Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Processing and characterization of hydroxyapatite-based bioceramic pastes Chae, Taesik 2007

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

Item Metadata

Download

Media
831-ubc_2007-0046a.pdf [ 11.25MB ]
Metadata
JSON: 831-1.0078477.json
JSON-LD: 831-1.0078477-ld.json
RDF/XML (Pretty): 831-1.0078477-rdf.xml
RDF/JSON: 831-1.0078477-rdf.json
Turtle: 831-1.0078477-turtle.txt
N-Triples: 831-1.0078477-rdf-ntriples.txt
Original Record: 831-1.0078477-source.json
Full Text
831-1.0078477-fulltext.txt
Citation
831-1.0078477.ris

Full Text

PROCESSING AND CHARACTERIZATION OF HYDROXYAPATITE-BASED BIOCERAMIC PASTES by TAESIK CHAE B.Sc., Department of Materials Engineering, Chungnam National University, Daejeon, Korea, 2004 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Materials Engineering) THE UNIVERSITY OF BRITISH COLUMBIA April 2007 © Taesik Chae, 2007 ABSTRACT Repair of skeletal defects and filling fracture gaps requires bioactive, biocompatible, biodegradable and easy to handle bone grafts. Although synthetic hydroxyapatite (HAp) based pastes are very good candidates for such bone grafting applications (as they meet most of these requirements), there is a need for improved handling of such pastes. In the present research, a processing technology was developed for improving the injectability of HAp paste through surface modifications of HAp with surfactants. A novel syringe-based system for simple, practical assessment of paste's flow and injectability was developed to correlate viscosity of the modified paste with its flowability. HAp-based bioceramic pastes were prepared by combining HAp powder and surfactants in an intensive planetary ball milling in three kinds of liquid media: 1) distilled water, 2) binary solution of water and ethylene glycol (EG) and 3) poly(dimethyl siloxane) (PDMS). In order to observe the effect of tri-sodium citrate (TSC) on flowability of the pastes, it was dissolved in each of the water-containing liquids. The pastes were carefully loaded into a novel syringe-based practical viscometer for injection tests using Instron Universal Testing System. It was found that TSC was very effective in reducing viscosity of the HAp pastes. EG was also found to be helpful for homogenization of the pastes. It is hypothesized that the citrate ions from TSC induced negative surface charges on HAp particles, which resulted in homogeneously dispersed HAp in the pastes. The hydrocarