hard tissues from animals), and synthetic substitutes like metals, polymers or ceramics. These include: 1. easy to synthesize in large quantity and at low cost, in comparison with autografts. 2. no risk of disease transmission and host body rejection such as from xenografts, 3. negligible wear or adverse tissue response, e.g., side effects, in comparison with metals and polymers. 4. superior biocompatibility and bioactivity to human tissues as compared to those structural ceramics such as AI2O3, ZrCh [17,18] (which are classified as bioinert materials) although these ceramics generally exhibit better mechanical properties. Accordingly, it is desirable to produce HAp with controlled Ca/P molar ratio in a range between 1.5 and 1.67, and structurally resembling that of natural apatite. Most existing synthetic strategies require high-temperature treatment to develop well-crystalline H A p in bulk, particulate, or coating form. However, well crystalline H A p may not be an advantage in terms of biological activity, such as bioresorption, as compared to natural biocrystals. In 2 contrast, biocrystals are nano-crystalline and also poorly crystalline structure, and are developed under physiological environment. Therefore, it is expected that the lower the temperature for the synthesis of apatite, the more resemblance in microstructure to that of natural apatite would occur. With the advancement of materials technology, a soft chemical route, termed "sol-gel process" (referred to as SG-HAp), to synthesize H A p has been proposed since 1990 and received considerable attention till today. Sol-gel process has several advantages, i.e., lower processing temperature, greater chemical and physical homogeneity and wider shape-forming capability. Therefore it would be favorable i f the sol-gel route could be applied to the synthesis of HAp. It is therefore the main purpose of this work to employ the sol-gel route for H A p synthesis. The common features observed from those existing sol-gel H A p synthesis, were solvent-based diluting media and medium-to-elevated heat treatment temperature, e.g., 500°C - 900°C, for a time period of several hours and prolonged sol preparation, e.g., 24 h were reported. Those processing features may cause adverse effect such as environmental pollution, energy consuming, and when applied for coating application onto metallic substrate, heating at elevated temperatures over few-hour duration may cause undesirable degradation of mechanical property of the underlying metallic devices as a result of phase transformation, oxidation or undesired chemical interaction between the metal substrate and chemical precursors. Therefore, it is more desirable to find a new sol-gel synthesis process that enables HAp phase to be developed at relatively low temperatures, either as powder form or as coating 3 onto metallic substrates, without causing adverse impact on the environment. With this in mind, in this work, we attempt to find an alternative process using water as the only diluting medium, lower synthesis temperatures and shorter period of synthesis time. This newly-developed sol-gel HAp process is subsequently applied to form coatings on metallic substrates. The coating properties included interfacial strength, interfacial structure, and in-vitro activity are systematically investigated. 4 Chapter 2: Literature Review • Chemical and structural resemblances between the natural apatite (i.e. mineral i component of bone) and synthetic HAp allow strong chemical bonding to occur at the material-bone interface [17]. This results in a strong fixation of HAp implants [31,32] which is particularly critical over the early stage of post-implantation. The mineral phase in the natural bone consists mainly of -70 wt% of poorly-crystalline apatite and a small amount of carbonated apatite. The poorly-crystalline HAp, such as naturally-occurring apatite, is prone to be more soluble in physiological environment than the highly-crystalline HAp which is commonly considered to be non-resorbable. However, due to brittleness of pure synthetic HAp ceramics, either composite [33] or coating [34] applications are frequently employed. 2-1 H A p Structure The similarity in crystallographic structure between bone mineral and hydroxyapatite (HAp) was first observed in 1926 by DeJong [35]. Later, a refinement of the spatial arrangement of the constituent groups of Ca + 2 , PO4" 3, and OH" ions in HAp structure was further given by Posner et al. [36]. HAp has a hexagonal spatial symmetry and the unit cell for HAp is a right, rhombic prism with a length along each edge of the basal plane of the cell of a = 0.9432 nm and a height of c = 0.6991 nm. 5 Figure 2-1 shows the arrangement of the constituent atoms as projected along the c axis onto the basal plan of the HAp. The O H ions lie at the corners of the rhombic base of the unit cell, at equal distance of one-half the height of the cell along columns parallel to the c axis. Six calcium ions in the cell form two triangles rotated by 60° with respect to each other, centered on and perpendicular to the axis of OH" group. The other four calcium ions in the unit cell lie along separate columns along the c-axis. These columnar calcium ions are coordinated by oxygens from orthophosphate tetrahedral, i.e., PO4" 3, that construct the bulk structure of the unit cell. Figure 2-1 Hydroxyapatite structure projected down the c-axis. The a and b axes, intersecting at 120°, are perpendicular to the c axis. [37] 6 One important feature of the HAp structure is that it is easily subject to isomorphous substitution. Ca positions can be substituted by numerous divalent or trivalent cations. Some anions such as F", CI", CO3" 2 , etc, are also known to replace OH" as well as PO4" 3 ions to form isomorphs [38]. In bone apatite, cationic substitution by ions, such as Mg , Sr, F, Pb, is limited in quantity. However, carbonate substitution is more critical as it is the third most abundant ion in bone apatite. Figure 2-2 shows X-ray diffraction patterns of H A p from different sources. Amorphous characteristic is a microstructural feature of the natural bone HAp, which is due Figure 2-2. X-ray diffraction patterns of powdered human bone femur diaphysis (bottom curve); poor-crystalline, synthetic HAp with small crystals (middle curve); and well-crystalline, synthetic HAP, with reflection peaks indexed (top curve) [39]. 7 primarily to both nanometric crystallite and poor crystalline structure, in comparison to crystalline, synthetic HAp. The former has been known to offer more biological activity than the latter in physiological environment [39]. 2-2 Conventional Techniques for HAp Synthesis Numerous routes have been developed to synthesize HAp. For instance, hydrothermal preparation uses various calcium phosphate precursors such as monocalcium phosphate, dicalcium phosphate as starting chemicals [40]. Pressure and temperature within the reactor need to be suitably manipulated in order to convert the intermediate products such as P-calcium pyrophosphate, calcium monohydrogen phosphate into H A p phase. However, among the numerous methods available for HAp powder syntheses, wet precipitation is the most frequently used method. One aqueous solution containing calcium salt is mixed dropwise with phosphorus-containing solution. Upon mixing, chemical reactions take place between calcium and phosphorus ions under a controlled pH and temperature of the solution. The resulting white precipitated powder is typically calcined at 400-600°C (or even at higher temperature) in order to obtain a stoichiometric, apatitic structure. In some cases, well-crystallized H A p phase was developed after sintering at temperatures of 1000-1200°C. However, fast precipitation during phosphate solution titration (to calcium solution) leads to chemical inhomogeneity in the final product. Slow titration and diluted solutions must be used to improve chemical homogeneity and stoichiometry of the resulting HAp. Careful control of the solution condition is critical in the wet precipitation. Otherwise, a decrease of 8 solution pH below about 9 could lead to the formation of Ca-deficient H A p structure. Another low-temperature synthesis route is through hydrolysis of suitable calcium phosphate precursors such as dicalcium phosphate dihydrate (CaHP04.2H20) [41] or B-tricalcium phosphate (B-Ca3(P04)2) [42] in a water-containing environment, wherein the resulting HAp showed good crystallinity. Unfortunately, H A p is inherently brittle and shows relatively poor mechanical strength, which further restricts its applications to those areas that require little or no load-bearing ability such as mandibular and maxillary augmentation, etc. H A p has a fracture toughness not exceeding about 1.0 MPa.mVi, in comparison to 2-12 MPa.m'/i for human bones. The Weibull modulus of HAp has been determined in the range of 5-18, suggesting a continuous improvement of the reliability of HAp implants may still be required_[43,44]. Therefore, over the past decades, a number of techniques and concepts were employed to fabricate HAp-based composites by using secondary reinforcements such as ceramic/metallic fibers/whiskers [45,46], platelets [47], or tougher ceramic particulates [48,49]. One attractive alternative that has received considerable attention is through coating. By taking advantage of the bioactivity of HAp, together with the superior mechanical properties of biocompatible metals (mainly Ti , Ti-6A1-4V, Co-Cr-Mo alloy, and 316 stainless steel), HAp-coated implants have enjoyed commercial and clinical success for many years. 9 2-3 Hydroxyapatite Coatings Several techniques have been developed for deposition of H A p coatings, namely, electrophoretic deposition [50,51], sputtering [52], hot-isostatic pressing [53], thermal spraying [21,22, 34, 54, 55,56], electrochemical deposition [57-59], laser deposition [60], metallorganic vapor deposition [61], and sol-gel process [23-25, 62-77]. One of the most frequently used methods for deposition of HAp is plasma (or thermal) spraying. Plasma spraying is a violent, difficult-to-control, and line-of-sight process. The starting HAp powders are subject to extremely high temperatures during the deposition. For instance, a temperature typically ranging from 6,000°C to 10,000°C is applied to melt the HAp powders before impinging onto cold metallic substrates. Once the partially melted H A p powder reaches the substrate, it solidifies at an extremely rapid rate. One advantage of the plasma spraying is that no subsequent sintering of the coating is needed. However, such a violent process causes significant de-stabilization of the plasma sprayed HAp structure. Numerous reports [21, 22, 34, 54, 55, 56] indicated decomposition of the sprayed H A p to form other phases, such as tetracalcium phosphate, tricalcium phosphate, calcium oxide, and amorphous calcium phosphate. A subsequent heat treatment has usually been applied in order to restore crystalline HAp phase in the final product. However, some of these compounds are undesirable in vivo. Additionally, severe cracking of the plasma-sprayed layer frequently causes accelerated failure of implants. The plasma-sprayed coatings are relatively thick, typically ranging from 10 20 um to 200 pm. As it is line-of-sight process, only those surface areas that are visible to the plasma stream can be covered. Calcium hydroxyapatite, Caio(P04;6(OH)2, coatings on metallic or ceramic substrates have been subject of many studies because of their chemical similarity to mineralized phase in human hard tissues. This enables affinity between HAp and surrounding host tissue, possibly including formation of a chemical bond [17,19,20]. Usually H A p coatings are thermally processed, such as thermal/plasma sprayed [21,22] and more recently through sol-gel route [23,24,25]. The resulting synthetic apatite is usually having better crystallinity, coarser grain, and more chemical stoichiometry, as compared to that of natural bio-crystals. The natural apatite is poorly crystalline, finely divided ("nano-structured"), non-stoichiometric and includes carbonate dopant. Such fine, non-stoichiometric nature makes bio-crystals more efficient in their metabolic interactions with surrounding physiological environments, as compared to the synthetic apatite crystals. Therefore, it is desirable to synthesize apatite with a structure closely resembling that of bio-crystals. This prompted work on new techniques for deposition of H A p in recent years, in particular room-temperature biomimetic methods. These man-made processes simulate growth of HAp bio-crystals that naturally develop in the physiological environment [26-30]. For instance, Wen et al. [27] employed a two-step chemical process to achieve biological affinity of metallic surfaces (such as Ti and its alloys) allowing subsequent deposition of calcium phosphate layer in physiological solution. Although a relatively fast deposition of the calcium phosphate of 1 pm/h could be reached, the 11 resulting phase was macro-porous and mostly octacalcium phosphate rather than apatite. A further modification of this process resulted in dense apatite layer grown at about 0.4 um/day [26]. A n alternative technique called surface-induced mineralization has recently been reported by Campbell et al. [28], who modified metallic surface with a mono-layer assembled short-chain polymer, following immersion into physiological solution for calcium phosphate layer formation. 2-4 Sol-Gel H A p Process 2-4-1 Process Overview Much effort has been done to synthesize HAp coatings with improved chemical and physical properties. In particular, the ability to fully cover implant surface may be important for fixation of medical devices. With the advancement and better understanding of H A p synthesis, the sol-gel process has received considerable attention in the recent decades, particularly for coatings. The sol-gel method provides a soft chemical route that ensures a significant improvement in the chemical and physical homogeneity of the final product. In conventional sol-gel syntheses, metal alkoxides, which can be represented as M(OR) n where M is metal atom and R represents the alkyl group, are frequently used as starting precursors. In 1990, a research group in Japan had first disclosed an alkoxide-based sol-gel method to form H A p powder by using calcium diethoxide Ca(OC2H5)2 and triethyl phosphite P(OC2Hs)3 as precursors [62]. This sol-gel method offers a molecular-level mixing of the 12 calcium and phosphorus precursors, which improves chemical homogeneity of the resulting HAp, in comparison with conventional methods such as solid state reactions [78], wet precipitation [79,80], and hydrothermal synthesis [81]. For the sol-gel process developed by Masuda et al. [62], a sol pH range of 5-8 was found suitable for a phase-pure HAp formation; impure phases would develop only if the pH is above or below this range. Later on, numerous researchers focused on a similar subject but have extended the regime of coating [23, 24, 63-77]. Based on these results, it appears that the versatility of the sol-gel method opens a great opportunity to form thin film coatings and has currently widely used for biomedical applications [82,83]. The sol-gel approach provides much milder conditions for the synthesis of HAp films. This results in a much better structural integrity whereas the defects typical for plasma spraying can be largely avoided [82]. Furthermore, the lower temperature synthesis particularly benefits the metal substrates where the mechanical degradation or phase transition of the underlying alloy (e.g. a—» B phase transition in Ti , occurring at 883°C and 960°C, respectively) can be prevented. However, thermal treatment of H A p sol-gel films in vacuum is frequently required to avoid metal oxidation. This leads to structural instability of the H A p coating (i.e., evolution of structural water) during vacuum thermal treatment. Therefore, from both economic and practical points of view, thermal treatment of the HAp coating should be performed in air. In order to minimize oxidation of the underlying substrate, thermal treatment temperature should be selected at a minimum level, which still assures sufficient quality of the H A p film, in terms of crystallinity, 13 film integrity and adhesion to the substrate. At temperatures below ~500°C oxidation of the underlying Ti or Ti alloy is negligible due to the presence of a natural, dense oxide layer on commercial titanium implants [84,85]. Therefore, the primary focus of this research is a low-temperature sol-gel process for the synthesis of HAp. Recently, H A p has been used as vehicle for controlled release of drugs [31, 86-91]. A number of combinations between calcium and phosphorus precursors were employed for sol-gel HAp synthesis. For instance, Gross et al. [65], who followed the work done by Masuda et al. [62], used calcium°diethoxide (CaCOEfh) and triethyl phosphate (PO(OEtj3) to form pure H A p phase at temperatures above 600°C. They also found that ageing time longer than 24 h is critical for the solution system to stabilize such that a monophasic HAp can be produced [92]. Otherwise, large weight loss during pyrolysis and undesirable phases, such as CaO, could be observed. Jillavenkatesa et al. [93] synthesized a mixture of HAp and CaO at 775°C using calcium acetate (Ca(C2H302)2) and triethyl phosphate as precursors. A further hydrochloric acid leaching was required in that process to eliminate CaO, leading pure HAp phase. Brendel et al. [64] obtained HAp at a temperatures as low as 400°C using calcium nitrate (Ca(N03)2.4H20) and phenyldichlorophosphite (C6H5PCI2) as precursors. However, the resulting H A p had low purity and poor crystallinity. Further increase of the synthesis temperature to 900°C resulted in pure, well-crystallized H A p phase. Takahashi et al. [94] developed a gel route using calcium nitrate and phosphonoacetic acid (HOOCCH2PO(OH)2) in an aqueous solution and obtained a pure H A p powder at 700°C. The 14 crystallinity of H A increased with temperature up to 1100°C. Chai et al. [82] compared two calcium precursors, namely calcium diethoxide and calcium propionate, reacted with triethyl phosphite to form HAp coating. They found that HAp phase appeared at 500°C for calcium propionate solution, but no H A p formed when calcium ethoxide was use. However, they did not explain the influence of chemical nature of the precursors on phase formation. Qiu et al. [95] used calcium nitrate and ammonium dihydrogen phosphate (NH4H2PO4) to synthesize HAp in highly basic solution. They obtained HAp at calcination temperatures of 500°C -1100°C and indicated that the crystallinity of the H A p was improved with increasing temperature. Haddow et al. [69] used calcium acetate together with a number of phosphorus precursors, namely phosphoric acid (H3PO4), phosphorus pentoxide (P2O5), and triethyl phosphite for HAp coating applications. They found the films prepared from triethyl phosphite and calcium acetate showed the best wetting characteristic and the temperature required to form an apatitic phase was greater than 600°C. Lopatin et al. [24] used hydrated solution of N-butyl acid phosphate mixed with calcium nitrate tetrahydrate dissolved in 2-methoxyethanol to synthesize HAp. The authors indicated that HAp phase was able to develop at a temperature as low as 300°C. However, the crystallinity of the HAp phase was rather poor and improved only when the heat treatment temperature was above 500°C. However, increase in the firing temperature caused an increase of the amount of CaO and tri-calcium phosphate, which are non-desirable impurity phases if present in mixture with HAp. Weng et al. [70] synthesized HAp using a mixed ethanol solution of calcium nitrate and phosphorous pentoxide. A highly 15 crystalline HAp coating with dense morphology was obtained after heat treatment at 500°C for 12-24 hours. Phosphorus alkoxides have frequently been used as the phosphorus precursors for sol-gel H A p synthesis in recent years. Triethyl phosphate and triethyl phosphite are major precursors among them [62-65, 6S,69, 93,94, 96]. The hydrolysis activity of the triethyl phosphate is relatively poor and a higher solution temperature together with a prolonged time period (of several days) is needed to form HAp phase [93]. Alternatively, triethyl phosphite offers a much higher activity for hydrolysis [95,97] and a recent study by means of 3 I P N M R revealed a valance transition from P(III) to P(V) upon ageing with Ca precursor to form HAp within 24 h [65]. This indicates a nucleophilic addition of negatively-charged OH" groups to the positively-charged metal P, leading to an increased coordination number of the phosphorus atom which is essentially an indication towards the polymerization reaction [97]. After subsequent protonation of the alkoxide ligands (-OR) and removal of the charged ligand (-OR) +, P-(OR) is hydrolyzing to form P-(OH) [78,24], following interaction with Ca precursor to develop the apatitic structure. Besides the difference in chemical activity of the precursors, such as hydrolysis, polycondensation, etc., the temperature that is required to form the apatitic structure depends largely on the chemical nature of the precursors. Table 2-1 provides a comparison among the chemical precursors and solvents used, together with the synthesis parameters such as temperature and time of annealing employed to form a crystalline H A p via the sol-gel process 16 by different research groups worldwide over the past decade. As indicated in Table 2-1, the sol-gel process parameters vary considerably for synthesis of phase-pure HAp. However, lower-temperature, environmentally-friendly, and shorter processing time are currently most desirable for the sol-gel synthesis. Therefore a water-based, low temperature synthesis of the HAp is one of the focuses of this research. 2-4-2 Hydrolysis of Sol-Gel HAp Precursors In conventional sol-gel synthesis, metal alkoxides are frequently used as the starting materials. In this process the alkoxides are hydrolyzed in the presence of water, hydroxyl groups nucleophilically substitute alkyl groups bonded to the metal atom (M) and alcohol molecules (ROH) are released as by-product: M-(OR) + H 2 0 -> M-(OH) + R O H (2-1) A subsequent polymerization-condensation reaction between M-(OH) molecules leads to the formation of - M - O - M - bond and water as by-product: M-(OH) + M-(OH) -> M - O - M + H 2 0 (2-2) Table 2-1: Chemical precursors, mutual solvents, and the synthesis parameters, i.e., temperature and time, required to form phase-pure crystalline H A p in the various sol-gel HAp 17 processes developed over the recent decade. Chemicals Solvent Temperature Time Reference Ca(OEt)2 PO(OEt)3 Ethanol >450°C >24h Y. Masuda, 1990 [62] Ca(N0 3) 2 C 6 H 5 PC1 2 Ethanol >500°C T. Brendel, 1992 [64] Ca(N0 3) 2 N H 4 H 2 P 0 4 Water >500°C Q. Qiu, 1993 [76] Ca(N0 3) 2 HOOCCH 2PO(OH) 2 Water 700°C H.Takahashi 1995 [94] Ca(OEt)2 P(OPr)3 Ethanol 500°C 24h CS Chai 1995 [82] Ca(OEt)2 PO(OEt)3 Ethanol >600°C >24h K.A. Gross 1998 [65] Ca(C 2 H 3 0 2 ) 2 PO(OEt)3 Water >775°C >48h Jillarenkatesa 1998 [93] Ca(C 2 H 3 0 2 ) 2 H3P04/P205/P(OEt)3 Ethanol >600°C D.B. Haddow 1998 [68] Ca(N0 3) 2 C 4 H 9 (H 2 P0 4 ) 2-methoxy-ethanol 300-500°C 24h C M . Lopatin. 1998 [24] Ca(N0 3) 2 P 2 0 5 Ethanol 500°C 48h W. Weng 1998 [70] Ca(OEt)2 H 3 P 0 4 Methanol >600°C 24h P. Layrolle 1998 [98] Ca(N0 3) 2 PO(OEt)3 2-methoxy ethanol 600°C >16h M.F. Hsieh 2001 [99] Where Et = C 2 H 5 and Pr = C 3 H 7 18 Reactions (2-1) and (2-2) can be accelerated by acid or base catalysts. For most sol-gel glasses/oxides, Reaction (2-2) proceeds spontaneously at room temperature to form a three-dimensional network structure, resulting in a solid-like gel. However, for orthophosphates, Reaction (2-2) requires a moderate heat treatment to solidify the sol solution [44,47,63]. Recently, a few examples demonstrated gelation of the orthophosphates at ambient environment i f calcium ethoxide [62] or calcium glycolate [100] was employed as a starting material under highly acidic conditions. A number of combinations between phosphorus and calcium of various chemical forms have been adapted for HAp formation, as presented in Table 2-1. Triethyl phosphite has been widely used as one of the precursors because of its rapid hydrolysis, as opposed to trialkyl phosphate precursor. Subsequent interaction of the hydrolyzed phosphite with Ca precursors, for instance, calcium diethoxide [62], calcium acetate [93], or calcium acetate-glycolate [100], proceeds slowly to form Ca-P-containing derivatives. During the reactions, an increase in the coordination number of phosphorus from III to V was detected through 3 1 P N M R , an evidence of polymerization [62]. Non-aqueous solvents are frequently employed for dilution of triethyl phosphite, together with a small amount of water or acetic acid for hydrolysis. Triethyl phosphite is immiscible with water and forms emulsion phase after mixing with water. The emulsion turns into a clear solution after a certain period of time, then the phosphite odor disappears, indicating a complete hydrolysis [97]. Westheimer et al. [97] 19 proposed that the trialkyl phosphite proceeds rapidly to form dialkyl phosphate in acid. For instance, in the case of trimethyl phosphite, the unshared electron pair in trimethyl phosphite will react rapidly with proton to form protonated phosphite, followed by deprotonation to form a product: ( C H 3 0 ) 3 P : + H + <-> ( C H 3 O H ) 3 P + H ( C H 3 0 ) 3 P + H + H 2 0 -> (CH 3 0) 2 P(H)=0 + C H 3 O H + F f (2-3) However, the chemistry of extended hydrolysis of dimethyl (or diethyl) hydrogen phosphite has not been fully identified. According to the reactions proposed by Masuda et al. [62], diethyl hydrogen phosphite underwent further hydrolysis or chemical modification to form monoethyl phosphite, followed by interaction with Ca to form a complex containing Ca and P. Crystalline H A p then developed after heat treatment of the chemical complex at > 600°C. It is known that polymerization reaction wil l usually accompany hydrolysis. Therefore, a reaction to form oligomeric phosphorus compounds during synthesis is possible. This would result in the formation of calcium phosphate materials other than HAp and accordingly, have a lower Ca/P ratio than stoichiometric HAp. 20 2-4-3 Ageing of the Sol-Gel H A p In the sol-gel synthesis of HAp, alkoxides or metal salts are frequently used as either calcium or phosphorus precursors. In most cases, phosphorus alkoxides were employed as one of the major constituents, together with stoichiometric amount of calcium alkoxides or calcium salts, to form HAp. However, a long period of the sol preparation time, e.g., 24 h or longer, is commonly reported in order to form a desirable product, refer to Table 2-1. This is because of slow reaction between Ca and P precursors in the sol phase. Obviously, the reactivity depends on the chemical nature of the precursors. For instance, it took at least 24 h for the most often-used, highly-active phosphorus alkoxide, i.e., triethyl phosphite, to interact with calcium alkoxide in order to obtain HAp [65]. In the case of less-active triethyl phosphate, longer time and higher temperature treatment is necessary to activate the interaction with calcium acetate to form crystalline H A p [93]. Unfortunately, in either case, details of the chemical pathways have not yet been fully understood and the chemical reactions proposed so far in the literature may be oversimplified in terms of "real" reactions [62, 65, 100, 101, 102, 103]. 2-4-4 Structural Evolution of Sol-Gel-Derived H A p Low-temperature formation and fusion of the apatitic crystals has been the main advantage of the sol-gel process in comparison to conventional methods. For instance, temperatures higher than 1,000°C are usually required to sinter the fine apatite crystals prepared from wet precipitation, whilst 700-900°C is needed to density sol-gel-derived HAp 21 [71,98]. Moreover, sol-gel H A p usually has fine-grain microstructure, better accepted by the host tissue. The recent communication of Schrotter et al. [104] reported the structural changes of phosphite (P(OC2H5)3) precursor by monitoring of ' H , 1 3 C, and 3 1 P N M R during hydrolysis. They indicated that the phosphite rapidly reacts with water to form (HPO(OC2Hs)2), and then forms a mixture of HPO(OC 2 H 5 )(OH) and HPO(OH) 2 . Several days were needed to complete the hydrolysis reaction, i.e., to produce HPO(OH)2. It is not presently clear what is the exact product of the reaction of phosphite with water in this system. Therefore, we used a form of HPO(OC2H5)2-x(OH)x in this work to represent the mixture [105]. A subsequent reaction with Ca precursor is critical in developing a desirable apatitic structure. A recent study indicated that a specific aging time either at ambient or higher temperature is necessary to form a phase-pure apatite [106]. Insufficient ageing causes appearance of impurity phases, such as CaO or CaC03. This suggests incomplete reaction between HPO(OC 2H5)2- x(OH) x and Ca [106]. Therefore, structural evolution during the sol-to-gel transition of HAp requires further attention and is investigated systematically in this work. However, the ageing effects have scarcely been explored in sol-gel HAp processing. One recent report by Chai et al. [92] indicates that a critical ageing time for the sol containing triethyl phosphite and calcium diethoxide of at least 24 h is necessary to form a monophasic HAp. This prolonged ageing time is needed for P and Ca precursors to form desirable intermediate compound suitable for HAp formation. A corresponding increase of the coordination number from P(III) to P(V) was detected during ageing, indicative of 22 polymerization process. 2-4-5 Sol-Gel HAp Coatings Metallic biomaterials, such as titanium and its alloys, have enjoyed clinical successes because of their superior strength, biocompatibility, durability, and resistance to corrosion in physiological environment [107]. The high mechanical strength and toughness of these metals are the most important advantages over bioactive ceramics, which are inherently weak and brittle. Upon implantation, a close contact of the metal prostheses with surrounding host tissue is required for a subsequent in-growth of bone tissue into the pre-design cavities on the implant surface. The change of relative position between implant and the surrounding tissue is highly undesirable and therefore immobilization of a patient would be required before the implant fixation is strong enough to bear load. Eulenberger et al. [108] reported about 100 days for this recovery time period for bone fixation applications. This time can be reduced to only 20 days through the use of HAp, leading to rapid bond development between HAp and surrounding bone tissue. However, mechanical weakness of the HAp limits its practical application to those requiring little or no load bearing locations. Therefore, to extend the applicability of the HAp to implants bearing substantial load, such as dental or hip implant, a composite system including H A p coating on the metallic implant have been used [38]. This system combines the mechanical advantages of the underlying (metallic) substrate and 23 biological affinity of the HAp surface to natural tissue. It is critical however that this system does not substantially deteriorate during its lifetime, e.g. through loss of interfacial integrity. Interfacial bonding strength for plasma sprayed H A p on titanium alloys is sufficient, e.g., 20 - 30 MPa, possibly due to development of a chemical bonding between HAp and Ti through calcium titanate phase and/or Ti-P compounds [22, 54, 34]. However, plasma spraying operates under extremely high temperatures, e.g., 6,000 - 10,000°C, and thus can easily destabilize the crystal structure of the HAp, causing decomposition into a mixture of HAp, CaO, tricalcium phosphate, tetracalcium phosphate, and considerable amount of amorphous phases [21, 22, 110]. Severe cracking of the coating layer is frequently observed, primarily due to rapid temperature fluctuations and solidification of the coating. Additionally, transformation of the amorphous phase to crystalline oxide, while the plasma-sprayed coating is subject to annealing, gives rise to cracking, as a result of volume change (shrinkage) [111]. These cracks may also cause loss of adhesion of the coatings, leading to delamination, premature wear and finally implant detachment. However, published reports regarding the adhesive strength of the sol-gel HAp coating on metal or ceramic substrates are relatively limited. Most reported works focused on structural evolution of sol-gel HAp [66, 67, 69, 82, 112]. According to Weng et al. the adhesive strength of 10.5 MPa can be achieved for a porous HAp coatings on alumina substrate [70], and > 14 MPa on Ti6A14V alloy [73], after annealing at 750°C of Ca(N03)2.4H20 - P2O5 sol-gel system. Breme et al. [109] minimized the thermal expansion 24 mismatch between Ti and sol-gel HAp coating (prepared from CaO - PO(OC2Hs)3 system) by introducing small amount of M n to form Ti-8Mn alloy and claimed a bonding strength > 70 MPa after 600 - 800°C annealing. 316L steel has long been used for prosthesis devices such as plate, screw, etc, in orthopedic surgery. Therefore, in this work we prepare the sol-gel H A p coatings on stainless steel substrate with sandblasted surface. Surface morphology and interfacial microstructure of the coatings on the 316L substrates were also examined in terms of annealing temperature, and further correlated with their mechanical behavior. 2-5 In-Vitro Deposition of Calcium Phosphate Bioactive materials such as bio-glasses, calcium hydroxyapatite, tri-calcium phosphate, promote bone-tissue formation at the interface and form chemical bonding with osseous tissues. Such remarkable properties allow bioactive materials use in a number of medical applications in dentistry and orthopedics [1,17,86]. Formation of calcium phosphate layer (CPL) on the surface of the bioactive materials in solutions that mimic blood plasma has been recognized as a prerequisite for the materials to form bond with living tissue [19, 113]. Therefore, the formation of C P L in vitro on the surface of the materials is considered to be bioactive in-vivo [131]. Biomimetic deposition process was used recently to deposit calcium phosphate layer (CPL) at room or body temperature for a variety of biomedical applications, including drug 25 delivery [26, 114 129,130]. This forming mechanism is driven by supersaturation of desirable ionic species, primarily C a 2 + and PO4 3", under an appropriate solution pH. The apatitic crystals form through nucleation and growth on specially prepared ("activated") material surface. A direct incorporation of biological active agents, such as antibiotics, anti-cancer drugs, anti-inflammatory agents, etc., enhances clinical success of the CPL . The most attractive rationale of such biomimetic deposition process is that C P L can be developed at ambient environment, i.e. heat treatment is no longer required, as opposed to other techniques such as plasma spraying [21, 22], physical/chemical vapor deposition [50, 52, 59, 60], and sol-gel methods [23-25, 70]. Recently, a number of reports have emphasized the formation of C P L onto metallic substrates, such as Ti or its alloys. For instance, Wen et al. [27] deposited C P L onto porous Ti surface subjected to simple chemical treatments, following a pre-calcification procedure that activated the Ti surface. Deposition of C P L , at a rate as fast as 1 um/h, has been observed. However, the deposited phase was predominantly composed of loose octacalcium phosphate plates, which had poor mechanical properties and was unsuitable for drug incorporation and release. Later, they modified the process to deposit dense C P L with an apatitic structure onto the porous Ti surface at a deposition rate of about 3 um over 8-day in-vitro immersion [26]. They suggested that strong adhesion of this C P L variant could be expected due to chemical bonding and mechanical entanglement of the calcium phosphate crystals on the porous surface of Ti. Campbell et al. [28] activated the substrate surface via a surface-induced mineralization technique, where macromolecules containing functional 26 groups were employed to form a self-assembled monolayer on substrate surface. These molecules "anchor" to the underlying substrate and can further interact with other organic molecules with functionalized end groups that are able to induce mineral nucleation and growth. The deposition rate of C P L on such activated surface was found to be solely dependent on the solution chemistry. The deposition rates were 0.4-0.5 um/h for solution of sufficient degree of supersaturation (with respect to the formation of apatite crystals) and about 1 um per day for only slightly supersaturated solution. As disclosed in a recent patent literature, drug molecules can be introduced into such C P L film via a cyclic deposition-adsorption process [114]. One promising feature of this technique is that it allows substrates of different materials to be coated with CPL. The method can thus be used to coat prosthetic devices with complex geometry. It seems conceivable that the formation of C P L imparts not only bioactive property to those biocompatible materials, e.g., native Ti surface, but may also serve for drug delivery purposes. In other words, C P L may essentially be a dual-functional coating. 2-6 Applications of Calcium Phosphate for Drug Delivery As indicated in previous section, the solution-deposition method for Ca-P films may include incorporation of drugs or biologically active agents, such as antibiotics, anti-cancer agents, bone morphogenetic proteins, etc, into the layer. Preferably, both the Ca-P crystals and the drug molecules are deposited simultaneously or in a cyclic deposition-adsorption [114], 27 resulting in an in-situ formation of drug-containing Ca-P coating. However, high-concentration dosage of the incorporated active agents is difficult to achieve in this process. This is especially critical for prosthetic devices in orthopedics where a high concentration of antibiotics is required at bone-coating interface to prevent acute inflammation in early stage post-implantation. Additionally, the physiological solutions for Ca-P layer formation are naturally water-based. This makes it impossible to encapsulate hydrophobic bioactive agents into the coatings using this approach. Therefore, it is important to find an alternative method suitable for drugs or active agents of different natures to be encapsulated and release in a controlled manner. 28 Chapter 3 : Objectives and Scope 3.1 Objectives 1. To develop a new sol-gel process for HAp formation, having the following characteristics: (i) sol can be prepared under ambient environment, (ii) water can be used as diluting medium, (iii) HAp synthesis temperature is below 500°C. Lower synthesis temperature should permit an annealing/calcination environment shift from controlled atmosphere (e.g., vacuum or oxygen-free gases) to air environment, without causing thermal or oxidative degradation of the underlying metallic substrates. 2. To explore the related sol-gel chemistry through investigation of the possible reaction pathways of the newly-developed synthesis scheme. In particular, the effects of hydrolysis, ageing, and calcination temperatures on H A p formation need to be systematically investigated. 3. To establish the relationship between microstructural development and phase evolution of the new sol-gel HAp, in terms of annealing temperature and other process variables. 4. To explore the use of the novel HAp sol-gel scheme for thin film synthesis. Microstructure and bonding strength at the interface between the thin-film HAp and underlying substrate wil l be examined. In-vitro bioactivity of the H A p coating will be evaluated. 29 3.2 Scope In this investigation, triethyl phosphite, P(OC2H 5 ) 3 , and calcium nitrate hydrate, Ca(N03)2 4 H2O, are selected as P and Ca precursors for the sol-gel HAp synthesis, respectively. Processing parameters that are critical to the resulting H A p included annealing temperature, hydrolysis time, use of catalyst, and ageing time. The range of these parameters is as follows, 1. Temperature: 300 - 500°C. 2. Heat treatment environment: ambient 3. Hydrolysis time: as short as possible, e.g., few minutes. 4. Ageing time: as short as possible, e.g., < 24 h. As a result, H A p is achieved with 1. phase composition having phase-pure hydroxyapatite with Ca/P ratio in the vicinity of stoichiometry, 1.67. The crystallinity of resulting H A p should be controllable, for instance, from poorly-crystalline to well-crystalline HAp. 2. application to form thin-film coating on various metal substrates, typically, Ti alloy or stainless steel. 3. sufficient bioactivity that can be evaluated via in-vitro test, by immersing into a simulate body fluid with ion concentration mimicking that of human blood plasma, for a period of 14 days. 30 The resulting H A p was further characterized in terms of elemental analysis through Inductive Couple Plasma- Atomic Emission Spectroscopy (ICP-AES), microstructure using scanning electron microscopy (SEM), crystal phase via X-ray diffraction analysis (XRD) and Infra-red spectroscopy (FTIR), thermal behavior using thermogravimetric analysis (TGA) and differential thermal analysis (DTA), and mechanical property. 31 Chapter 4: Experiments 4-1 Sol-Gel Synthesis of HAp In the synthesis of sol-gel HAp, two diluting media are first employed for comparison purpose. One is ethanol that is widely used in many sol-gel processes, the other is water.Triethyl phosphite (Sigma, USA) sol was first diluted in anhydrous ethanol and then small amount of distilled water was added for hydrolysis. The molar ratio of water to the phosphorus precursor was kept at 3. The mixture was sealed in a glass beaker immediately after solvent addition, then stirred vigorously. Due to the immiscibility between the phosphite and water, the mixture initially appeared opaque, light being scattered by the emulsion phase. However, after approximately 30 min of mixing, the emulsion phase transformed into a clear solution suggesting the phosphite was completely hydrolyzed. This was confirmed by the loss of phosphite odor of the mix. Stoichiometric amount (i.e. to maintain Ca/P = 1.67) of 3 M calcium nitrate (Aldrich, USA) dissolved in anhydrous ethanol, was subsequently added dropwise into the hydrolyzed phosphorus sol. Vigorous stirring continued for additional 10 minutes after the titration. As a result of this process, a clear solution was obtained and aged at room temperature for 16 h before drying. The solvents were then driven off at 60°C until a viscous liquid was obtained. The corresponding HAp concentration changed from 3.6 vol% (in the solution) to 13.6 vol% (in viscous liquid), the calculation based on the final volume of HAp powder after calcination. Further drying the viscous liquid at 60°C resulted in a white gel. The gel was ground with a mortar and pestle into fine powder and subjected to different 32 calcination temperatures from 200 to 800°C with 25°C intervals for 2 h. In addition to the above-mentioned ethanol-based synthetic scheme, an aqueous-based process was employed using distilled water as diluting medium. The preparation procedure was identical to that of ethanol-based process described above. The powders derived from both types of gels were analyzed using an X-ray diffractometer ( M A C Science, C u K a , M18X) at a scanning speed of l°20/minute from 20 to 40°. Scanning electron microscopy (SEM, Hitachi 800, Tokyo Japan) was used for microstructural examination. The gel powders were also examined by thermal gravimetric analysis (TGA, Perkin Elmer 7 series Thermal Analysis) at a heating rate of 15°C/minute from 25°C to 1200°C under a flowing air of 10 ml/minute to monitor the weight loss of organic residues. Elemental analysis of the Ca and P was conducted using energy dispersive x-ray spectroscopy (EDS) and inductively coupled plasma-atomic emission spectroscopy (ICP-AES, USA, Jarrell-Ash, ICAP-9000). A reagent-grade commercially available HAp powder (Fisher Scientific, USA) was employed as reference materials for comparison purpose. Figure 4-1 shows the flowchart of the water-based sol-gel process. Hydrolysis of phosphorus sol and ageing of the mixed sol were systematically investigated in order to gain better understanding of possible reaction paths and for processing optimization. For hydrolysis study, the phosphorus sol was first hydrolyzed in water with and without the presence of nitric acid. In the absence of the acid catalyst, the phosphite sol was hydrolyzed in water for 0.5, 1, 2, 33 4, 8, 16, and 24 hours, whilst catalyst was added, the phosphite sols were hydrolyzed using acid solution with nitric acid concentration of IN, 2N, and 5N. The calcium-containing solution was added dropwise into the hydrolyzed phosphite sol. The mixture was vigorously stirred for 10 min, followed by ageing for 24 h at ambient environment. The pH value of the solution without acid was monitored during the stages of hydrolysis and ageing. The resulting white, dried gel was calcined at 375°C for up to 6 hours or at 500°C for 10 minutes. Phase identification of the calcined powder was examined using X R D . The gel powders were also examined by T G A to monitor the weight loss of organic residues. The microstructure of the gel powders was examined using scanning electron microscopy. Fourier transform infrared (FTIR, Impact 400D Nicholet Spectrometer) in an absorption mode was used to examine the structure in those dried and calcined gels. For IR sample preparation, 2 mg of powder sample was thoroughly mixed with 200 mg of K B r using a mortar and pestle, following pressing at 100 MPa to form pellets. For ageing study, aliquot amount of the mixed sol (without catalyst) was taken at an ageing time period of 0.25h, 0.5h, lh , 2h, 4h, 8h, 16h, and 24h at ambient temperature, 25°C. Additional ageing experiments were performed at 35, 45, 55, 65, and 80°C, following the same procedure as that for the room-temperature ageing. These aged sols were then subjected to thermal treatment at 80°C for 16 h until a white dried gel was obtained. The dried gels were further calcined at 400°C or 500°C for 10 minutes. X-ray diffraction and FTIR analyses were used for structural characterization. 34 P Sol (Hydrolysis) Ca Sol y 1 ' . Mixing Ageing Drying / Dip-Coating Calcination (a) 350-500C Figure 4-1 Flowchart of the water-based sol-gel H A p synthesis. 35 4-2 Sol-Gel HAp Coatings Metallic substrates, i.e., 316L stainless steels (50 mm x 50 mm x 2 mm) were sandblasted with 150 grit SiC for improvement of coating adhesion. These substrates with surface roughness of ~3 um were dip coated with the sol solution, with a withdraw speed of 5 cm/min. The coatings were then dried at 80°C, followed by annealing in air at 375°C, 400°C, and 500°C for 60 min, 20 min, and 15 min, respectively. For metallographic examinations the samples were mounted in epoxy resin and polished along cross-section direction down to 0.5 pm alumina grit. Pull-out test was performed for interfacial bonding strength determination of the HAp-coated 316L specimens. The bonding strength of the coatings to substrate was tested according to the procedure of A S T M standard C-633-79. The coatings are glued by DP-460 thermoset epoxy (3M product) with two plain steel rod substrates (1 inch diameter). The rod surface is wiped with acetone, and dried in open-air to evaporate solvent before applying adhesive. The test specimens were cured at 175°C for 60 min, and the pull-out test performed at a crosshead speed of 1 mm/min, using a universal mechanical tester (Instron, Model 3600). At least 6 test specimens were used for each experimental point. The bonding strength of the epoxy adhesive was about 50 MPa, as determined through a blank test. The coatings were further characterized using reflection mode of Fourier transform infrared (FTIR, Impact 400D Nicholet Spectrometer), and X-ray diffractometer (XRD) at a scanning speed of 1° 2G/min from 25° to 40° was used for phase identification (however, thin film X R D is unsuitable for phase identification of the coating on such a roughened surface) Cross-section examination of 36 the coatings was performed by slicing, mounting, and polishing a set of 5 cm x 5 cm substrates, followed by examining with scanning electron microscopy. The coatings, having surface area of about 1.5 cm 2 , were then immersed into 50 ml of a water-based fluid with ionic composition nearly identical to that of human blood plasma, in a sealed polyethylene container. The ionic composition (in units of mmol/1) of the fluid was 142 Na + , 5.0 K + , 2.5 C a 2 + , 1.5 M g 2 + , 103 CI", 25 HC0 3 " , 1.4 HP0 4 2 " , and 0.5 S 0 4 2 " .This "simulated body fluid" (hereafter termed SBF) was buffered at pH 7.4 with tris(hydroxymethyl)aminomethane and H C l [131]. In-vitro test was performed in a static condition ( i.e. the SBF was not renewed during the test period) at 37°C for 14 days. The solution pH was monitored periodically and the samples of C P L were taken out at 3, 7 and 14 day periods for microstructural examination, through the use of S E M . For cross-sectional microstructural observation, samples were mounted in an epoxy resin (Industrial Formulators of Canada, Ltd.), thermally cured at 75°C for 2 h, and polished down to 1 um AI2O3 powder suspension. FTIR with reflectance mode was used for analysis of the deposit phase before and after the in-vitro test. 4-3 H A p for Drug Encapsulation For encapsulating drugs, 316 stainless steel substrates are first coated with a 0.6-0.8 um thin layer of sol-gel apatite (SG-HAp). Colloidal slurry containing calcium phosphate precursors, such as Ca(OH)2 and calcium phosphate salts, i.e., monocalcium phosphate 37 anhydrate, was ball milled in ethanol. The starting inorganic ingredients of Ca(OH)2 and monocalcium phosphate anhydrate were submicrometer in size, i.e. 0.3 um and 0.5 um, respectively. The initial Ca/P ratio in the suspension was kept at 1.5. The thin film SG-HAp surface-modified sample was dip coated in the suspension. After dip coating, a 10-30 um thick layer of porous powder mixture developed on the substrate due to rapid evaporation of ethanol. Subsequently, a small amount of sodium phosphate water-based solution (0.25 M) was dropwise added into the deposited powder mixture, allowed to soak into the open pores of the film, and then the sample was placed in an incubator at 37°C, 100% relative humidity, for 24 h. This resulted in the formation of an apatite layer, termed as CPC-HAp. In order to assess the possibility of use of the CPC-HAp for controlled drug release, amethopterin (Sigma Chemicals, USA) , having a chemical structure C20H22N8O5 and M W = 454.4, was employed in an amount of 5% based on solid phase content of CPC-HAp precursors. Amethopterin is frequently used to treat certain cancers and arthritis. In clinical applications, amethorpterin is given once per week to those with rheumatic disease, but there are conditions in cancer therapy where the drug is given daily. So far, amethopterin can be taken by injection under the skin or by mouth. Another important rationale for using the amethopterin is because it is highly soluble in water or preferably in buffered saline. To incorporate the amethopterin into the novel C P C - H A P film-type vehicle, the drug dissolved in aqueous sodium phosphate solution, was added dropwise onto the precursor film of CPC-HAp, following incubation for a period of 38 several hours. After encapsulation, a drug release study was conducted by immersion of the substrates into 20 ml of phosphate buffer saline (PBS, pH 7.4) at constant ratio of (CPC-HAp coating weight)/(volume of PBS) of 1 mg/ml. The entire PBS was refreshed in 4, 8, 24, 48, and 72 h of time and the used PBS was subjecting to U V - V i s spectroscopy (USA, Backman, Model DU-600) with a characteristic adsorption peak at 352 nm to examine the amount of the amerthopterin in the solution according to a pre-determined concentration - intensity calibration curve. A poly(vinyl alcohol) (PVA, Aldrich, molecular weight = 13,000-23,000) solution with a concentration of 5 weight percent was prepared for a post-coating purpose in order for an alternative comparison of release control. 39 Chapter 5: Results and Discussion 5-1 Process Development 5-1-1 Characterization of the Mixed Sol After initial preparation of the sols, pH = 2.85 was detected for the ethanol-based sol and pH = 2.77 for the aqueous-based one. In anhydrous ethanol, alkyl groups may form Ca(OR)y(N03)2-y and replace some of the nitrate groups while the calcium nitrate is dissolved in water to form ionic species. Both the ethanol-based and aqueous-based sols are relatively stable and would not transform into a solid gel at ambient temperature, for at least 5 days. A decrease in solution pH was observed for the aqueous-based system to pH = 1.680 after 24 h whereas the pH level in ethanol-based sol was relatively constant over the entire ageing period. Both sols transformed into clear, viscous liquid after approximately 76 vol% of the solvents were removed upon heating at 60°C. This corresponded to a solid concentration increase from 3.6 vol% to 13.6 vo l% (calculation based on the resulting solids content after removal of all organic residuals). Upon cooling to ambient temperature, a remarkable increase in viscosity was observed for both the viscous sols. A solid-like translucent gel (i.e. no sign of flow was observed at ambient temperature) was seen only when greater than 80 vol% of the solvent was removed, corresponding to a solid concentration of about 15.8 vol%, upon cooling to ambient temperature. It is interesting to note that both the translucent gel and the viscous sol (for both aqueous- and ethanol-based systems) showed a reversible and reproducible flow behavior, i.e., from low viscosity on heating to high viscosity on cooling. However, the viscous sol could be 40 re-dissolved into the solvent to form a clear sol solution as initially prepared, whereas the translucent suspension was observed when the translucent gel was "re-dispersed" in the solvent. This suggests that the colloidal particles (which are large enough to scatter incident visible light) once developed can hardly be dispersed due to strong interparticle bonding. Although the bonding nature is unknown, this observation suggests a particulate nature of the resulting sol. According to the study of Livage et al. [115, 116] and Gross et al. [65], the polymerization of the phosphite cannot proceed indefinitely. Instead, it wil l be limited to a certain degree of polymerization due to progressively weakly charged hydroxyl ligands. The polymerization stops when the partial charge of the hydroxyl groups approaches zero or positive charge, according to the partial charge model proposed by Livage et al. [115]. The colloidal particles can thus be considered as an aggregation of oligomeric particles. Extended heating of the viscous sol at 60°C resulted in a white solid gel. Upon ageing, the hydrolyzed phosphorus sol (which may be in a form of phosphoric ester [62, 97], HPO(OEt)2, or more generally, P(OEt)3_x(OH)x) interacted with Ca sol, possibly in the form of Ca(OEt) y(N03)2- y in anhydrous ethanol and Ca'"" in water, to form oligomeric derivatives containing Ca-O-P bonds. For the ethanol-based process, the reaction may proceed as follows: P(OEt) 3 . x(OH) x + Ca(N0 3 )2- y (OEt) y -> -> (OEt)y(N03)2-y'-Ca-0-HPO(OEt)3-x- + H 2 0 + C 2 H 5 O H 41 (5-1-1) For aqueous-based process, an ionic derivative may be developed by: P(OEt)3-x(OH)x + C a + 2 + NOV 1 -> -> (N0 3)" 1(OH)-Ca-0-PHO(OEt) 3-x' + H + + C 2 H 5 O H + H 2 0 (5-1-2) Reactions (5-1-1) and (5-1-2) are rather simplified and idealized chemical reactions. Although not strictly precise, they offer some understanding on phase formation based on forthcoming analysis. The release of protons in Reaction (5-1-2) seems to account for a decreased pH level of the sol, and was experimentally verified. Further heating causes removal of the solvents, accompanied by accelerated thermal dehydration [116] or polymerization/condensation [62] between these derivative units, resulting in the formation of more (-Ca-O-P-)-containing bonds in the dry gels. 5-1-2 Characterization of the Dry Gels Results of thermogravimetric analysis of both gels are shown in Fig. 5-1-1, where three weight loss stages can be easily distinguished (the NE-gel represents ethanol-based gel and NA-gel, aqueous-based gel). The characteristic temperature regions are almost identical, i.e., 30-100°C, 100 - 300°C, and 300 - 520°C, for both gels. In the first region, a sharp weight loss by 16% for NA-gel and 18% for the NE-gel was observed, indicating the evaporation of 42 ethanol and adsorbed water. The second region reveals a slow removal of residuals, such as structural water, and the third stage results due to nitrate decomposition, leading to approximately 23 wt% loss for both gels. No further weight loss was seen at temperatures higher than ~520°C (till 1000°C) for the NE-gel, indicating complete removal of the residues. However, upon close inspection, a slight and gradual decrease in the T G A curve was detected for the NA-gel from 520°C to about 730°C, corresponding to a weight loss below 1%, suggesting a further removal of structural water due to decomposition. However, no further change in the weight loss was detected upon further heating to 1000°C, indicating the decomposed phases are thermally stable. The NA-gel shows a total weight loss of 43 wt%, and the NE-gel, 45 wt%. This corresponds to a yield of 57% and 55%, for NA-gel and NE-gel, respectively. Large amount of weight loss due to removal of residuals could be potentially dangerous for structural integrity in coating applications of the HAp. However, when the gels were dried at 100°C, yield can be enhanced to about 73% for both gels, which is suitable for coating purposes. 43 0 500 1000 Temperature (°C) Figure 5-1-1 Thermogravimetric analysis of both N A and N E gels, showing a similar weight loss behavior The small difference in weight change commenced at the first heating stage, indicating the NE-gel contained greater amount of residual ethanol. This is reasonable because this gel was derived from ethanol-based system. The striking similarity between the weight change curves for both gels suggests a close resemblance of residual composition evolved during the course of the gel formation, as indicated in the Reactions (5-1-1) and (5-1-2). The X R D patterns of both gels after different temperatures of calcination are given in Figs. 5-l-2a and 5-l-2b, for NA-gel and NE-gel, respectively. Both gels have similar diffraction patterns, irrespectively of the diluting medium employed. Several major peaks, i.e., (002), (211), (112), (300), and (202), are identical to those of the reference HAp structure 44 (top pattern in Fig. 5-1-2a, and JCPDS file card #9-432), indicating the formation of apatitic structure. Small amount of tricalcium phosphate (TCP) phase was detected for both gels calcined at different temperatures, especially pronounced for the NA-gel-derived HAp at 800°C, suggesting decomposition of the HAp. However for those samples calcined at lower temperatures, the presence of trace TCP indicates that it may be produced during the course of synthesis. One possible reason is the high acidity of the sol solution that favors formation of acidic calcium phosphates, having a lower Ca/P ratio than stoichiometry. Elemental analysis shows that the dried N E gels have a Ca/P molar ratio of 1.663, a value slightly smaller than stoichiometric HAp of 1.67 and a value of 1.656 for NA-gel (derived from more acidic solution), smaller than the stoichiometric value. These provide supporting clue for above argument. The NA-gel seems more thermally unstable at elevated temperatures, e.g. 800°C, than those derived from the NE-gel. Together with the above elemental analysis, suggest formation of a calcium-deficient apatite from N A gels. Differential thermal analysis (DTA) shows a broad endothermic region over temperatures of 640°C - 730°C for the NA-gel, suggesting a slow decomposition of the defective, Ca-deficient HAp, Fig. 5-1-3. 45 20 25 30 35 40 45 50 55 (a) Figure 5-1-2 The X R D patterns for (a) NA-gel and (b) NE-gel after different calcination temperatures. A predominant apatitic phase can be obtained, with a minor amount of TCP phase which developed at high temperatures. An X R D pattern (top curve) of commercial HAp is also provided for comparison. 46 47 2 i -2 i 1 1 1 1 1 0 200 400 600 800 1000 . Temperature (°C) Figure 5-1-3 Differential thermal analysis of N A and N E gels However, for the NE-gel, no appreciable change in D T A curve was observed in high-temperature region, (>600°C) indicating formation of a stable HAp phase. The differential thermal analysis is consistent with the weight loss curve (Fig. 5-1-1), and also confirmed by the phase identification in Fig. 5-1-2, wherein a strong endothermic peak at ~100°C indicates evaporation of residual volatiles such as water or ethanol, followed a small exothermic peak at 350°C for H A p phase crystallization, and an endothermic peak at about 500°C as a result of H A p phase formation. It should be noted that the apatitic phase could be obtained at a temperatures as low as 350°C, which is one of the lowest temperatures ever 48 reported in the literature for HAp by means of sol-gel process. This apatite-forming temperature is also lower than that observed in the electrochemical deposition process recently proposed by Shirkhanzadeh [57], who detected an apatitic structure appearing at 425°C after 6 h. It should be noted however that the diffraction intensity of the major peaks in both gels was improved with increasing temperature, indicating an increase in HAp content that is attributed to both the removal of the organic residuals and improved crystallinity of the apatite. Eanes et al. [117] examined width of the (002) peak at 29 = 25.8° of the apatitic structure for both types of gels calcined at different temperatures. The half-intensity width of the (002) peak decreased gradually (accompanied by improved sharpness of those major peaks) as the temperature increased from 350°C to 800°C for both gels. This observation indicates an increase in crystallite size and/or improved crystal structure of the annealed H A p powders. However, previous X R D analysis of the NA-gel (Fig. 5-l-2a) may suggest that the change in half width of the (002) peak is attributed mainly to crystal growth (this observation is confirmed by microstructural analysis presented later). For the H A p phase derived from NE-gel, the apatitic structure is thermally stable. A decreased intensity of half width for (002) crystal plane is then suggestive of improved crystallnity along the crystallographic c-axis at elevated temperature (Fig. 5-1-2b). The X R D patterns for both dried gels before calcination are identical (these are represented by the pattern shown in the bottom in Fig. 5-l-2b). These peaks indicate the 49 presence of crystalline phase (e.g., nitrate at 29 ~ 20.5° and -25°, suggesting an incomplete reaction between hydrolyzed phosphorus and calcium precursors). However, these peaks disappeared after calcination at temperatures > 350°C. This crystalline phase was found to be water-soluble, i.e. disappeared after washing the gel with distilled water. Most interestingly, no apatitic structure was detected by calcination of this water-rinsed powder at 375°C or even at 400°C. Therefore, it is apparent that the Ca-P-containing derivatives in the dried gels are water soluble, as suggested by the presence of hydrophilic groups in the derivatives in Reactions (5-1-1) and (5-1-2). We therefore believe that the formation of the resulting HAp phase is essentially related to the nitrate-containing crystalline phase (which may act as a pre-HAp derivative, herein referred to as pre-HAp). Significant loss of Ca and P can be incurred upon washing the gel with water, with the consequent loss of HAp. Further investigation of the phase content of gel calcined at temperatures below 350°C revealed a gradual decrease in reflection intensity of pre-HAp with increasing temperature until 350°C. At this temperature the crystalline structure of the dried gel (i.e. pre-HAp) completely disappeared and an apatitic structure developed. This indicates that the apatite phase is essentially a re-crystallized form of poorly-crystalline or possibly a transient amorphous calcium phosphate phase. It should be further noted that the crystallization of H A p occurs at 350°C, a temperature that is lower by about 200°C to 300°C compared to those reported in the literature [65, 98, 92, 118]. A greater amount of hydroxyl species (primarily from initial water addition) 50 evolved in the gel phase may be responsible for the low-temperature crystallization [118]. Based on the above observations, the novel sol-gel route proposed in this investigation seems to provide an attractive alternative for HAp synthesis, especially compared to the all-alkoxide-based process [62, 92]. Although the current method resulted in a calcium-deficient H A p phase via the entirely aqueous-based system, this achievement is very encouraging. This new process opens a possibility to form a biphasic calcium phosphate, which has been recently recognized as offering an optimum balance in solubility in clinical applications [119]. 5-1-3 Microstructural Examination The crystal size of the dried gel is inversely proportional to the X R D peak half-intensity width according to the Scherrer equation [120]: A(2^) = - ^ (5-1-3) •Dcos^f?) Where A(20) represents the peak width at half maximum intensity of the reflection (002), X is the wavelength for C u K a , (k=0.15418 nm), and D is the crystal size in nm. Figure 5-1-4 shows the resulting D calculated for both gels as a function of calcination temperature. It is evident that the crystal growth contributes to the increase in reflection intensity, as depicted in Fig. 5-1-2. Both gels exhibit similar crystal size for a given temperature, increasing from -120 nm 51 at 350°C to -210 nm at 800°C. S E M micrograph of the NA-gel calcined at 800°C shows that the average crystal size is -200 nm, Fig. 5-1-5, close to the size determined from X R D analysis via Eqn (5-1-3). By closer examining this micrograph, it can be observed that some initial sintering occurred between the H A p crystals (as indicated by arrows). This sintering behavior seems not only limited to the particle-particle necking, typical for the early stage of sintering. Some well-sintered sections can also be observed, which is believed to be due to smaller H A p crystallites providing better sinterability at lower temperatures. Figure 5-1-4 Crystal size of the N A and N E gels, calculated based on the Scherrer equation, calcined at different temperatures. 52 Figure 5-1-5 Scanning electron micrograph of the microstructure for the NA-gel calcined at 800°C an average size of approximately 200 nm, in agreement with calculation, was observed. Some sintering behavior is also seen (as arrows indicated). 53 5-2 Effect of Hydrolysis 5-2-1 Effect of Hydrolysis Time (Without Catalyst) Figure 5-2-1 shows the X R D patterns of the gels hydrolyzed for different time durations, followed by calcination at 375°C for 4 h. The calcined gels showed a mixture of CaCC>3, calcium nitrate, and a small amount of apatite for short-term hydrolysis, i.e., 0.5 h and 1 h. The gels showed a predominant apatitic structure together with a small amount of tricalcium phosphate (TCP) when the phosphite was hydrolyzed for > 4 h. Several characteristic peaks, such as (002), (211), (112), (300), and (202) of the H A p (JCPDS card #9-432) are clearly identified. The gelled powder after calcinations had dark gray color for those powders subjected to short-term hydrolysis (SH) and was light gray for those powders subjected to long-term hydrolysis (LH). An extended heating of the light-gray L H gel for additional 2 h produced white but gray color persisted for the SH gels. This suggests that the L H gelled powders contained less amount of organic residues. Thermogravimetric analysis results shown in Fig. 5-2-2 for both the S H and L H gelled powders provide quantitative evidence of the changes in the calcined powders. A relatively large weight loss of about 16% was observed in the range of 25 - 100°C for the L H powder, whilst the weight loss was only about 5% for the SH powder. 54 o H A P <) TCP A CaCO • Ca(N03) O | U Q3)2 | U o | 2 4 h * H\ rP Q # 16h i i. A 40 35 30 29 25 Figure 5-2-1 X R D patterns for the gels calcined at 375°C for 4h for hydrolysis time, 0.5 to 24 h, of the phosphite sol, without acid catalyst. 55 This weight loss is primarily due to the removal of adsorbed water and free solvent, i.e., C2H5OH. Although both powders showed similar total weight loss, i.e., -43% above 550°C, the weight loss patterns of the powders are different, indicating different pyrolytic mechanisms. 100 90 ^ 80 <D 60 C ca 70 O £ '5 60 50 4 40 0.5-h hydrolysis, SH 24-h hydrolysis L H Acid Catalyst (2N) 200 400 600 800 1000 1200 Temperature (°C) Figure 5-2-2 Thermal gravimetric analysis of the dried gels prepared under different conditions (with and without acid catalyst) of hydrolysis of the phosphite. At the same temperature, for instance 375°C, the SH gelled powders demonstrated smaller weight loss (-12%), as compared to the L H powder (-25%). This means that the residues in the SH powder are more difficult to remove than in the L H powder. Therefore, it is reasonable to believe that sufficient amount of organic residues can still appear as 56 chemically-bonded groups, such as C2H5O-, in the SH gelled powder. Therefore a higher temperature is required to remove those residues. This may also suggest that a more extensive amorphous background was seen in the SH gel than that in the L H gel (Fig. 5-2-1). The increased peak intensity of the X R D pattern with temperature is then suggestive of an increase in HAp concentration and crystallite growth rather than improved crystallinity, because the half-intensity width of the characteristic (002) peak remains unchanged for L H gels. However, an appreciable increase in the intensity and sharpness of the diffraction peaks was revealed by raising the calcination temperature to 500°C, Fig.5-2-3, indicating an increase in HAp concentration and improved crystallinity. Elemental analysis of the 500°C - calcined gel using the energy dispersive X-ray spectroscopy (EDS) showed a Ca/P molar ratio of 1.65-1.66, suggesting that the synthetic apatite is a Ca-deficient apatite. The gels calcined at 500°C showed white color in appearance, indicating no residues left. Small amount of TCP was detected for all the samples and the diffraction intensity of the TCP peak was found to decrease with increasing time of hydrolysis. This indicates that an extended hydrolysis favors H A p formation. Change in solution pH during hydrolysis shows that the sol became more acidic with time, Fig. 5-2-4. Therefore, incorporation of more acidic phosphate ions, such as HPO42", rather than more basic group, PO4 ", into the resulting gel structure becomes more favorable and this may account for the decreased crystallinity of the resulting apatite. TCP is unlikely to result from thermal decomposition of the apatitic phase because the 57 calcination temperature is too low (eg. 375°C) to destabilize the apatitic structure. One supporting evidence is that no appreciable change in the relative intensity of the major diffraction peak for TCP and apatite phase has been observed over an extended heating for 24 h at the same temperature, even after a higher temperature calcination at 500°C (Fig.5-2-3). The transformation from the Ca-deficient apatite to a mixture of H A p and TCP as indicated by Slosarczyk et al. [80] may be a possible alternative paths. However, the temperature used here is lower by 250-400°C than that observed by Slosarczyk et al., so it is thermodynamically unfavorable to trigger these reactions. Even i f decomposition does occur at such a low temperature region, a detectable change in the relative intensity of the reflection peak between the TCP and HAp phases should be easily observed, which is not the case. Therefore, one plausible explanation is a result of transformation from amorphous Ca-P-containing intermediate phases developed during synthesis. 58 o o HAp 0 TCP A CaO o o A. A 1 ^ ^ ^ ^ hiv f^kjK i I 1 lh 40 35 29 30 25 Figure 5-2-3 X R D patterns for the calcined gels (500°C/10min) derived from the phosphite sol hydrolyzed for lh , 4h, and 24 h. A small amount of TCP was detected in all cases. 59 1 1 1 1 1 1 0 4 8 12 16 20 24 Hydrolysis Time (h) Figure 5-2-4 pH change in phosphite sol solution during hydrolysis. A mixture of amorphous apatite and TCP phase could be co-existing in the dried gel and transformed into crystalline phases upon thermal treatment. Therefore, to obtain HAp gel with sufficient crystallinity and minimum amount of impurities of such as TCP and CaO, a sufficient time of hydrolysis for at least 4 hours seems necessary for the alkoxide phosphite. It is also interesting to note that the formation of apatitic phase becomes much less time-dependent at 500°C, although stronger time dependence is observed at 375°C. This is suggestive of a kinetically-controlled phase formation at lower temperatures, whilst thermodynamics dominating at higher temperatures. 60 5-2-2 Reactions During Sol Preparation Triethyl phosphite can be hydrolyzed in the presence of water or air moisture to form diethyl phosphorus ester [62]: P ( O C 2 H 5 ) 3 + H 2 0 -> HPO(OC 2 H 5 ) 2 + C 2 H 5 O H (5-2-1) Masuda et al. [44 62] indicated that Reaction (5-2-1) proceeds exceedingly slowly, e.g. takes 10 h or more to complete. However, a convenient and direct qualitative observation of the hydrolysis is the loss of phosphite odor, together with the disappearance of emulsion phase (which appeared immediately after initial mixing), to form a clear solution [97]. From experimental observation, both odorance and emulsion phase can be removed in about 20 minutes, which is in qualitative agreement with that reported by Westheimer et al. [97]. A higher molar ratio of water to phosphite in our case may provide explanation for this discrepancy, where higher hydrolysis rate can be expected than that observed by Masuda et al. (where r = 1 was studied). However, a continuous decrease in solution pH was detected for up to 24 h (Fig. 5-2-4), suggesting an ongoing hydrolysis, by release of proton according to Reaction (5-2-3) below. This strongly suggests that an extended hydrolysis of the diethyl phosphorus ester may proceed where more -OR groups are replaced by -OH groups, i.e. (5-2-2): 61 H P O ( O C 2 H 5 ) 2 + H 2 0 -> HPO(OC 2H 5) 2-x(OH) x + x C 2 H 5 O H (5-2-2) The hydrolyzed phosphite interacts with Ca ions in aqueous solution through a condensation polymerization reaction to form a Ca-P intermediate: 2HPO(OC 2 H 5 ) 2 - x (OH) x + C a ^ -> Ca-P intermediate + 2FT" (5-2-3) The liberation of proton in Reaction (5-2-3) can be monitored by the change of the solution pH, as shown in Fig. 5-2-5, for the phosphite sol hydrolyzed for 4 h, where the solution pH decreases as a function of ageing time for over a period of 24 h. This also indicates a slow reaction between hydrolyzed phosphite and Ca ions. Phosphite sol hydrolyzed for 4 hours before mixing with Ca sol 12 16 Ageing Time (h) 20 24 Figure 5-2-5 Solution pH in a mixed sol containing both Ca and P precursors for 24 h of ageing. 62 For the SH gelled powders, some organic ligands (e.g., alkyl groups) may still chemically bond to the phosphorus, e.g., (OC2H 5) 2-xO(H)P-0-Ca, rather than being liberated as a by-product, such as C2H5OH, which is much easier to burnout. The poor crystallinity of apatitic phase observed in Fig. 5-2-1 at lower temperatures also suggests that these residues may retard crystallization. The change in color and the phase evolution illustrated in Fig.5-2-1 for the SH powders seem to be supporting evidences. Instead, sufficient hydrolysis causes the formation of greater amount of solvent by-product (i.e. C 2 H 5 O H in Reaction 5-2-2) and is expected to remove easily, resulting in a white-color gel powder with improved crystallinity after moderate calcination. The presence of the TCP phase may be a result of condensation between hydrolyzed phosphite molecules to form oligomeric derivatives such as: 2HPO(OC2H5)2- x(OH) x->(OC2H5)2.xO(H)P-0-P(H)0(OC 2H5)2-x + 2 H 2 0 (5-2-4) The oligomeric derivative further hydrolyzes (Reaction 5-2-5 below) and reacts with Ca to form calcium phosphate derivatives, according to Reaction 5-2-6: (OC 2 H 5 ) 2 . x (H)OP - O - P(H)0(OC 2 H 5 ) 2 . x + H 2 0 - » (OC 2 H 5 ) 2 - y (OH) y (H)OP - O - P(H)0(OC 2 H 5 ) 2 - y (OH) y + 2 (x-y)C 2 H 5 OH (5-2-5) 63 (OC 2 H 5 ) 2 . y (OH) y (H)OP - O - P(H)0(OC 2 H 5 ) 2 . y (OH) y + C a + + -> C a - 0 - ( O C 2 H 5 ) 2 . y ( H ) O P - 0 - P ( H ) 0 ( O C 2 H 5 ) 2 . y - 0 - C a + 2 H + + H 2 0 (5-2-6) Although above equations describe simplified and somewhat speculative chemical paths, they do propose the formation of the Ca-P derivatives with lower Ca/P ratios than stoichiometric HAp. From X R D analysis, the lower Ca/P calcium phosphates are present only in a small amount in all cases, indicating that Reaction (5-2-6), i f it does occur, is relatively minor. Both the apatite and the lower Ca/P calcium phosphates can develop as amorphous intermediates in the gel, which further transform into crystalline phases upon heating. The presence of tricalcium phosphate, TCP, (Ca/P = 1.5) illustrated in Figs. 5-2-1 and 5-2-3 is a supporting example. However, the exact chemical paths to form the T C P (or possibly also other low Ca/P calcium phosphates that are undetectable due to low resolution of X R D ) are not clearly understood at present. Ironically, this "impure" TCP phase may be not an undesirable component according to the biphasic calcium phosphate concept recently proposed by Daculsi [119]. From practical viewpoint, it is desirable that an implant material can be designed in such a way that it can be completely replaced by the host tissues upon a resorption-apposition mechanism. Moreover, bioresorption has been a process that is able to accelerate the growth of defective hard tissues and this mechanism implies the significant role of resorption in biological response. Therefore, 64 the less-resorbable nature of synthetic HAp can be further modified by incorporation with more-bioresorable secondary phase. Daculsi [119] took advantage of the resorption/dissolution abilities of different calcium phosphates (i.e. A C P or TCP) by combining more soluble tricalcium phosphate (TCP) with non- or less-soluble HAp to form biphasic calcium phosphate (BCP). A n optimization of dissolution balance was reported in vitro for the BCP. Therefore, an optimized combination of different calcium phosphates of varying degree of solubility may be a promising alternative for some practical purposes. Based on this concept, a minor fraction of the secondary TCP phase developed in situ under current HAp synthesis conditions can be considered as an advantage rather than an obstacle. 5-2-3 Effects of Hydrolysis Time (Systems With Catalyst) The use of acid catalyst to accelerate hydrolysis and/or polymerization is commonly exercised in conventional alkoxide-based sol-gel processes [115, 121]. The experimental observations showed that an immiscible (emulsion-like) phosphite phase formed when diluted nitric acid water was added. The mix became a clear solution within a few seconds, accompanied with the loss of phosphite odor. However, it took about 20 minutes to obtain a clear, odor-free solution with distilled water alone (pH = 6.8). This means that the hydrolysis rate can be accelerated by 2-3 orders of magnitude when the acid catalyst is introduced. Therefore, the time for hydrolysis described in previous section can reasonably be reduced by several orders of magnitude through use of acid catalysts. After conducting a series of 65 experiments, we found that for these systems a time period of about 5 minutes is long enough for the development of crystalline HAp. Figure 5-2-6 shows the X R D patterns of the gelled samples prepared with acid catalyst of different concentrations, IN, 2N, and 5 N , calcined at 375°C for 4 h. The major apatitic peaks, e.g., (002), (211), (112), (300), and (202), are clearly distinguishable. The calcined powder derived using 2N and 5N acid catalyst showed white appearance; indicating a minimum concentration of 2N is sufficient for H A p synthesis However, the one derived using 1 N acid catalyst showed gray color, indicating residues left. A comparative study of the X R D patterns between the powder (2N) and the standard JCPDS file (not shown) showed that a small shift of the diffraction peaks (especially the characteristic peaks such as 002, 211, 112, 300, and 202) towards smaller 20 by 0.06° - 0.07°, relative to the standard pattern. This indicates an expansion in the crystal lattice along both a and c axes, which is suggestive of lattice imperfection by larger ions. Accordingly, this defect may be introduced as a result of the higher acidity of the solution, where larger acidic phosphate HPO4" 2 may substitute smaller P C V 3 . The resulting material may, based on electrical neutrality, have a chemical form of Caio-x(HP04)x(P04)6-x(OH)2-x, i.e., a calcium-deficient apatite, which agrees with the Ca/P ratio previously determined. Trace amount of TCP was also observed, but the one with 2N catalyst shows little TCP phase and was used for a subsequent coating practice (refer to Chapter 6) It appears that hydrolysis (5-2-1) and polymerization reactions (5-2-2 to 5-2-5) 66 concurrently take place at a much faster rate with the help of acidic catalyst. Similarity in the phase evolution in these acid-catalyzed powders implies that the reaction path is similar to those powders without using catalyst. The hydrolysis reactions (5-2-1 and 5-2-4) are believed to proceed completely for the cases of higher acid content, i.e. > 2N. This results in liberation of greater amount of C 2 H 5 O H and H2O as by-products that can be removed easily during the subsequent heat treatment. However, a T G A curve of the acid-catalyzed (2N) powder shows a weight loss behavior (Fig.5-2-2) somewhat different, in the temperature region of 100 - 430°C, from those without catalyst, where a steady removal of the residues was observed, rather than a slow removal at 100 - 350°C followed by a fast removal at 350 - 430°C for the acid-free powders. Microstructural examination revealed that both powders (with and without catalyst) calcined at 375°C for 4 h show similar equi-axial geometry with particle size of about 120 -150 nm, Fig. 5-2-7a and 5-2-7b, respectively. The close resemblance of the particle morphology between both powders strongly suggests (i) the reaction paths during synthesis remaining unchanged and (ii) a particulate nature of the water-based sol-gel-derived apatite. 67 Figure 5-2-6 X R D patterns of the gels (hydrolyzed for 5 min) calcined at 375°C for 4 h. The gels were prepared under the presence of nitric acid catalyst of IN, 2N, and 5N concentrations. Commercial HAp pattern is given for comparison purpose. 68 Figure 5-2-7 The morphology of the gel powders prepared (a) with and (b) without acid catalyst, after calcination at 375°C. 69 5-3 Effect of Ageing of Sol 5-3-1 Aged Sol From visual observations, the mixed sol solution (i.e., phosphite and calcium sols) remains transparent regardless of the ageing time. The mixed sol was stable and no any gelation occurred over an ageing time period as long as even several weeks. This is consistent with observations by Livage et al. [116], who attributed this stability to a limited degree of polymerization based on partial charge model [122]. A continuous pH measurement of the sol solution during ageing showed a decrease of pH value, normally from pH ~ 0.8-1 at the beginning to pH < 0.2 after ageing. This decrease in solution pH during ageing is similar to that previously observed for hydrolysis of the phosphite alone (Fig. 5-2-4). It is indicative of liberation of H + upon polymerization reaction between Ca and hydrolyzed phosphite precursors. Based on the partial charge model, the release of the positively-charged group, i.e., H + , further provides supporting evidence of the polymerization of the sol. The extent of pH reduction increases considerably as the ageing temperature rises, suggesting a thermally-activated reaction. 5-3-2 Phase Evolution: Room-Temperature Ageing Figure 5-3-1 presents X R D patterns for gels calcined at 400°C for 2 h after different time periods of ageing. The calcined gels derived from short-term ageing (< 4 hours) contain a small amount of poorly-crystalline apatite together with considerable quantity of impurity 70 phases consisting of CaO, Ca2?207, Ca3(P04)2, and nitrates. With increase of the ageing time, a gradual improvement in crystallinity (in terms of peak sharpness) and concentration (peak intensity) of the apatitic phase was observed whilst impurity phases gradually disappeared. These changes continued up to 8h, after then, the calcined gel is predominantly apatitic phase. However, for the case of 24-h ageing, calcined gels show the presence of small amount of tricalcium phosphate (TCP). After 500°C calcination for 2 h (Fig.5-3-2), considerable improvement in diffraction intensity is observed. The CaO (peak at 29=37.5°) seems to be the only impurity phase in the calcined gels for < 8 h ageing. The impurity phases such as Ca2P207 and nitrate, originally present in short-term aged gels (Fig.5-3-1), disappeared at 500°C. For long-term ageing, TCP appears to be the only trace impurity. To this point, it can be concluded that the gels are phase-pure apatite after the initial sol solution ageing for a time period of about 8 hours. These observations suggest that the critical ageing time period of about 8 hours is essentially required to obtain HAp through the route proposed in this work [25, 101,102]. This is shorter by a factor of approximately 3-4 than those reported previously [92]. The hydrolyzed phosphite may react with Ca precursor in aqueous environment through the following overall reaction: 2HPO(OC2H5)2-x(OH)x+ Ca + + -> [(OC2H5)2-xO(H)P-0]2Ca + 2H + (5-3-1) 71 Although the liberation of proton in Reaction (5-3-1) was verified by the change of the solution pH, Reaction (5-3-1) may still be an oversimplified form since some unknown reactions may also be concurrently taking place, which makes a detailed investigation of the chemical mechanism more difficult. As a first approximation, we assume that Reaction (5-3-1) represents the typical chemical path during synthesis. For those gels subjecting to short-term ageing, the presence of CaO phase results primarily from incomplete reaction (5-3-1), leaving a residual nitrate decomposing to oxide upon calcination. However, further examination of the X R D patterns for the 4 hrs and 16 hrs aged and dried (i.e. uncalcined) gels shows different crystalline phases. For shorter-term ageing, i.e., 4 hours (pattern A in Fig. 5-3-3), the dried gel shows broad diffraction lines consisting of a mixture of Ca2P207, Ca3(P04)2, CaC03, and Ca(N03)2, together with a strong amorphous background. These crystalline phases imply the complexity of the "real" reaction pathways, which are certainly much more complicate than that simply suggested by Eq. (5-3-1). Therefore, it would be pertinent to re-phrase Eq. (5-3-1) to a more general expression, i.e., Eqn (5-3-2): HPO(OC 2 H 5) 2-x(OH) x + C a + + + N03"-> Ca-P intermediate + H + (5-3-2) The "Ca-P intermediate" is X-ray amorphous, rather than crystalline phase. According to Figs. 5-3-1 and 5-3-2, "Ca-P intermediate" transforms to crystalline apatite at elevated temperatures. Longer ageing results in even stronger amorphous background, with little or no 72 crystalline reflections, as illustrated for 16 hrs aged sample, pattern B in Fig. 5-3-3. Two rather broad diffraction lines at 20 of -27° and -30.5° are displayed in this pattern, suggesting that the reaction (5-3-2) is more or less complete after 16 hrs ageing. o HAp 0 TCP A CaO • Ca(N03)2 V Ca2P207 4h 0.5h 4 0 3 5 3 0 29 2 5 Figure 5-3-1 X R D patterns of the gels calcined at 400°C for 2 h, prepared from different ageing time periods. Impure TCP phase appeared after ageing for greater than 16 h. 73 o o HAp 0 TCP A CaO A 40 35 20 30 25 Figure 5-3-2 X R D patterns of the gels, prepared from different ageing time periods, after calcinations at 500°C for 2 h, showing a better crystallinity. 74 A comparative inspection of Figs. 5-3-1 or 5-3-2 and 5-3-3 suggests that the apatite can evolve from either the transformation of the amorphous Ca-P intermediate or the reactions among Ca2P207, Ca3(P0 4)2, CaC03, and Ca(N03)2, or a combination of both when the calcination temperature reached about 500°C. This may be grossly represented as follows: 1 / 2 C a 2 P 2 0 7 + Ca 3 (P0 4 ) 2 + Ca(N0 3 ) 2 + C a C 0 3 + V* H 2 0 -> Ca 5 (P0 4 ) 3 (OH) + CaO + 2 N 0 2 + C 0 2 (5-3-3) o HAp 0 TCP A CaO • C a ( N 0 3 ) 2 V Ca 2 P 2 0 7 (B) 16 h ageing (A) 4 h ageing ? 901 v • 40 35 30 20 25 20 Figure 5-3-3 X R D patterns of the dried gels after (A) 4 h and (B) 16 h of ageing at ambient temperature. 75 Thermo-gravirnetric analysis of the gels, Fig.5-3-4a, shows greater weight loss by approximately 5% for the 4 hrs aged sample than that for the 16 hrs aged sample, at temperatures up to 500°C (the values in parenthesis are the total weight change). This provides supporting evidence for Eq. (5-3-3), where greater weight loss for the 4 hrs aged gel is due to the liberation of nitrogen dioxide and carbon dioxide. The calcite within the gels may result from reaction between air-supplied carbon dioxide and calcium ions during sol-gel synthesis. Although the real mechanisms for the development of these impurity phases are not fully understood, it is interesting to note that poorly-crystalline, low-Ca/P calcium phosphates can develop at such a low temperature, i.e. 80°C. As postulated above, the X R D pattern for the 16 hrs aged gel (pattern B in Fig. 5-3-3) suggests that progress of the reaction towards the right side of Reaction (5-3-2) leads to the formation of amorphous Ca-P intermediate. The absence of crystalline phases suggests that the Reaction (5-3-2) to form the Ca-P intermediate is approximately complete, and most of the Ca and P precursors are believed to consolidate into the amorphous structure. Further support to this hypothesis is obtained from analysis of the D T A curves, Fig. 5-3-4b. A smaller endothermic peak at lower temperature (~512°C) is observed for the 16 hrs aged gel, as compared to that of the 4 hrs aged gel (~546°C). This indicates that lower energy is required to form crystalline apatite from the amorphous Ca-P intermediate rather than from the reactants involved in Reaction (5-3-3). 76 0 500 1000 1500 Temperature (°C) Figure 5-3-4 Dried gels prepared after 4 h and 16 h ageing periods show different weight loss and thermal behaviors under (a) thermogravimetric and (b) differential thermal analysis. 77 5-3-3 Phase Evolution During Thermal Ageing Thermal treatment is expected to accelerate the Reaction (5-3-2) to form Ca-P intermediate complex. The mixed sol solutions were aged at 45°C for different time periods. Figure 5-3-5 shows the resulting phase evolution for the gel calcined at 400°C for 2 hrs. In comparison with that obtained at room temperature (Fig. 5-3-1), phase-pure apatite evolves after about 2 hours of thermal ageing, which is about four times shorter than that derived after room-temperature ageing. This indicates a fourfold increase of the reaction rate by heating the sol to 45°C. However, for longer-term ageing at 45°C, e.g. > 8 hrs, a decreased diffraction intensity and broadened peaks are found, Fig. 5-3-5. One extreme case is for 24 hrs thermal ageing, Fig. 5-3-5, where only a relatively small amount of the apatite remains. Instead, the crystalline phases in the calcined gel include Ca2P207, Ca3(P04)2, and CaC03, a mixture somewhat similar to that observed for the short-term aged "dried" gels (pattern A in Fig. 5-3-3). Figure 5-3-6 illustrates X R D patterns for the aged sol for 24 hrs at 45°C, and then dried (pattern A) and calcined at 400°C and 500°C for 2 hrs (patterns B and C respectively). The X R D pattern of the dried gel (pattern A in Fig. 5-3-6) is somewhat similar, but with much poorer diffraction intensity, to that of 400°C-calcined gels (pattern B). This suggests that the impurity phases within the 400°C-calcined gel may readily evolve upon drying. These calcium phosphates impurities have lower Ca/P ratios than the stoichiometric apatite, and can be categorized as acidic Ca-P compounds. According to an earlier study [104], these compounds 78 are prone to develop under more acidic conditions. Calcination at 500°C brings the dried gel into a mixture of crystalline apatite, TCP, and Ca2P20*7 (pattern C in Fig. 5-3-6), rather than o HAp 0 TCP A CaC0 3 V Ca 2 P 2 0 7 2h lh 40 35 20 30 25 Figure 5-3-5 X R D patterns of gels, derived from sols after different ageing periods at 45°C, after 400°C calcination for 2 hrs. A decrease in apatite associated with a corresponding increase in TCP was observed as aging time increased. 79 apatite and CaO as observed in the short-term aged samples (Figs. 5-3-2 and 5-3-3), suggesting that the reaction pathways did not follow Eq. (5-3-3). O o HAp 0 TCP A CaC0 3 V Ca 2 P 2 0 7 (B) 400°C gel A 0 (A) dried gel 40 35 30 2 9 25 Figure 5-3-6 X R D patterns of the gels, derived from 24 h-aged sol at 45°C, prepared after (A) 25°C (dried gel), (B) 400°C, and (C) 500°C calcinations. A considerable amount of the apatite appeared at 500°C. An appreciable amount of apatite phase appeared whilst the carbonate peak almost disappeared, associated with a small reduction of Ca2P207. This suggests that the apatite is derived mostly from the transformation of the amorphous Ca-P intermediate, together with a 80 smaller fraction being a product of reaction between Ca2P 2 0 7 and C a C 0 3 , according to Reaction (5-3-4); 3/2 C a 2 P 2 0 7 + 2 CaCOs + 1/2 H 2 0 -» Ca 5(P0 4 ) 3 (OH) + 2 C 0 2 (5-3-4) The presence of Ca2P20? phase indicates that it is a residual reactant of Reaction (5-3-4), once the CaC03 is completely consumed to form apatite. In fact, both Ca2P207 and TCP phases are not necessarily detrimental components of HAp implants, since they provide faster dissolution rate than pure apatite in physiological environments. According to the experimental observation of Dacusi et al. [119], this would be an advantage from the viewpoint of bioresorption. 5-3-4 H A p Phase Evolution Map Based on the phase development mentioned above, it is conceivable that the time of ageing can be considerably reduced by increasing the ageing temperature. The formation of phase-pure apatite can therefore be optimized by appropriate control of the ageing parameters. Developing a phase evolution map for apatite formation in terms of these parameters can further consolidate this hypothesis. The map can be constructed by examining the phase composition of different gels aged at various temperatures and time, followed by calcinations at 400°C for 2 hrs. Figure 5-3-7 illustrates the resulting phase evolution map, 81 where the shadowed region indicates optimal ageing parameters for single-phase apatite formation. The apatite evolved in the shadowed area has a Ca/P ratio of 1.67 + 0.02 based on EDS and ICP-AES measurements. HAp in the lower-bound region is accompanied by CaO, resulting primarily from incomplete reaction (5-3-2), whilst the upper-bound region indicates the formation of impurity phases, such as Ca2P207, TCP, and CaC03, together with poorly-crystalline apatite. The shadowed area of the map suggesting evolution of pure HAp narrows to about 0.5 hr for ageing above 65°C, as compared to about 2 hrs for ageing between 45°Cand 55°C. From a practical viewpoint, important conclusion results that a significant reduction in the time required for apatite synthesis can be achieved by appropriate thermal treatment. A new, finely tuned process, can therefore lead to sol-gel derived pure HAp phase in a period of time by over an order of magnitude shorter than the existing synthetic methods. 82 Figure 5 - 3 - 7 Phase evolution map for apatite formation in terms of ageing time and ageing temperature. The sol prepared without using nitric acid catalyst, and the resulting gels were calcined at 400°C for 2 hrs. 5 - 3 - 5 Ageing Kinetics Upon ageing, the formation of the intermediate Ca-P phase is produced as a result of polymerization between hydrolyzed phosphite and calcium precursors. This results in proton liberation, which causes the sol solution to become more acidic with time. Although a decreased pH of the phosphite alone was also observed upon hydrolysis, both reactions (i.e. hydrolysis and polymerization) may concurrently affect the solution pH. A clear differentiation between these two reactions is difficult. Therefore, in the process presented in 83 this work a two-step reaction is adapted, i.e. the phosphite precursor was first hydrolyzed, followed by mixing with calcium precursor to initiate the second-step reaction (5-3-2). A prolonged hydrolysis of the highly reactive triethyl phosphite in the first step (i.e. 24 hrs hydrolysis) would allow an accurate evaluation of the 2 n d step reaction. Figure 5-3-8 shows the kinetics of sol acidification as determined through monitoring the change in pH (plotted in terms of equivalent concentration of Ff , i.e., [H] = 10"pH) in the sol solution during ageing at different temperatures (for clarity, long-term ageing data at 25°C are not included). The plots include the experimental points and least squares best-fit lines, with correlation coefficient (R 2) indicated. The best-fitted curve indicated that the [H] values increase logarithmically with time and also with the ageing temperature, indicating the endothermic nature of Reaction (5-3-2). The curves can further be expressed as: [H] = kj [P] n [Ca]m In t (5-3-5) - E and *, = k0 exp — (5-3-5a) [P] and [Ca] represent initial P and Ca concentrations, respectively; n and m are reaction orders; ki can be defined as the apparent rate constant, and E is referred to as the apparent activation energy for Reaction (5-3-2). As all these parameters can be treated as constants under current investigation, Eq. (5-3-5) can be simply re-written as; 84 [H] = kln\ (5-3-6) 0 50 100 150 200 250 300 Ageing Time (minute) Figure 5-3-8 Ageing kinetic shows an exponential correlation with time and turns to be more pronounced with aging temperature. The solid curves are obtained through the use of best-fitted least-squared fit of Eq. (5-3-6). The correlation coefficients are indicated in parentheses. The rate constant has a value (mole/min.L) of 0.04, 0.159, 0.248, and 0.294 for 25°C, 45°C, 55°C, and 65°C, respectively. Equation (5-3-5) represents an apparent kinetic expression; however, a detailed analytical solution of Eq. (5-3-5) requires a set of different [P] and [Ca] values and the concentration of the Ca-P intermediate in order to determine the rate constant and reaction orders, i.e., n and m. However, this wil l be far beyond the scope of this study and complicate the analysis to a considerable extent, thus no detailed kinetic analysis was made in this work to solve Eq. (5-3-5). Instead, a relatively simplified kinetic form of Eq. (5-3-6) was employed, 85 which is essentially an experimentally-derived kinetic model. On this basis, the apparent rate constant k, based on Fig. 5-3-8, can be determined according to Eq. (5-3-6), and is plotted in an Arrhenius form as shown in Fig. 5-3-9. The apparent activation energy E for Reaction (5-3-2) can be further calculated from the slope of the straight line, as given in Fig.5-3-9, having a value of 10.35 kcal/mole, which is lower by -60% than that for hydrolysis of triethyl phosphite [97]. Figure 5-3-9 Arrhenius plot of the rate constant vs. ageing temperature shows good correlation with an apparent activation energy calculated to be 10.35 kcal/mole. The resulting rate constant k can be further examined by correlating the relative value of the obtained rate constants; 86 k. 45 0.159 0.04 4.0 k. 25 k, 55 0.248 6.2 k 25 0.04 k, 65 0.294 7.4 Jfc. 25 0.04 The subscript number indicates the ageing temperature. For instance, the reaction rate increases by approximately 4 times at 45°C as compared to 25°C, which is in excellent agreement with previous observations. These ratios are sufficiently high, suggesting that the reaction rate is strongly temperature-dependent, which is essentially indicative as a result of high activation energy. 5-3-6 Optimal Solution Chemistry To examine the optimal solution chemistry in terms of apatite formation, the ageing time data were selected according to the shadowed region of the evolution map at specific temperatures (Fig. 5-3-7) and marked on the [H]-t curves of Fig. 5-3-8. The resulting graph is given in Fig. 5-3-10, and highlighted by the two dash lines determined as an average value of the three sets of data points (indicated by squares). The range of solution pH suitable for phase-pure apatite formation is from approximately [H] = 0.4 M to 0.6 M , corresponding to an approximate pH range of 0.2 - 0.4. This optimal pH range seems only slightly affected by temperature, i.e., widening for higher ageing temperatures. It is, however, expected that the reaction pathway remains unchanged within the 40-60°C ageing temperature. 87 0.8 55°C pH~0.22 (Phase-pure apatite region) pH~0.38 25°C 0 0 50 100 150 200 250 300 Ageing Time (minute) Figure 5-3-10 Solution pH, which is most suitable for apatite formation under current aqueous sol-gel system, can be determined as highlighted between two dash lines, based upon the phase evolution map depicted in Fig.5-3-7. The pH range suitable for apatite formation in the sol system under investigation is exceptionally low in comparison to other studies involving both aqueous precipitation [123-127], e.g., pH ~ 7-10, and sol-gel routes [62, 93], e.g., pH ~ 6-12. This may explain the formation of defects presented in the structure, i.e., Ca-deficient apatite, as observed in this study. Although fundamental reasons for this phenomenon are unclear, little or no additional phases other than apatite were observed after low-temperature calcinations of the "optimally-prepared" gels. However, recent investigation revealed that the ability for apatite phase formation is strongly dependent on the chemical nature of the starting precursors, e.g., different forms of calcium salts, and nature of the solvent(s) used in sol preparation [128]. 88 Chapter 6: Sol-Gel HAp Coatings 6-1 Characterization of the Coatings 6-1-1 X R D Analysis Figure 6-1-1 shows the X R D patterns of the coatings annealed at 375°C, 400°C, and 500°C. The weak diffraction peaks together with an amorphous background were observed for all the coatings. A small thickness (-0.6 um) of the HAp layer structure 500°C O HAp o ^ -^-^  o 400°C 375°C 40 35 30 25 Two Theta, (degree) Figure 6-1-1 X-ray diffraction patterns of the sol-gel HAp coating annealed in air at various temperatures. associated with an extensive morphological roughness of the underlying surface (having a measured roughness of 3 urn) is likely to account for the observed weak diffraction patterns. It 89 can still be distinguishable that a broad diffraction peak appears in the range of 31.8 - 32.5° 29, which represents the characteristic peak of apatitic phase (according to JCPDS card #9-432). Some characteristic peaks at, for instance, (211), (300), (212) planes were shown for coatings annealed at > 400°C. This suggests that the apatite coatings evolved from poor crystalline to good crystalline structure when the annealing temperature is increased from 400°C to 500°C. A thin film X-ray diffractometry was also tried in this investigation in order to get a clearer diffraction pattern than Fig. 6-1-1, however severe surface roughness (of the underlying sandblasted surface) gave a much poorer resolution in diffraction pattern and was finally abandoned. 6-1-2 Infrared Analysis Figure 6-1-2 shows the IR spectra of the coatings annealed at different temperatures. For those annealed at 375°C, the characteristic band for PG*4 groups is not clearly discernable at 600-550 cm"1, instead, showing a broad band, in comparison to those at 400 and 500°C,which, according to Russel et al. [66], can be attributed the absorption modes associated with PO4 groups. With increasing annealing temperature to 400°C, the spectra clearly illustrate characteristic v 4 PO4 bands at 563 cm"1 and 600 cm"1, vi P 0 4 band at 942 cm"1, and a strong v 3 P 0 4 absorption band in the range 1100-1000 cm"1, which are typical of apatitic structure. 90 These adsorption spectra become stronger in intensity and better in resolution as the annealing temperature increases to 500°C, indicating an improved molecular arrangement, such as PO4 polyhedra, in the crystal structure. These findings suggest that at 375°C, the atomic arrangement in the apatitic phase is virtually a short-range order, in agreement with the X R D spectra in Fig. 6-1-1. T 1 1 1 1 1400 1200 1000 800 600 400 Wavenumber (cm'1) Figure 6-1-2 Fourier transform infrared spectra for the sol-gel H A p coatings prepared at different temperatures. 91 However, above 400°C, albeit hardly visible in the X R D patterns, FTIR shows a considerable improvement of the molecular arrangement manifested through clearly distinguishable absorption peaks. A shoulder at band 631cm"1 between 400°C and 500°C is attributed to O H group, indicating the presence of bonded water in the film structure. However, this O H absorption band can hardly be observed for samples annealed at 375°C. It should be noted that absorption band at 871 cm"1, which is attributed to CO3 group, appeared for all the annealing temperatures. This band is assigned to a B-type carbonate substitution for PO4 group in the apatite structure, indicating a carbonated apatite coating. This CO3 band is getting stronger in intensity for higher temperatures, suggesting increasing carbonate concentration in the film. The resolution of the absorption bands within the apatite film improves considerably with temperature, even though the dwell time is decreased by a factor of 3, i.e., from 60 min to 20 min, when the temperature increases from 375°C to 400°C. This is further evidenced that the arrangement of the PO4 tetrahedra (which construct the crystal column along the c-axis of the apatite lattice) is becoming more ordered and symmetric. 6-1-3 Surface Morphology of the HAp Coatings Figures 6-1-3a, b, and c show surface morphology of the H A p coatings subject to annealing temperatures of 375°C, 400°C and 500°C respectively. Although the coatings appear relatively dense, some surface microcracks, as arrows indicated (0.1 - 0.2 um wide and 1 - 5 um long), are visible for the 400°C and 500°C samples. These cracks are less extensive 92 as compared to those produced by plasma spraying [21, 22], and are confined to the pockets of relatively thick (few um) coating within ~ 10 um large dents produced in the sandblasted sample surface. These cavities act as reservoirs during dip coating, leading to thicker coating, and therefore these thicker-coated areas are more susceptible to form cracks due to excessive drying and sintering strain. In contrast, those surfaces with thin-coated areas are free of cracking. The surface features of the underlying substrate (i.e. grooves and cavities of irregular geometry) appear to be replicated by the HAp coating. 93 (a) Figure 6-1-3 Surface microstructure of the coatings onto the sandblasted 316 stainless steel substrates annealed at (a) 375°C, (b) 400°C, and (c) 500°C. 94 6-1-4 Interfacial Microstructure of the HAp Coatings Figures 6-1-4 and 6-1-5 illustrate cross-section micrographs for the 375°C-annealed coating at smooth and roughened (eroded) areas, respectively. As expected, the HAp coating covers the entire surface of the substrate, irrespective of the complex surface features produced by sandblasting (e.g. compare Fig. 6-1-5). A minimum thickness of about 0.6 um is estimated for the coating on smooth surfaces. However, 1 - 5 um-thick layer was found within the surface cavities. No interfacial de-bonding was observed between the coating and the stainless steel substrate, although the mounting epoxy-coating interfacial fracture was typical. No microcracks were observed within the 375°C annealed coating, consistent with surface morphology examination aforementioned (i.e. compare Fig. 6-1-3a). Figure 6-1-4 Cross-section view of the coatings annealed at 375°C shows a uniform thickness of about 0.6 pm. 95 Figure 6-1-5 Cross-section view of the coatings annealed at 375°C shows the ability of the sol-gel method to coat cavities/grooves of different geometry on the substrates. Figure 6-1-6 shows the 500°C-annealed, 2 to 4 um thick coating deposited within a cavity in substrate surface. The cross-section examination reveals no detectable cracks, especially for those through or across the coating layer. It is therefore reasonable to assume that the microcracks observed on coating surface are limited in population. These pores were absent for coatings annealed at 375°C and 400°C, therefore, they may be originating from the grain growth at higher temperatures. Some larger voids (-0.5 um) with irregular geometry may be a result of grain pullout during sample preparation. The coatings, although somewhat porous after 500°C annealing, are still firmly attached to the substrate, as indicated by fracture occurring only at coating-epoxy interface. 96 Figure 6-1-6 Cross-section view of the 500°C-annealed sol-gel H A p coating. Figure 6-1-7 Some sparsely distributed nano-scale pores were found within the HAp coating annealed at 500°C. 97 Some sparsely distributed nano-pore approximately 40 nm to 70 nm were also observed in the coating, as shown in Fig.6-1-7. 6-1-5 Interfacial Bonding Strength Figure 6-1-8 illustrates the bonding strength of the H A p coatings on the 316L stainless steel substrates after annealing at different temperatures. Taking into account the average and standard deviation of the test results (as indicated by the error bars) the interfacial bonding strength appears to be relatively constant at about 44 MPa + 5 MPa. The average values of bond strength increase with annealing temperature from 375°C to 450°C, and then decrease slightly for 500°C. This coating strength is slightly lower than the average strength of the cured epoxy, which was determined through a blank test on sand-blasted samples without coating to be 50 ± 5.8 MPa. Mixed failure modes of epoxy failure, epoxy-coating interface failure, and failure within the coating were found in all the test specimens. However, little or no coating substrate interfacial fracture was detected through visual and optical microscopy examination. This indicates that the "true" bonding at the coating-substrate interface should be stronger than or at least comparable to that of the epoxy itself. It can then be considered that the coating-substrate interfacial bonding is similar in the annealing temperature range of study. The presence of 98 numerous small pores in the apatite layer at 500°C may be one of the factors that deteriorate the mechanical integrity of the coating, resulting in a slight decrease in bonding strength. £ 45 •{ M 40 •} c 1 35 -o m 30 300 350 400 450 500 550 Annealing Temperature (°C) Figure 6-1-8 The bonding strength of the sol-gel H A p coatings on 316 SS prepared at different annealing temperatures. Bars represent the standard deviation of measurement. 6-1-6 Sol-Gel HAp Coatings on Dental Implants-Practical Application Figure 6-l-9a shows the porous surface morphology of the dental implant (supplied by Endopore Inc., Toronto, Canada), composed of numerous Ti beads of 100 - 200 um in diameter, attached to the solid Ti core. Macropores ranging from several tens to about 150 um can be observed between the beads. These macropores allow an ingrowth of bone tissue, providing interlocking for implant fixation and has been reported to receive good mechanical strength after implantation [129]. A thin layer of the sol-gel H A p coating (approximately -0.2 um thick as estimated from a fractured surface, not shown) was deposited by single dipping the porous-surfaced dental implant into the acid-catalyzed solution (2N), followed by oven 99 drying. The coating was calcined at 375°C in air for 2 h. Figure 6-l-9b shows the surface morphology of the H A p coated implant. The coating on the Ti beads is relatively smooth. Numerous cracks were observed in the coating over the regions of the necking area between Ti beads (arrows in Fig. 6-1-9c). This is due to the fact that the aqueous sol solution flowed down along the surface of the sphere during drying of the coating, and accumulated at the necking areas (and triple-junction areas as well) as a result of surface tension. In consequence, the layer deposited on the sphere surface is relatively thin (< 1 pm) while those at the necking areas are substantially thicker (> 10 pm), hence leading to cracking and porosity. A closer look at the necking area shows that the pores have size ranging from approximately 1 - 5 pm, which may be advantageous for the circulation of physiological fluids in practical applications. A thin and dense H A p layer covers all other surface features of the T i beads throughout the porous part of the implant. The dense, fine structure developed at such a low temperature suggests the ease of densification of the gel particles and is consistent with a number of sol-gel-derived oxides [121]. The test indicates that the aqueous sol-gel HAp coating is suitable for biomedical devices having complex geometry. In this respect, the HAp coated surface reproduces the surface feature of the underlying porous titanium, retaining the bone interlocking capability of the surface. This is impossible to achieve using other coating techniques in particular plasma spraying. It is interesting to note that the coating on the sphere surface is relatively thin and is electronically (also optically) transparent which allows the structural features of the underlying metal to be 100 observed. A finger-nail scratching test was performed on the coated sample and showed nothing to come off from the sample surface, suggesting an adhesive coating. 6-1-7 Discussion The use of sol-gel technique for deposition of bioactive H A p coating is simpler and more accessible in comparison to other processing alternatives. The process results in a high quality thin H A p layer that imparts biological affinity to the underlying substrates. A temperature between 375°C and 400°C was required to develop apatitic phase for the coatings on roughened stainless steel, which is slightly higher than the temperature, i.e., 350°C required for H A p formation in powder form using the same sol-gel route, as described in Chapter 5. The volume contraction of the coating film gives rise to tensile stresses within the gel structure, which damages the structural integrity of thicker (few um) sections of the coating by the formation of surface microcracks for annealing at temperature > 400°C, as shown in Figs. 6-l-3b and c. A corresponding phase transformation from amorphous to crystalline structure takes place at these temperatures, Figs. 6-1-1 and 6-1-2. 101 Figure 6-1-9 The morphology of the porous surface dental implant (a) before and (b) after sol-gel HAp coating. A closer look of the coated surface (c) shows pores and cracks (arrows indicated) distributed along the neck and triple-junction areas. The surface feature of the underlying Ti beads can be observed and reproduced, indicating a thin, dense, and adhesive HAp coating. 102 The bonding strength for the coatings, -44 MPa, is superior to those produced by some alternate sol-gel depositions [70, 73] and also seems superior to those obtained using plasma-spraying deposition [130]. However, a direct comparison of bonding strength between the sol-gel H A p layer and those plasma-sprayed H A p layers may be misleading. This is because large difference in layer thickness, e.g., > 50 um for the plasma-sprayed coating but only < 1 urn for current sol-gel film, and a decreased bonding strength with increased thickness, as disclosed in the plasma-spraying literature. Penetration of the epoxy glue to the substrate through the surface microcracks could have also contributed to the measured bond strength. However, because the relatively small population of the through-thickness cracks (especially for 375°C where no cracks were visually observed from the surface and in cross-section regions), and because of relatively high viscosity of the epoxy glue, this contribution can thus be ignore. The interfacial bonding between ceramic coating and metallic substrate can be improved at higher temperatures, for instance, by forming solid solution such as calcium titanate and/or T i - P compounds [19, 22, 34, 54, 132]. In general, for most coatings, the adhesive strength is a combined result of mechanical interlocking and chemical bonding between the coating and the underlying substrate. For the presently investigated HAp/ 316L stainless steel system, the interlocking component of adhesion was maximized through surface roughening, similar to that observed for plasma spray coating reported in the literature. There is no direct evidence supporting the chemical bonding contribution to adhesion in this 103 system at present. However, an increase in absorption intensity of the CO3 group at band 871 cm"1 with temperature (Fig. 6-1-2) may suggest increased incorporation of the carbonate into the apatite coating, resulting in a carbonated apatite which may form bonds with surface oxides on the stainless steel substrate. 104 6-2 In-Vitro Test 6-2-1 Deposit Formation Figures 6-2-la, b, and c show the microstructural evolution the calcium phosphate layer (CPL) deposited on the HAp-coated substrate which was prepared by annealing the SG-HAp film at 400°C for 20 min. (hereafter termed 400-ACS) after 3, 7, and 14 days immersion in SBF, respectively. It is apparent that the C P L layer covered most of the surface area of the substrate after 3 days of immersion. Few surface pores are present in the deposition layers. These pores develop as a result of poor assembly of smaller particles, since the deposit layer, upon a closer examination, is essentially an assembly of small spherical particles of about 100-150 nm in diameter. These small particles remained nearly identical in size and shape over the entire time frame of immersion, suggesting a constant growth rate of the crystals and a corresponding growth mechanism remains unchanged over the immersion duration. The deposit layer thickness (HCPL) was found to increase linearly with time of immersion (t) according to Fig. 6-2-2, with the correlation coefficient of 7^=0.995 and 0.988 for 375-ACS (annealed at 375°C for 60 min. hereafter termed, 375-ACS) and 400-ACS substrates, respectively. Such a linear correlation suggests a constant growth rate of the deposit layer within the time span of immersion, irrespective of the difference in the apatitic structure of the underlying SG-HAp coating. 105 Figure 6-2-1 Surface microstructural evolution of in-vitro calcium phosphate layer on the 400-ACS substrate for an immersion time period of (a) 3, (b) 7, and (c) 14 days. Bar: 5 pm. 106 I Days of Immersion Figure 6-2-2 C P L deposit layer thickness as a function of time of immersion, showing a linear relationship for both substrates. This linearity can be further described via a simple mathematical expression of Eq. (6-2-1): HCPL=k't (6-2-1) The slope k\ referred to as a rate constant, suggests a constant deposition rate which has a value of £ j 7 5 = 0.22 um/day for the 375-ACS and km = 0.43 um/day for the 400-ACS. The latter shows a deposition rate slightly faster than that observed by Wen et al. [26]. It should be noted that the solid lines in Fig. 6-2-2 intercept the point near the origin, suggesting a relatively short induction time for the nucleation of the C P L to develop on the SG-HAp layer, which is consistent with the findings of Radin et al. [133]. Energy dispersive 107 x-ray spectroscopy (EDS) across the C P L thickness on the 400-ACS substrate shows an increase in Ca/P ratio from 1.20 at the region 1 um above the interface to 1.55 at 1 urn below the outermost layer, Fig.6-2-3, indicating a calcium-deficient apatite, which is similar to the structure of the underlying SG-HAp thin film [134]. The pH of the SBF increases with immersion time, as shown in Fig. 6-2-4. Such a pH-time correlation may explain the resulting increased Ca/P ratio of the deposit layer by the fact that greater amount of C a 2 + ions tend to adsorb onto the surface with higher negatively-charged character (i.e. in the higher pH environment) [134-137]. 1.7 -i 0 1 2 3 4 5 6 Distance from Interface (um) Figure 6-2-3 The variation of the Ca/P ratio along the C P L deposit layer. 108 7.35 -I - , , , 0 5 10 15 Days of Immersion Figure 6-2-4 The pH of the simulated body fluid increases with time of immersion and reaches plateau after about 10 days of test. 6-2-2 Microstructural Examination of the CPL Figure 6-2-5 illustrates the surface morphology of the crystalline C P L deposit grown on the underlying SG-HAp film after 14 days of immersion in SBF. This film shows extensive surface cracking resulting from drying of the gel-like deposit materials, as has frequently been observed in in-vitro studies [29, 135]. However, higher magnification micrographs, Fig. 6-2-6a and b, show different morphological features of the C P L on 375-ACS and 400-ACS substrates, respectively, after 14 days of immersion. Loose packing of the deposited particles 109 was observed for the 375-ACS, resulting in a pore size of about 190 nm. These apatitic particles are uniformly distributed and are about 0.32 - 0.35 pm in diameter. Figure 6-2-5 Surface morphology of the C P L deposit layer, 14-day immersion in simulated body fluid. For the 400-ACS substrate, denser packing with cauliflower-like particles was observed, Fig. 6-2-6b. However, the individual apatite particles are hardly distinguishable, as in case of Fig.6-2-6a, but show as clusters. The clusters have an apparent size of about 0.4 - 0.5 pm, which is somewhat larger than those observed in Fig.6-2-6a. Few crevice-like pores, about 60 nm wide and 0.1-0.5 pm long, are found on the surface of the C P L . Cross-sectional examination shows the C P L has an average of about 3 - 3.2 pm in thickness for the 375-ACS substrate and 6 - 6.2 pm for the 400-ACS substrate, as illustrated in Fig.6-2-7a and b respectively. Porous structure is seen for the 375-ACS substrate (consistent with surface morphology in Fig.6-2-6a), rather dense structure is observed for the 400-ACS substrate. It 110 should be noted that interfacial cracking appeared on the 375-ACS substrate possibly due to rather porous and weak particulate structure of the deposit layer. However, adhesive deposit was found for the 400-ACS substrate, Fig. 6-3-7b. Figure 6-2-6 Surface microstructure of the deposit layer in-vitro on (a) 375-ACS and (b) 400 A C S substrates, after 14-day immersion in SBF. Bar = 2 um. Figure 6-2-7 Cross-sectional view of the deposit C P L layer on (a) 375-ACS and (b) 400-ACS substrates, after 14-day immersion in SBF. Bar = 3 um. Ill 6-2-3 IR Analysis Figures 6-2-8a and b illustrate the FTIR spectra of the 375-ACS and 400-ACS substrates, respectively, before and after 14 days immersion. For the 375-ACS substrate, the bands at 711 cm"1 and 740 cm"1 attributed to NO3 residue were observed before immersion. However, they disappeared after the immersion, and instead nstead, clearly distinct absorption bands at 563 cm"1 and 600 cm"1, corresponding to V4 PO4 groups in apatitic environment, associated with a shoulder at 630 cm"1, assigned to O H groups, were detected. (a) 375-ACS substrate rrtw^f^- before immersion .after 2000 1600 1200 800 400 Wavenumber (cm-1) (b) 400-ACS substrate ^ r m i H B k before immersion .after "HIM 2000 1600 1200 800 400 Wavenumber (cm'1) Figure 6-2-8 FTIR spectra for the (a) 375-ACS and (b) 400-ACS substrates before and after the in-vitro immersion test for a time period of 14 days. The presence of the above-mentioned splitting absorption bands in the region of 600 -500 cm"1 suggests the development of crystalline apatitic structure in C P L . This is in contrast to the spectrum before immersion, where an ill-defined absorption. band in the same wavenumber region was detected, indicative of amorphous or poorly crystalline apatitic structure. For the 400-ACS substrate, Fig. 6-2-8b, the IR spectrum after immersion is identical 112 to the spectrum observed in the 375-ACS substrate subject to the same immersion. The residual NO3 groups also disappeared as a result of dissolution. However, the absorption bands at 600 - 500 cm"1 and 1100 - 950 cm"1 regions corresponding to the PO4 groups remain unchanged after the immersion, suggesting that the deposit crystals have similar crystalline structure as the underlying SG-HAp. A broad band over the region of 1500 - 1400 cm"1 together with an absorption band at 871 cm"1 appeared after immersion suggests the incorporation of carbonate groups in the apatite crystals, indicating that the deposit layer is essentially a carbonated apatite, resembling that of human mineralized tissue. 6-2-4 Discussion Calcium phosphate minerals are unable to nucleate and grow on native (i.e., oxidized) metallic surfaces, e.g. titanium alloys or stainless steel, typically used for biomedical devices. At the same time in-vitro deposition of the films at near-ambient temperature opens up a possibility for drug encapsulation to affect tissue response to the surface of an implant. Therefore, modification of such native metallic surfaces is necessary, and has been undertaken in the past using variety of chemical and/or physical treatments. In this work the formation of apatite-like layer (CPL) on stainless steel has been imparted onto the surface through a pre-coated thin layer of apatite using sol-gel route. This result suggests that such an in-vitro surface bioactivity can be achieved and is considered as the prerequisite for in-vivo bond formation [19, 132]. 113 The formation of C P L has been well recognized as a result of nucleation-growth mechanism on substrate surface in simulated physiological environment. However, it is important to form a calcium phosphate layer with continuous structure, rather than discrete particles, on the substrate. In a recent study, Wen et al. [26] observed various modes of C P L deposit, from discrete particles to uniform Ca-P layer, on chemically modified Ti or Ti alloys surfaces after 14 days immersion into Hank's solution. The solution had C a 2 + and HPO4 2 " concentration about half of that used in the present SBF protocol. The C P L formation was controlled by the nucleation sites present on the metal surface, which is esentially a function of surface topology [145] and surface OH" groups [138]. The latter can be understood as a result of C a 2 + adsorption, following interaction with the surrounding phosphate ions to form calcium phosphate precipitates. Here we have undertaken an entirely different approach to surface modification of metallic substrates to encourage in-vitro deposition of calcium phosphates, i.e. through deposition of an intermediate thin sol-gel HAp. The interfacial SG-HAp (0.3-0.6 pm thin) coating showed high biological activity by inducing formation of a uniform layer of CPL, starting immediately after immersion in SBF. For the 375-ACS substrate, a porous C P L deposited with a growth rate slower than that formed on the 400-ACS substrate, suggesting a higher activity of the 400-ACS. Close examination of the IR spectra of the 400-ACS substrates before the immersion reveals a discernable shoulder at 630 cm"1, corresponding to vibrational mode of OH" groups in the apatite structure. This shoulder band became weaker (but can still 114 be seen, Fig.6-2-8) for the 375-ACS substrate. This indicates higher population of the OH" groups on the 400-ACS substrates, resulting in a greater amount of nucleation sites for C P L formation. Since OH" groups attract C a 2 + ion in the solution, improved surface activity is certainly expected for the 400-ACS substrate. This could be the reason where a denser C P L structure developed on the 400-ACS substrate as compared to that developed on the 375-ACS substrate (Fig. 6-2-7). The growth of nuclei in a direction parallel to the substrate surface would be constrained because of the abundant nucleation sites. This hypothesis may explain both the faster deposition rate and an assembly of particulate clusters observed on the 400-ACS substrate. The observed deposition rate of the C P L may be related to the growth kinetics of the nuclei once they were stably developed on the surface. According to classic kinetics of surface controlled crystal growth process, the overall crystal growth rate is a combination of bulk diffusion and surface reaction. Mullin [138] expressed such a surface mineralization by simple empirical equation; R = tea" (6-2-2) where the growth rate (R) is a function of k, kinetic constant, and s is a parameter proportional to the number of nucleation sites (m) on the surface, and a is the relative supersaturation ratio. The exponent n is the order of reaction, i.e. for surface reaction controlled growth, n = 2, and for a bulk diffusion controlled process, n = 1. In a recent study, Combes et al. [139] 115 investigated the growth kinetics of calcium phosphate layer on titanium particles and defined crystal growth rate (R) as; {Atm0As) where ^ / ^ t represents the weight change of the particulate substrate of initial weight mo and surface area A S during deposition, and C is the effective concentration of the solution with respect to the deposit phase. By analogy to the growth rate of the deposit phase, Eq. (6-2-3) can be modified and used for current study by taking the term ^/^t a s a ^ cP'^CPiPcply/^t, where AQPL is the surface area of the deposit phase on the model substrate and is the same as AS for dense structure, H c p i is thickness, and P C P L is the density of the deposit layer. Therefore, Eq (6-2-3) can be rearranged as; _ ^HCPL ACPLpCPLC (6-2-4) Ar m0As In combination with Eq. (6-2-1), Eq. (6-2-4) becomes m0A. or for dense deposit structure, 116 R = K , ACPLPCP,C ( 6 2 5 A ) R = k> Rent- (6-2-5b) It is apparent from Eq. (6-2-5a) or (6-2-5b) that the crystal growth rate (R) in current system should be constant over the entire immersion time period of study, since all the parameters, in particular k\ appear to be constant. The observed uniform size and shape of the deposited particles is supporting evidence. Furthermore, since the kwo = 2kJ75, it is reasonable to expect from Eqn. (6-2-5b) thati? 4 0 0 = 2R375, indicating that the crystal growth rate on the 400-ACS substrate is nearly double that on the 375-ACS substrate. This means that a coarser grain size should be attained for the 400-ACS than that for the 375-ACS in a given time period of deposition, and this is consistent with surface microstructural evolution observed in Fig. 6-2-6. However, as mentioned before, higher population of the nucleation sites on the 400-ACS substrate would limit the developing crystal size in a direction parallel to the substrate surface, on the contrary, resulting in a higher growth rate along the direction perpendicular to the substrate surface. Further nucleation and growth on the surface of the deposit crystals results in a cluster-like surface morphology (Fig. 6-2-6a). According to Eq. (6-2-2), crystal growth rate (R) should be proportional to the number of nucleation sites (m) on the substrate surface because both k' and a are constant under current experimental conditions. Therefore, it is concluded that by analogy, m 4 0 0 = 2w 3 7 5 , i.e. nearly twice as many nucleation sites developed on the 400°C-annealed substrate than on that prepared at 375°C. This is in good agreement with previous observation on the O H groups, 117 possibly acting as nucleation sites, via the IR spectra shown in Fig. 6-2-8. However, it is difficult to determine whether there are about twice as much as O H groups on the 400-ACS substrate in comparison to the 375-ACS substrate. It should be noted that the analysis given above should only be valid during the initial time period of immersion, where the nature of substrate surface significantly affects deposition. Once the first C P L layer develops, further nucleation and growth of the calcium phosphate crystals should rely solely on the physical and chemical properties of the "first" C P L layer, rather than the original substrate surface. However, microstructural examination as revealed in Fig.6-2-7 suggests the growth of the C P L is microstructurally consistent across the layer for each case. This implies that a subsequent nucleation and growth of the calcium phosphate crystals on the "first" layer of C P L would be identical to that occurred in developing the "first" layer. If this assumption is true, then further deposition of the "second" or subsequent layers should follow the same mechanism that occurred in the first one, i f the surrounding environmental conditions are held constant. However, dissolution of residual nitrate from the SG-HAp films (as indicated by IR spectra in Fig. 6-2-8), may locally decrease pH, especially near the region slightly above the substrate-SBF interface. Such a localized pH reduction may lower Ca/P ratio in the CPL, since it would favor the formation of acidic calcium phosphate deposits (i.e., with lower Ca/P), especially in early period of deposition. The Ca/P ratio increases rapidly for C P L thickness between 2 pm and 4 pm (i.e. during a time period of growth from 4 days to 10 days), Fig. 118 6-2-3. Subsequently, Ca/P increase is slower, which may reflect equilibration of the film with the corresponding solution (as confirmed by pH monitoring, Fig. 6-2-4). This observation may be utilized to tune the Ca/P ratio to a desirable value during C P L formation via manipulating solution pH, i.e., higher solution pH promotes C P L with higher Ca/P ratio and vice verse. 6-3 Hydroxyapatite Coatings for Drug Delivery In this work we report a new approach to achieve the active Ca-P layer on metallic surface, capable of encapsulating drugs that can be hydrophilic or hydrophobic character. This approach combines deposition of several layers of Ca-P, each fulfilling a specific role in the resulting functionally gradient Ca-P coating. Specifically, a 0.6-0.8 um thin layer of sol-gel HAp (SG-HAp) was first deposited onto metallic substrate and heat treated at a temperature of 400°C as described previously in Chapter 6-1. This constitutes a well-adhesive interfacial HAp film of low porosity and good crystallinity. The role of this film is to modify the metal surface to provide nucleation sites, for a subsequent room-temperature deposition of a second thin layer for drug delivery purpose. We found that subsequent deposition of a second, and later third etc., apatite layer can be achieved at ambient environment through a controlled chemical pathway. In this application, a conventional calcium phosphate cement composition was employed. Accordingly, it is expected that a hardening effect as a result of acid-base neutralization reaction occurring in the cement composition [140-142] is able to provide 119 strength as well as adhesion to the SG-HAp film. Here, Ca(OH)2 was used as basic precursor and Ca(H2P04)2 .H20 as acidic precursor. Figure 6-3-1 shows a comparison of the coatings appearance on the two substrates (i.e. with and without SG-HAp film). The layer dip-coated in C P C slurry directly on 316 steel (i.e. without the SG-HA pre-treatment) showed essentially no adhesion and the thick CPC-HAp film largely spalled. In contrast, the thick CPC-HAp film shows good adhesion with the SG-HAp surface pre-treatment. (without SG-HAp pre-coating) (with SG-HAp coating) Figure 6-3-1 The C P C - H A p layer showed extensive debonding for the model substrate without pre-SG-HAp-coated treatment; however, for the substrate with the SG-HAp pre-treatment, a well-attached CPC layer was occurred. Both C P C layers have the same thickness of about 20 urn. 120 X-ray diffraction analysis of the thick CPC-HAp layer shows a poorly crystalline apatitic structure together with trace amount of starting precursors, indicating incomplete reaction, Fig. 6-3-2. The poorly-crystalline apatite has X R D characteristics similar to that of naturally occurring apatite. Since the starting Ca/P ratio in the colloidal suspension is set at 1.5, the apatite obtained is then a calcium-deficient apatite as further illustrated in IR spectrum in Fig.6-3-3b. The reaction can be expressed as a result of interaction between Ca(OH)2 and Ca(H2P04)2 in the layer matrix: . 3Ca(H 2 P0 4 ) 2 + 6 Ca(OH) 2 -> Ca 9(P04)5(HP0 4)(OH) + 11H 2 0 (6-3-1) 40 35 30 Two Theta, (degree) 25 Figure 6-3-2 X R D pattern of the CPC layer (bottom curve), showing a poorly-crystalline apatitic structure. The X R D pattern of the commercial H A p (top curve) is used for comparison purpose. 121 Equation (6-3-1) illustrates an acid-base interaction that is frequently observed in many calcium phosphate cement systems (CPC) [140-142]. Accordingly the coating is termed CPC coating, or CPC-HAp in short. The pull-out test performed according to A S T M C-633 standard revealed adhesion strength of the thick CPC-HAp film to the thin SG-HAp film of 6 + 2.3 MPa (test on 6 specimens with -20 urn thick coating). This bonding strength is greater than the tensile strength of bulk CPC-HAp reported in literature, i.e. generally < 2 MPa [142]. The fracture mode was predominantly at the SG-HAp/CPC-HAp interface, indicating that the interface is the weakest link region. Interfacial examination (Fig.6-3-3) shows that the CPC-HAp coating is in intimate contact with the underlying SG-HAp thin layer. Therefore, it is reasonable to hypothesize that the pre-coated thin SG-HAp film acts as a template for the nucleation and growth of precipitated apatite in the CPC-HAp coating. Such mechanism allows an interfacial bonding to develop between the SG-HAp and the CPC-HAp, as evidenced by the measured bonding strength. The IR spectrum of the coating is shown in Fig.6-3-3b. The bands ranging from 600cm"1 to 500 cm"1 and 1100 cm"1 to 900 cm"1 indicate characteristic absorption band of PO4 groups in the apatitic environment. This broadness of the absorption bands, consistent with that observed in the X R D spectrum (Fig.6-3-2), suggests the randomness of the crystal structure, indicating an amorphous or fine, poorly crystalline apatite. The absorption band at 871 cm"1, which is assigned to CO3 group and/or HPO4 group, because the absorption bands of 122 both groups are overlapped at 871 cm" . Therefore, it can be concluded that the coating is essentially a calcium-deficient, CO3 and HPG-4-containing apatite. (b) CO3/HPO4 (a) Figure 6-3-3 Cross-section examination of the CPC-HAp coating confirms the development of a well-bonded SG-HAp - CPC-HAp, (as well as an adhesive SG-HAp - substrate interface), resulting in an interfacial strength of 6 + 2.3 MPa. Infrared analysis (b) shows that the CPC-HAp layer is essentially a calcium-deficient, carbonate apatite, closely resembling that of human bone mineral. 123 The band at 871 cm"1 indicates a B-type (for P 0 4 site) carbonate apatite. The end product may then have a chemical form as; Ca9(P04)5-x-y(C03)y(HP04)x(OH),.y/3 The CPC-HAp layer had a porosity of about 45% and an average pore size of about 16 nm as determined by B E T nitrogen sorption (AutoPore 9100). It is believed that these fine pores can form conduits for the release control, i.e., by diffusion, of drug molecules. As some of these pores have 2-5 nm in diameter, these nanopores in the C P C - H A p film may physically immobilize the biomolecules of similar dimensions [143, 144]. A preliminary evaluation of the drug release from the C P C - H A p layer was performed according to the protocol described in Section 4-3. Figure 6-3-4 illustrates the drug release kinetics for the time period of 72 hours, from the drug-loaded C P C - H A p . Although a "burst" effect was detected for both coatings over the initial period of about 8 h, a slower release is evident for the sample post-coated (by dipping) with a thin layer of poly(vinyl alcohol) film. Such a burst effect may be due to a rapid dissolution and release of the drug that being deposited near the outermost surface of the CPC-HAp layer. The presence of drug agglomerates due to in-homogeneous distribution upon drug loading may be a factor of the resulting burst effect. However, in some aspects, the burst effect may offer an advantage, for instance, in the early period of recovery after orthopedic/dental surgeries i f anti-inflammatory 124 agents were incorporated into the implant devices to avoid acute or severe inflammatory response. However, a further investigation on the interaction between drug molecules and CPC-HAp matrix, together with the effect of drug concentration in the CPC-HAp layer on a subsequent release behavior wil l be a part of future work. A linear relationship was observed (inset, Fig.6-3-4) between the amount of drug released and (time) for the release time greater than 8 h. It is therefore concluded that the sustained release period for 8 to 60 hours should be well described by Fick's law of diffusion. This also confirms the presence of nanometric conduits within the C P C - H A p layer that regulates the release rate of the drug from within the layer. The release kinetic is further modified by a post-coated thin PVA film, suggesting the diffusion of the drug being further reduced when across the P V A thin layer. Therefore, on this basis, it is believed that the CPC-HAp layer developed in this novel application is able to act as a drug depot for drug delivery purpose, although a pertinent drug release pattern needs to be further systematically studied for varying practical requirements 125 Figure 6-3-4 A sustained release of the model drug (amethoperin, 5% by weight with respect to the weight of the CPC-HAp layer, 20 mg) from the CPC-HAp coating for a time period of 72 h. A rapid release (the "burst" effect), was detected for the first 8 h, after then, following a linear t l / 2 - dependent behavior (inset). With a post-coated PVA film on the CPC-HAp layer, the burst effect was reduced to a certain extent, indicating the release pattern can be adjusted for practical needs. 126 Chapter 7: Conclusions Water-based sol-gel hydroxyapatite (HAp) was successfully synthesized and characterized using the novel alkoxide-salt precursor system including triethyl phosphite and calcium nitrate salt. To our knowledge, this is the first and original report among the related studies disclosed in literature. This was also recognized by US Patent Office through award of two patents: (1) "Novel Sol-Gel Hydroxyapatite Ceramic Coatings and Method of Making Same", US Patent No. 6,426,114 B l , awarded Jul. 30, 2002 (the mirror Canadian patent has No. 2,345,552); (2) "Biofunctional Calcium Phosphate Coatings and Microspheres for Drug Encapsulation", US Appl . No. 2002/0155144 A l , Oct. 24, 2002. Both patents have been already licensed to local biotechnology company. A number of related publications followed inRefs. 25,102,106,134, and 146-159. The following principal conclusions result from this work: 1. The proposed process for HAp synthesis allows for relatively low temperature for HAp phase formation, i.e. < 400°C, as compared to at least 500°C reported in literature for sol-gel H A p using other precursor systems. Such a low synthesis temperature provides a new avenue for future research leading to better understanding of HAp sol-gel chemistry, where a strong and rapid interaction occurs between the calcium ions and hydrolyzed triethyl phosphite. 2. The time of phase-pure HAp synthesis through triethyl phosphite / calcium nitrate route is 127 shortened by up to 2 orders of magnitude upon the use of an acidic catalyst, i.e. to ~5 minutes, as compared to the conventional sol-gel hydroxyapatite synthesis, i.e. 10-24 hours. Such a significant shortening of the time span for HAp synthesis suggests that both the hydrolysis and condensation rates are considerably enhanced in the presence of the acidic catalyst. 3. The use of the highly chemically active triethyl phosphite P(III), as compared to less active triethyl phosphate P(V) employed in many reports, is responsible for the enhanced reactivity in the triethyl phosphite / calcium nitrate route. We suggest that an increase in ligand coordination from P(III) to P(V) upon hydrolysis facilitates further interaction with calcium ions. 4. Reaction paths in the triethyl phosphite / calcium nitrate system, from sol preparation to the final phase-pure H A p formation, are proposed and experimentally verified. We propose that monitoring of proton release during the reaction provides a valuable evidence for process evaluation through simple in-process pH monitoring. Incomplete reaction leads to a lower proton concentration in the sol, resulting in a multi-phase structure, including HAp, TCP, and some residual nitrate phase. Such systems require higher temperatures, e.g. 500°C, to convert the mixed phases into single-phase crystalline HAp. 128 5. Thorough studies of sol ageing were performed and allowed to chart the phase formation map for the triethyl phosphite / calcium nitrate precursor system. The phase formation map allows rapid identification of the synthesis conditions leading either to the phase-pure HAp or formation of diphasic HAp-TCP composite for different practical needs. Optimum ageing of the sol, easy to identify using the phase formation map, is necessary to allow phase-pure HAp formation at temperatures below 400°C. 6. The novel triethyl phosphite / calcium nitrate precursor system has been utilized to deposit through dip-coating thin films (-0.3-0.8 um) of H A p on stainless steel and titanium substrates, both flat and roughened by sand blasting. The phase transformations in coatings during sol-gel processing were essentially the same as for the bulk materials. 7. A strong interfacial bonding strength of about 44 MPa was achieved by dip-coating stainless steel substrate by the sol following by annealing at 400°C. Such a strong bonding achieved at such a low temperature (believed to be achieved for the first time for sol-gel HAp) may suggest development of a chemical bond between the H A p and the substrate. Although no direct evidence can be provided in this study, it is hypothesized that interaction of surface oxide or surface hydroxyl group of the substrate with the hydroxyl ligands of the sol takes place. However, confirmation of this hypothesis requires further work. If the density of the hydroxyl groups between the substrate and the sol can be controlled via either 129 chemical and physical means, it should possible to further enhance the bonding strength at lower temperature. In such a case, the use of an organic polymer (or bio-polymer) substrate for H A p fdms may be a possible alternative. 8. Bioactivity of the H A p coatings has been tested in vitro. It is concluded that annealing of the HAp coatings at 400°C should be sufficient to impart bioactivity to the coated substrate, where a sufficient amount of the hydroxyl groups on the surface of the H A p coating is responsible for the in-vitro bioactivity. Based on these results, it can then be expected that in-vivo bioactivity is achievable for the sol-gel HAp coatings processed via the novel triethyl phosphite / calcium nitrate precursor system. 9. The thin films of sol-gel H A p processed at elevated temperatures have been used for room-temperature nucleation and growth of secondary, thicker calcium phosphate layers (CPL) through biomimetic route, in preparation for future work on drug encapsulation and cells activity. The slow growth kinetics of C P L on H A p coating surface suggests a diffusion-controlled mechanism. The method is slow and requires several days to achieve appreciable (several micrometers) thickness of the film. A n important finding of this work is that the microstructural evolution of the C P L depends on the chemical nature (particularly presence of the surface hydroxyl groups) of the HAp coating. This implies that the surface texture of the C P L can be controllable at a micrometric or even nanometric scale. If this is 130 technologically feasible, then, it wil l be interesting to study in the future the effect of surface texture of the C P L on cell activity. 10. The thin films of sol-gel H A p processed at elevated temperatures have been used for room-temperature nucleation and growth of secondary, thicker calcium phosphate layers through calcium phosphate cement (CPC) route. The calcium phosphate cement converted to apatitic structure in the presence of water via a neutralization reaction between monocalcium phosphate monohydrate and calcium hydroxide. The method is rapid and allows deposition of > lOOum thick films in 24hr cycle. It has been demonstrated that such a room temperature deposition process allows in-situ drug encapsulation and a subsequent controlled release of the drug. 131 Chapter 8 Future Work Following the successful development and extensive investigation of the novel water-based sol-gel hydroxyapatite in this work, there are several research directions that need to be further explored as future work: 1. Future experimental and theoretical work should determine i f it is possible to further reduce the HAp phase formation temperature, using the current precursor system of triethyl phosphite / calcium nitrate route, or after some chemical modification. We believe that this is possible to achieve through control of solution chemistry, pH or ligand modification of the precursors. A change in precursor composition, especially the nitrate salt used in this work, may be a favorable alternative. If the phase formation temperature can be further reduced, then the reaction paths in both the sol solution and solid phase should be explored and the process applied for coating surfaces of polymeric materials. 2. In-vivo bioactivity of the hydroxyapatite coating is important object to explore in future studies. This is especially interesting if a thin calcium phosphate layer of different surface texture can be first developed using in-vitro deposition on the coating surface for a subsequent study of cell behaviors (such as osteoblast). 3. Further enhancement of the interfacial bonding strength between the CPC-HAp layer and its underlying SG-HAp coating is required for applications subject to mechanical loads, such as dental or hip implants. 132 REFERENCES 1. M . Jarcho, "Calcium Phosphate Ceramics as Hard Tissue Prosthetics", Clin. Orthop., 157, 259-278 (1981). 2. H . Aoki , K . Kato, M . Ebihara, and M . Inoue, "Studies on the Applications of Apatite on Dental Materials", J. Den. Eng (Jpn), 17, 200 (1976). 3. P. Ducheyne, L . L . Hench, A . Kagan, M . Martens, A . Bursens, and J. C. Mulier, "Effect of Hydroxyapatite Impregnation on Skeletal Bonding of Porous Coated Implants", J. Biomed. Mater. Res., 14, 225 (1980). 4. S. N . Bhaskar, J. M . Brady, L . Getler, M . P. Grower, and T. Driskell, "Biodegradable Ceramic Implant in Bone", Oral. Surg., 32, 336 (1971). 5. J. L . Berry, J. M . Geiger, J. M . Moran, J. S. Skraba, and A . S. Greenwald, "Use of Tricalcium Phosphate of Electrical Stimulation to Enhance the Bone-Porous Implant Interface", J. Biomed. Mater. Res., 20, 65 (1986) 6. G. Daculsi, R. Z. LeGeros, E. Nery, K . Lynch, and B . Kerbel, "Transformation of Biphasic Calcium Phosphate Ceramics in Vivo: Ultrastructural and Physico-Chemical Characterization", J. Biomed. Mater. Res., 23, 883 (1989). 7. M . Jarcho, J. F. Kay, K . I. Gumbaer, R. H . Doremus, and H . P. Drobeck, "Tissue, Cellular, and Subcellular Events at a Bone Ceramic Hydroxyapatite Interface", J. Bioeng., 1, 79 (1977). 8. A . S. Posner, "Crystal Chemistry of Bone Mineral", Physio. Rev., 49[4], 760-92 (1969). 9. D. McConnell, "Apatite", New York, Springer-Verlag, 1973. 10. E. E. Berry, "The Structure and Composition of Some Calcium-Deficient Apatite", J. Inorg. Nucle. Chem., 29, 317-27 (1967). 11. J. E. Eastoe, "The Chemical Composition of Bone", pp. 715-720 in Biochemists' Handbook, edited by C. Long, Princeton, N . J. Van Nostrand, 1961. 12. M . J. Glimcher, "Recent Studies of the Mineral Phase in Bone and Its Possible Linkage to the Organic Matrix by Protein-Bound Phosphate Bonds", Trans. R. Soc. London, 304B 479-508 (1984). 133 13. H . Yuan, Z . Yang, Y. L i , X . Zhang, J. D., Bruijn, and K . DeGroot, "Osteoinduction by Calcium Phosphate Biomaterials", Mater. Med., 9, 723-26 (1998). 14. U . Ripamonti, "The Morphogenesis of Bone in Replicas of Porous Hydroxyapatite obtained from Conversion of Calcium Carbonate Exoskeletons of Coral", J. Bone Joint Surg., 73A, 692 (1991). 15. K . Vargerik, pp.112-20 in "Bone Grafts and Bone Substitutes", edited by M . B . Habal and A . H . Reddi, (W. B . Saunders, 1992). 16. H . Yamasaki and H . Sakai, "Osteogenic Response to Porous Hydroxyapatite Ceramics Under the Skin of Dogs", Biomaterials, 13, 308 (1992). 17. L . L . Hench, "Bioceramics: From Concept to Clinic", J. Amer. Ceram. Soc, 74, 1487-1510 (1991). 18. J. B . Park, Biomaterials Science and Engineering, Plenum Press, New York, 1987. 19. C. Ohtsuki, T. Kokubo, and T. Yamamuru, Mechanism of Apatite Formation on Calcium Oxide-Silica-Phosphorus Pentoxide Glasses in a Simulated Body Fluid", J. Non-Cryst. Solids, 143, 84-92 (1992). 20. P. Ducheyne, S. Radin, and K . Ishikawa, in Bone-Bonding Biomaterials, P. Duchenye, T. Kokubo, and C. A . van Blitterswijk (eds), Reed Healthcare Communications, Leiderdrop, pp. 213-18 (1992). 21. D. M . Liu , H . M . Chou, and J. D. Wu, "Plasma-Sprayed Hydroxyapatite Coatings: Effect of Different Calcium Phosphate Ceramics", J. Mater. Sci. Mater. Med., 5, 147-153 (1994). 22. K . de Groot, R. G. T. Greesink, C. P. A . T. Klein, and P. Serekian,, "Pasma-Sprayed Hydroxyapatite Coatings of Hydroxyapatite", J. Biomed. Mater. Res., 21, 1375-1381 (1987). 23. C. S. Chai, B . Ben-Nissan, S. Pyke, and L . Evans, "Sol-Gel Derived Hydroxyapatite Coatings for Biomedical Applications", Mater. Manuf. Processes, 10, 205-216 (1995). 24. C. M . Lopatin, V. Pizziconi, T. L . Alford, and T. Laursen, "Hydroxyapatite Powders and Thin Films Prepared by a Sol-Gel Technique", Thin Solid Films, 326, 227-232 (1998). 25. D. M . Liu, T. Troczynski, and W. J. Tseng, "Water-based Sol-Gel Synthesis of 134 Hydroxyapatite: Process Development", Biomaterials, 22, 1721-1730 (2001). 26. H . B . Wen, J. R. de Wijn, F. Z. Cui, and K . de Groot, "Preparation of Calcium Phosphate Coatings on Titanium Implant Materials by Simple Chemistry", J. Biomed. Mater. Res., 41, 227-36 (1998). 27. H . B . Wen, J. G C. Wolke, J. R. de Wijn, Q. Lin, F. Z. Cui, and K . de Groot, "Fast Precipitation of Calcium Phosphate Layers on Titanium Introduced by Simply Chemical Treatments", Biomaterials, 18, 1471-78 (1997). 28. A . A . Campbell, G E. Fryxell, J. C. Linehan, and G. L . Graff, "Surface-Induced Mineralization: A New Method for Producing Calcoium Phosphate Coatings", J. Biomed. Mater. Res., 32, 111-118 (1996). 29. Y. Abe, T. Kokubo, and T. Yamamuru, "Apatite Coating on Ceramics, Metals, and Polymers Utilizing a Biological Process", Materials in Madicine, 1, 233-238 (1990). 30. H . M . K im, Y. K im, S. J. Park, C. Rey, H . M . Lee, M . J. Glimcher, and J. S. Ko, "Thin fdm of low-crystalline calcium phosphate apatite formed at low temperature", Biomaterials, 21, 1129 (2000). 31. P. K. Stephenson, M . A . R. Freeman, P. A . Rovall, J. Germain, M . Tuke, and C. J. Piris, "The Effect of Hydroxyapatite Coating on Ingrowth of Bone into Cavities in an Implant", J. Arthoplasty, 6, 51-58 (1991). 32. K . Soballe, S. Overgaard, E. S. Hansen, H . Brokstear-Rasmussen, M . Lind, and C. Bunger, " A Review of Ceramic Coatings for Implant Fixation", J. Long Term Eff. Med. Implants, 9[l-2], 131-151 (1999). 33. W. Suchanek and M . Yoshimura, "Processing and Properties of Hydroxyapatite-based Biomaterials for Use as Hard Tissue Replacement Implants", J. Mater. Res., 13[1], 94-117 (1998). 34. P. Ducheyne and J. M . Cuckler, "Bioactive Ceramic Prosthetic Coatings", Clin. Orthopaedic Rel. Res., 276, 102-114 (1992). 35. W. F. DeJong, " L a Substance Mineral dans les oa", Rec. Trav. Chim, 45, 445-448, 1926 36 A.S . Posner, A . Perloff, and A . F. Diorio, "Refinement of Hydroxyapatite Structure", Acta. 135 Crys., 11, 308-309, 1958. 37. R. A . Young and J. C. Elliott, "Atomic-scale based for Several Properties of Apatites", Arch. Oral Biol . , 11, 699-707, 1966. 38. T.S.B. Narasaraju and D E Phebe, "Some Physico-chemical Aspects of Hydroxyapatite", J. Mat. Sci., 31, 1-31, 1996. 39. R. A . Harper and A . S. Posner, "Measurement of Noncrystalline Calcium Phosphate in Bone Mineral, Pro. Soc. Exptl. Bio l . Med., 122, 137-142, 1966. 40. P. W. Brown, N . Hocker, and S. Hoyle, "Variations in Solution Chemistry during the Low-Temperature Formation of Hydroxyapatite", J. A m . Ceram. Soc, 74, 1848, 1991. 41. M . T. Fulmer and P.W. Brown, "Hydrolysis of dicalcium phosphate dihydrate to hydroxyapatite", Mater. Med., 9, 197-202 (1998). 42. K . Sakamoto, S. Yamaguchi, A . Nakahira, K . Kijima, and M . Okazaki, "Transformation of a-TCP to hydroxyapatite inorganic media", Bioceramics, Vol.10, edited by L . Sedel and C. Rey, pp. 241-244 in Proceedings of the 10 t h International Symposium on Ceramics in Medicine, Paris, France, Oct. 1997. 43. G . DeWith, H . J. A . Van Dijk, N . Hattu, and K . Prijs, "Mechanical Behavior of Biocompatible Hydroxyapatite Ceramics", J. Mater. Sci., 16, 1592 (1981). 44. A . J. Ruys, M . Wei, C. C. Sorrell, M . R. Dickson, A . Brandwood, B . K . Milthorpe, "Sintering Effects on the Strength of Hydroxyapatite", Biomaterials, 16, 409 (1995). 45. G . DeWith and A . J. Corbijn, "Metal Fiber-Reinforced Hydroxyapatite Ceramics", J. Mater. Sci., 24, 3411 (1989). 46. T. Nonami and N . Satoh, "Preparation of Elongated Diopside/Hydroxyapatite Composite and Their Cell Culture Test", J. Ceram. Soc. Jpn., 103, 804 (1995). 47. T. Noma, N . Shoji, S. Wada, and T. Suzuki, "Sintering of Hydroxyapatite Powders With Silicon Carbide Platelets Dispersion", J. Ceram. Soc. Jpn., 100, 1175-178 (1992). 136 48. M . Takagi, M . Mochida, N . Uchida, K . Saito, and K . Uemats, "Filter Cake Forming and Hot Isostatic Pressing for TZP-dispersed Hydroxyapatite Composite", Mater. Med., 3, 199-203 (1992). 49. T. K . Chaki and P. E. Wang, "Strengthening Behavior of Hydroxyapatite-Silver Composite", Mater. Med., 5, 533 (1994). 50. P. Ducheyne, W. Van Raemdonck, J. C. Heughebaert, and M . Heughebaert, "Calcium phosphate ceramic coatings on porous titanium: Effect of Structure and Composition on Electrophoretic Deposition, Vacuum Sintering and in-vitro Dissolution", Biomaterials, 11, 244 (1990). 51. M . Wei, A . J. Ruys, B . K . Milthorpe, and C. C. Sorrell, "Solution Ripening of Hydroxyapatite Nanoparticles: Effects on Electrophoretic Deposition", J. Biomed. Mater. Res., 45, 11-19(1999) 52. L . Torrisi and G. Foti, "Ke V Ion Sputtering of Hydroxyapatite", App. Phys. Letters, 62, 237-39 (1993). 53. H . Hero, H . Wie, R. B . Jorgensen, and I. E. Ruyter, "Hydroxyapatite Coatings on Ti Produced by Hot Isostatic Pressing", J. Biomed. Mater. Res., 28, 343-48 (1994). 54. C. C. Berndt, G. N . Haddad, A . J. D. Farmer, and K . A . Gross, "Thermal Spraying for Bioceramic Applications-A Review", Mater. Forum, 14, 161-173 (1990). 55. R. G T. Geesink, "Hydroxyapatite-Coated to Total Hip Prostheses", Clin. Orthop., 261, 39-58 (1990). 56. P. K . Stephenson, M . A . R. Freeman, P. A . Rovall, J. Germain, M . Tuke, and C. J. Piris, "The Effect of Hydroxyapatite Coating on Ingrowth of Bone into Cavities in an Implant", J. Arthoplasty, 6, 51-58 (1991). 57. M . Shirkhanzadeh, "Calcium Phosphate Coatings Prepared by Electrocrystallization from Aqueous Electrolytes", J. Mater. Sci. Mater. Med., 60, 90-93 (1995). 58. H . Monma, "Electrochemical Deposition of Calcium-Deficient Apatite on Stainless Steel Substrate", J. Ceram. Soc. Jpn. Int. Edn., 101[7], 718-20 (1993). 59. S. Ban and S. Maruno, "Deposition of Calcium Phosphate on Titanium by Electrochemical 137 Process in Simulated Body Fluid", Jpn. J. App. Phys., part 2, Letters, 32[10B], 1577-80 (1993). 60. C. M . Cotell, "Pulsed Laser Deposition and Processing of Biocompatible Hydroxyapatite Thin Film", App. Surf. Sci., 69, 140-48 (1993). 61. Y. Gao, "Synthesis and Characterization of Calciium Phosphate Coating by Metalorganic Chemical Vapor Deposition", Mat. Res. Soc. Symp. Proc, vol.550, 361-66 (1999) 62. Y. Masuda, K . Matubara, and S. Sakka, "Synthesis of Hydroxyapatite from Metal Alkoxides Through Sol-Gel Technique", J. Ceram. Soc Japan, 98, 1266-1277 (1990). 63. L . D. Piveteau, M . I. Girona, L . Schlapbach, P. Barboux, J. P. Boilot, and B . Gasser, "Thin Films of Calcium Phosphate and Titanium Dioxide by a Sol-Gel Route:, A New Method for Coating Medical Implants", J. Mater. Sci. Mater. Med., 10, 161-167 (1999). 64. T. Brendel, A . Engel, and C. Russel, "Hydroxyapatite Coatings by a Polymeric Route", J. Mater. Sci. Mater. Med., 3, 175-179 (1992). 65. K . A . Gross, C. S. Chai, G. S. K . Kannangara, B . Bin-Nissan, and L . Hanley, "Thin Hydroxyapatite Coatings via Sol-Gel Synthesis", J. Mater. Sci. Mater. Med., 9, 839-843 (1998). 66. S. W. Russel, K . A . Luptak, C. T. A . Suchicital, T. L . Alford, and V. B . Pizzicoui, "Chemical and Structural Evolution of Sol-gel-Derived Hydroxyapatite Thin Films Under Rapid Thermal Processing", J. Am. Ceram. Soc, 79[4], 837-842 (1996). 67. K . Hwang and Y. Lim, "Chemical and Structural Changes of Hydroxyapatite Films by Using a Sol-Gel Method", Surf. Coating Tech., 115, 172-175 (1999). 68. D. B . Haddow, P. F. James, and R. Van Noort, "Sol-Gel Derived Calcium Phosphate Coatings for Biomedical Applications", J. Sol-Gel Sci. Tech., 13, 261-265 (1998). 69. D. B . Haddow, P. F. James, and R. Van Noort, "Characterization of Sol-Gel Surfaces for Biomedical Applications", J. Mater. Sci. Mater. Med., 7, 255-260 (1996). 70. W. Weng and J. L . Baptista, "Sol-gel Derived Porous Hydroxyapatite Coatings", J. Mater. Sci. Mater. Med., 9, 159-163 (1998) 138 71. K . Hwang, J. Song, B . Kang, and Y. Park, "Sol-Gel Derived Hydroxyapatite Films on Alumina Substrates", Surf. Coat. Tech., 123, 252-55 (2000). 72. C. S. Chai and B . Ben-Nissan, "Bioactive Nanocrystalline Sol-Gel Hydroxyapatite Coatings", Mater. Med., 10, 465-69 (1999). 73. W. Weng and J. L . Baptista, "Preparation and Characterization of Hydroxyapatite Coating on Ti6A14V Alloy by a Sol-Gel Method", J. Am. Ceram. Soc, 82[1], 27-32 (1999). 74. S. Kaciulia, G. Mattogno, L . Pandolfi, M . Caralli, G. Gnappi, and A . Montenero, "XPS Study of Apatite-Based Coatings Prepared by Sol-Gel Technique", App. Surf. Sci., 151, 1-5 (1999). 75. L . D. Piveteau, M . I. Girona, L . Schlapbach, P. Barboux, J. P. Boilot, and B . Gasser, "Thin Films of Calcium Phosphate and Titanium Dioxide by a Sol-Gel Route: a new method for implants", Mater. Med., 10, 161-67 (1999). 76. Q. Qin, P. Vincent, B . Lowenberg, M . Sayer, and J. E. Davies, "Bone Growth on Sol-Gel Calcium Phosphate Thin Films in Vitro", Cells Mater., 3[4], 351-60 (1993). 77. B . Mavis and A . C. Tas, "Dip Coating of Calcium Hydroxyapatite on Ti-6A1-4V Substrates", J. Amer. Ceram. Soc, 83[4], 989-91 (2000) 78. R. A . Young and D. W. Holcomb, "Variability of Hydroxyapatite Preparations", Calif. Tissue Int., 34, 17-32(1982). 79. A . Osaka, Y. Miuro, K . Takeuchi, M . Asada, and K . Takahashi, "Calcium Apatite Prepared from Calcium Hydroxide and Orthophosphoric Acid" , J. Mater. Sci. Mater. Med., 2, 51-55 (1991). 80. A . Slosarczyk, E. Stobierska, Z. Paszkiewicz, and M . Gawlick, "Calcium Phosphate Materials Prepared from Precipitates with Various Calcium: Phosphorus Molar Ratios", J. A m . Ceram. Soc, 79, 2539-2544 (1996). 81. M . Yoshimura, H . Suda, K . Okamoto, and K . Ioku, "Hydrothermal Synthesis of Biocompatible Whiskers", J. Mater. Sci., 29, 3399-3402 (1994). 82. C. S. Chai, B . Ben-Nissan, S. Pyke, and L . Evans, "Sol-Gel Derived Hydroxyapatite Coatings for Biomedical Applications", Mater. Manuf. Processes, 10, 205-216 (1995). 139 83. D. B . Haddow, P. F. James, and R. Van Noort, "Sol-Gel Derived Calcium Phosphate Coatings for Biomedical Applications", J. Sol-Gel Sci. Tech., 13, 261-265 (1998). 84. A . P. Ameen, R. D. Short, R. Johns, and G. Schwach, "The Surface Analysis of Implant Materials. 1. The Surface Composition of a Titanium Dental Implant Material", Clin. Oral Implantation Res., 4, 144 (1993). 85. C. C. Ting, S. Y. Chen, and D. M . Liu, "Structural Evolution and Optical Properties of Ti02 Thin Films Prepared by Thermal Oxidation of Sputtered Ti films", J. Applied Phys. 88[8], 4628-4633 (2000). 86. R. Z . Legeros, "Calcium Phosphate Materials in Restorative Dentistry: A Review", Adv. Dent. Res., 2, 164-168 (1988). 87. M . Itokazu, W. Yang, T. Aoki , and N . Kato, "Synthesis of Antibiotic-loaded Interporous Hydroxyapatite Blocks by Vacuum Method and in Vitro Drug Release Testing", Biomaterials, 19, 817-819, (1998). 88. F. Minguez, M . Agra, S. Luruena, C. Ramos, and J. Prieto, "Post-antibiotic Effect of Isepamicin Compared to that of Other Aminoglycosides", Drugs Exp. Clin. Res., 16[5], 231-235 (1990). 89. W. Paul and C. P. Sharma, "Development of Porous Spherical Hydroxyapatite Granules: Application Towards Protein Delivery", J. Mater. Sci. Mater. Med., 10, 383-388 (1999). 90. J. A . Jansen, J. P. C. M . Van de Waerden, J. G. C. Wolke, and K . de Groot, "Histologic Evaluation of the Osseous Adaptation to Titanium and Hydroxyapatite-coated Titanium Implants", J. Biomed. Mater. Res., 25, 973-989 (1991). 91. T. W. Bauer, R. G. T. Geesink, R. Zimmerman, and J. T. Mcmahon, "Hydroxyapatite-Coated Femoral Stems", J. Bone Joint Surg., 73A, 1439-1452 (1991). 92. C. S. Chai, K . A . Gross, and B . Ben-Nissan, "Critical Ageing of Hydroxyapatite Sol-Gel Solutions", Biomaterials, 19, 2291-2296 (1998). 93. A . Jillavenkatesa and R. A . Condrate, "Sol-Gel Processing of Hydroxyapatite", J. Mater. Sci., 33,4111-4119 (1998). 140 94. H . Takahashi, M . Yashima, M . Kakihana, and M . Yoshimura, "Synthesis of Stoichiometric Hydroxyapatite by a "Gel", Route from The Aqueous Solution of Citric and Phosphoneacetic Acids", Europ. J. Solid State Inorg. Chem., 32, 829-835 (1995). 95. Q. Qiu, P. Vincent, B . Lowenberg, M . Sayer, and J. E. Davies, "Bone Growth on Sol-Gel Calcium Phosphate Thin Films in Vitro", Cells and Mater., 3[4], 351-360 (1993). 96. A . Deptula, W. Lada, T. Olczak, A . Borello, C. Alvani, and A . di Bartolomeo, "Preparation of Spherical Powders of Hydroxyapatite by Sol-Gel Process", J. Non-Crystalline Solids, 147/148, 537-541 (1992). 97. F. H . Westheimer, S. Huang, and F. Coritz, "Rates and Mechanisms of Hydrolysis of Esters of Phosphorus Acids", J . A m . Chem. Soc, 110, 181-185 (1988). 98. P. Layrolle, A . Ito, and T. Takishi, "Sol-Gel Syntheis of Amorphous Calcium Phosphate and Sintering into Microporous Hydroxyapatite Bioceramics", J. A m . Ceram. Soc, 81 [6], 1421-1428 (1998). 99. M . F. Hsieh, L . H . Perng, T. S. Chin, and H . G Perng, "Phase Purity of Sol-Gel-Derived Hydroxyapptite Ceramic", Biomaterials, 22, 2601-2607 (2001). 100. W. Weng, L . Huang, and G. Han, "The Alkoxide Sol-Gel Process in The Calcium Phosphate System and its Applications", App. Organometallic Chem., 13, 555-564 (1999). 101. D. M . Liu and T. Troczynski, "Low-Temperature Synthesis of Hydroxyapatite Coating via a Sol-Gel Route", presented at the 102 n d Annual Meeting of the American Ceram., Soc, St. Louis (2000). 102. D. M . Liu, D. Hakimi, and T. Troczynski, "Effect of Hydrolysis on The Phase Evolution of Water-Based Sol-Gel Hydroxyapatite and Its Application to Bio-Coating", Mater. Med. 13, 657-665 (2002). 103. M.F. Hsieh, L . H . Perng, T. S. Chin, and H . G. Perng, "Phase Purity of Sol-Gel-Derived Hydroxyapatite Ceramics", Biomaterials, 22, 2601-2607 (2001). 104. J. C. Schrotter, A . Cardenas, M . Smaihi, and N . Hovnanian, "Silicon and Phosphorus Alkoxide Mixture: Sol-Gel Study by Spectroscopic Technics", J. Sol-Gel Sci. Tech., 4, 195-204 (1995). 141 105. R. J. Kirkpatrick and R. K . Brow, "Nuclear Magnetic Resonance Investigation of the Structure of Phosphate and Phosphate-Containing Glass: a review", Nuclear Mag. Reson., 5, 9-21 (1995). 106. D. M . Liu , T. Troczynski, and W. J. Tseng, "Ageing Effect on the Phase Evolution of Sol-Gel Hydroxyapatite", Biomaterials, 23, 1227-1236 (2002). 107. J. Breme, "Titanium and titanium Alloys-The Biomaterials of Preference", Mem Etudes Scient Rer-Metall, 86, 625-637 (1989). 108. J. Eulenberger, T. Kolle, A . Schroeder, S. G. Steinemann, Hoftung Zwischen Knochen und Titum, 4 DVM-Vortrags-reihe des Arbeitskreises, Implantate, 4, 131-140 (1983). 109. J. Breme, Y. Zhou, and L . Groh, "Development of a Titanium Alloy Suitable for an Optimized Coating with Hydroxyapatite", Biomaterials, 3, 239-244 (1995). 110. R. Z. LeGeros, J. P. Legeros, R. Kijkowska, R. Zheng, C. Bautista, and J. L . Wong, "Calcium Phosphates in Plama-Sprayed Hydroxyapatite Coatings", pp. 173-189 in Ceramic Transaction Vol. 48, Bioceramics: Materials and Applications, edited by G Fischman, A . Clare, and L . Hench, American Ceramic Society, Westerrille, O H 1995. 111. K . A . Gross, V. Gross, and C. C. Berndt, "Thermal Analysis of Amorphous Phases in Hydroxyapatite Coatings", J. Amer. Ceram. Soc, 81[1], 106-121 (1998). 112. S. Kaciulia, G. Mattogno, L . Pandolfi, M . Caralli, G. Gnappi, and A . Montenero, "XPS Study of Apatite-Based Coatings Prepared by Sol-Gel Technique", App. Surf. Sci., 151, 1-5 (1999). 113. P. Ducheyne, S. Radin, and K . Ishikawa, "The Rate of Calcium Phosphate Precipitation on Metal and Ceramics, and the Relatiionship to Bioactivity", in Bone-Bonding Biomaterials, P. Duchenye, T. Kokubo, and C. A . van Blitterswijk (eds), reed Healthcare Communications, Leiderdrop, pp. 213-18 (1992). 114. S. Lin and A . A . Campbell, US patent #5, 958, 430, 1999. 115. J. Livage, M . Henry, and C. Sanchez, "Sol-Gel Chemistry of Transition Metal Oxides", Prog. Solid State Chem., 18, 259-341 (1988). 116. J. Livage, P. Barboax, M . T. Vandenborre, C. Sichmutz, and F. Taulelle, "Sol-Gel 142 Synthesis of Phosphates", J. Non-Cryst. Solids, 147/148, 18-23 (1992). 117. E. D. Eanes, I. H . Gillessen, and A . S. Posner, "Intermediate States in The Precipitation of Hydroxyapatite", Nature, 208, 365-367 (1965). 118. K . A . Gross, V. Gross, and C. C. Berndt, "Thermal Analysis of Amorphous Phases in Hydroxyapatite Coatings", J . A m . Ceram. Soc, 8[1], 106-112 (1998). 119. G. Daculsi, "Biphasic Calcium Phosphate Concept Applied to Artificial Bone, Implant Coating and Injectable Bone Substitute", Biomaterials, 19, 1473-1478 (1998). 120. L . A . Azaroff, p. 552 in "Elements of X-Ray Crystallography", McGraw-Hil l , New York, 1968. 121. C. J. Brinker and G. W. Scherer, Sol-Gel Science, Chapter 2, Academic Press Inc. London, 1990. 122. C. Sanchez, J. Livage, M . Henry, and F. Babonneau "Chemical Modification of Alkoxide Precursors", J. Non-Crystal. Solids, 100, 65-76 (1988). 123. A . Ger and V. R. Deitz, "Pyrophosphate Formation Upon Ignition of precipitate Basic Calcium Phosphates", J. Amer. Chem. Soc, 2961-65, June 1955. 124. J. Arends, J. Christoffersen, M . R. Christoffersen, H . Eckert, B . O. Fowler, J. C. Heughebaert, G H . Nancollas, J. P. Yesinowski, and S. J. Zawacki, " A Calcium Hydroxyapatite Precipitaed From an Aqueous Solution", J. Crystal Growth, 84, 515-32 (1987). 125. E. D. Eanes and A . S. Posner, " A Note on The Crystal Growth of Hydroxyapatite Precipitated From Aqueous Solutions", Mat. Res. Bull . , 5, 377-84 (1970). 126. E. D. Eanes and J. L . Meyer, "The Maturation of Crystalline Calcium Phosphates in Aqueous Suspensions at Physiological pH", Calcif. Tiss. Res., 23, 259-69 (1977). 127. J. L . Meyer and B . O. Fowler, "Lattice Defects in Nonstoichiometric Calcium Hydroxyapatite, A Chemical Approach", Inorg. Chem., 21, 3029-35 (1982) 128. D. Hakimi, D. M . Liu , and T. Troczynski, "Effect of Alcoholic Solvents on Microstructural Evolution and Phase Formation of Sol-Gel Hydroxyapatite", Materials in 143 Medicine (submitted) 129. C. A . Simmons, N . Valiquette, and R. M . Pilliar, "Osseointegration of Sintered Porous-Surfaced and Plasma-Sprayed-coated Implants: A n Animal Model Study of Early Postimplantation Healing Response and Mechanical Stability", J. Biomed. Mater. Res., 47[2], 127-138 (1999). 130. B . C. Wang, E. Chang, D. Tu, C. H . Tsai, and C. Y. Yang, "Characteristics and Osteoconduction of Three Different Plasma-Sprayed Hydroxyapatite-Coated Titanium Implants", Surf. Coat. Tech., 58, 107-117 (1993). 131. T. Kokubo, H . Kushitani, S. Sakka, T. Kitugi, and T. Yamanuro, "Solution able to Reproduce in-vivo Surface-Structture Changes in Bioactive Glass-Ceramic A - W " , J. Biomed. Mater. Res., 24, 721-734 (1990). 132. J. I. Huaxia, C. B . Ponton, and P. M . Marquis, "Microstructural Characterization of Hydroxyapatite Coating on Titanium", Mater. Med., 3, 283-287 (1992). 133. S. R. Radin and P. Ducheyne, "The Effect of Calcium Phosphate Ceramic Composition anad Structure on in vitro Behavior, II. Precipitation", J. Biomed. Mater. Res., 27, 35-45 (1993). 134. D. M . Liu , Q. Yang, and T. Troczynski, "Sol-Gel Hydroxyapatite Coatings on Stainless Steel Substrates", Biomaterials, 23, 691-698 (2002). 135. P. L i , C. Ohtsuki, T. Kokubo, K . Nakanishi, N . Soga, and K . de Groot, "The Role of Hydrated Silica, Titania, and Alumina in Inducing Apatite on Implants", J. Biomed. Mater. Res., 28, 7-15 (1994). 136. H . P. Boehm, "Acidic and Basic properties of Hydroxylated Metal Oxide Surface", Disc. Faraday Soc, 52, 264-275 (1971). 137. K . E. Healy and P. Ducheyne, "Hydration and Preferential Molecular Adsorption on Titanium in vitro", Biomaterials, 13, 553-61 (1992). 1138. J. W. Mullin, "Crystallization", 2 n d ed. Butterworths, London, 1972, p.189. 139. C. Combes, M . Freche, C. Rey, and B . Biscans, "Heterogeneous Crystallization of Dicalcium Phosphate Dihydrate on Titanuim Surfaces", Materials in Medicine, 10, 231-237 144 (1999). 140. L . C. Chow, S. Takagi, P. D. Constantino, and C. D. Friedman, "Self-setting Calcium Phosphate Cements", Mat. Res; Soc. Symp. Proc, 179, 3-24, 1991. 141. B. R. Constantz, I. C. Ison, M . T. Fulmer, R. D. Doster, S. T. Smith, M . van Wangoner, J. Ross, S. A . Goldstein, J. B . Jupiter, and D. I. Rosenthal, "Skeletal Repair by In Situ Formation of the Mineral Phase of Bone", Science 267, 1796 (1995). 142. F. C. Driessens, M . G Boltong, O. Bermudez, J. A . Planell, M . P. Ginebra, and E. Fernandez, "Effective Formulations for the Preparation of Calcium Phosphate Bone Cements", Materials in Medicine, 5, 164 (1994). 143. L . M . Ellerby, C. R. Nishida, F. Nishida, S. A . Yamanakas, B . Dunn, S. J. Valentine, and J. I. Zink, "Encapsulation of Proteins in Transparent Porous Silicate Glasses Prepared by the Sol-Gel Method", Science, 255, 1113 (1992). 144. D. M . Liu and I. W. Chen, "Encapsulation of Protein Molecules in Transparent Porous Silica Matrices via an Aqueous Colloidal Sol-Gel Process", Acta Materialia, 47, 4535 (1999). 145. S. Mann, B . R. Heywood, S. Rajam, and V. J. Wale, "Molecular Recognition in Biomineraliztion" in Mechanism and Phylogeny of Mineralization in Biological Systems, S. Suga and H . Nakahara (eds), Springer-Verlag, Tokyo, pp.47-55 (1991). 146. D. M . Liu , Quanzu Yang, Tom Troczynski, and, Wenjea J. Tseng, "Structural Evolution of Sol-Gel-Derived Hydroxyapatite", Biomaterials, 23, 1679-1687 (2002). 147. D. M . Liu, Quanzu Yang, and Tom Troczynski, "In-Vitro Forming of Calcium Phosphate Layer on Sol-Gel Hydroxyapatite-Coated Metal Substrates", Materials in Medicine, 13, 965-971 (2002). 148. D. M . Liu, Quanzu Yang, and Tom Troczynski, " A Fast, Cost-Effective In-Situ Drug Encapsulation in Hydroxyapatite Coatings", Biomaterials, (submitted). 149. D. M . Liu , "Bioactive Coatings- An Overview", Materials in Medicine (in review) 150. Q. Yang, D. M . Liu , and T. Troczynski, "Influence of Apatite Seeds on the Synthesis of Calcium Phosphate Cements", Biomaterials, 23, 2751-2760 (2002). 145 151. D. Hakimi, T. Troczynski, and , D. M . Liu "Effect of Aliphatic Alcohol Solvent on Structural Evolution of Sol-Gel Hydroxyapatite", Materials in Medicine (accepted for publication). 152. D. M . L iu and T. Troczynski, "Low-Temperature Synthesis of Hydroxyapatite Coating via a Sol-Gel Route" paper presented at the 102 n d Annual Meeting of the American Ceram., Soc, (2000). 153. D. M . Liu , Q. Yang, and T. Troczynski, "Room-temperature synthesis of Calcium Apatite Microspheres fro Drug Delivery Applications", presented at the 102nd Annual Meeting of the American Ceramics Society (2001). 154. D. M . Liu, Q. Yang, and T. Troczynski, "Low-Temperature Sol-Gel Hydroxyapatite Coating: Interfacial Study", presented at the 102 n d Annual Meeting of the American Ceramics Society (2001). 155. Q. Yang, D. M . Liu , and T. Troczynski, "Development of Fast-Setting Calcium Phosphate Cements: Seeding Effect", presented at the 102 n d Annual Meeting of the American Ceramics Society (2001). 156. Q. Yang, D. M . Liu , and T. Troczynski, "Development of Novel Fast Self-Setting Calcium Phosphate Cements for Drug Delivery", presented at the San Diego Symposium on Controlled Release Conference, June 2001. 157. D. M . Liu, Quanzu Yang, Tom Troczynski, and Thomas Oxland, "Sol-Gel Hydroxyapatite Coatings: Adhesion and Microstructure Development", presented at the 102 n d Annual Meeting of the American Ceramics Society (2001). 158. Q. Yang, D. M . Liu, and T. Troczynski, "Development of a Fast-Setting Calciium Phosphate Cement", presented at the Annual Meeting of the Canadian Ceramic Society, Sep. 23-25 (2001). 159. T. Troczynski, Q. Yang, D. M . Liu, and D. Hakimi, "Novel Processing Routes and Applications for Hydroxyapatite Bioactive Ceramics", presented at the Annual Meeting of the Canadian Ceramic Society, Sep. 23-25 (2001). 146 Thesis-Related Publications Peer-review Journals 1. D. M . Liu, T. Troczynski, and W. J. Tseng, "Water-Based Sol-Gel Hydroxyapatite: Process Development", Biomaterials, 22, 1721-1730, (2001). 2. D. M . Liu, D. Hakeimi, and T. Troczynski, "Effect of Hydrolysis on the Phase Evolution of Water-Based Sol-Gel Hydroxyapatite and Its Application to Bio-Coating", J. Mater. Sci. Mater. Medicine, 13, 657-665 (2002). 3. D. M . Liu, T. Troczynski, and W. J. Tseng, "Ageing Effect on the Phase Evolution and Microstructural Development of Sol-Gel Hydroxyapatite", Biomaterials, 23, 1227-1236, (2002). 4. D. M . Liu , Quanzu Yang, Tom Troczynski, and, Wenjea J. Tseng, "Structural Evolution of Sol-Gel-Derived Hydroxyapatite", Biomaterials, 23, 1679-1687 (2002). 5. D. M . Liu , Quanzu Yang, and Tom Troczynski, "Sol-Gel Hydroxyapatite Coating onto Stainless Steel Substrates", Biomaterials, 23, 691-698 (2002). 6. D. M . Liu, Quanzu Yang, and Tom Troczynski, "In-Vitro Forming of Calcium Phosphate Layer on Sol-Gel Hydroxyapatite-Coated Metal Substrates", Materials in Medicine, 13, 965-971 (2002). 7. D. M . Liu , Quanzu Yang, and Tom Troczynski, " A Fast, Cost-Effective In-Situ Drug Encapsulation in Hydroxyapatite Coatings", Biomaterials, (submitted). 8. D. M . Liu, "Bioactive Coatings- An Overview", Materials in Medicine (in review) 9. Q. Yang, D. M . Liu , and T. Troczynski, "Influence of Apatite Seeds on the Synthesis of Calcium Phosphate Cements", Biomaterials, 23, 2751-2760 (2002). 10. D. Hakimi, T. Troczynski, and , D. M . Liu "Effect of Aliphatic Alcohol Solvent on Structural Evolution of Sol-Gel Hydroxyapatite", Materials in Medicine (accepted for publication). Conferences papers 147 1. D. M . L iu and T. Troczynski, "Low-Temperature Synthesis of Hydroxyapatite Coating via a Sol-Gel Route" paper presented at the 102 n d Annual Meeting of the American Ceram., Soc, (2000). 2. D. M . Liu , Q. Yang, and T. Troczynski, "Room-temperature synthesis of Calciium Apatite Microspheres fro Drug Delivery Applications", presented at the 102nd Annual Meeting of the American Ceramics Society (2001). 3. D. M . Liu, Q. Yang, and T. Troczynski, "Low-Temperature Sol-Gel Hydroxyapatite Coating: Interfacial Study", presented at the 102 n d Annual Meeting of the American Ceramics Society (2001). 4. Q. Yang, D. M . Liu , and T. Troczynski, "Development of Fast-Setting Calcium Phosphate Cements: Seeding Effect", presented at the 102 n d Annual Meeting of the American Ceramics Society (2001). 5. Q. Yang, D. M . Liu , and T. Troczynski, "Development of Novel Fast Self-Setting Calcium Phosphate Cements for Drug Delivery", presented at the San Diego Symposium on Controlled Release Conference, June 2001. 6. D. M . L iu , Quanzu Yang, Tom Troczynski, and Thomas Oxland, "Sol-Gel Hydroxyapatite Coatings: Adhesion and Microstructure Development", presented at the 102 n d Annual Meeting of the American Ceramics Society (2001). 7. Q. Yang, D. M . Liu, and T. Troczynski, "Development of a Fast-Setting Calciium Phosphate Cement", presented at the Annual Meeting of the Canadian Ceramic Society, Sep. 23-25 (2001). 8. T. Troczynski, Q. Yang, D. M . Liu, and D. Hakimi, "Novel Processing Routes and Applications for Hydroxyapatite Bioactive Ceramics", presented at the Annual Meeting of the Canadian Ceramic Society, Sep. 23-25 (2001). Patents US Patent application 2002155144 "Biofunctional hydroxyapatite coatings and microspheres for in-situ drug encapsulation" US Patent 6426114 B l "Sol-Gel calcium phosphate ceramic coatings and method of making same" 148 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0078658/manifest

Comment

Related Items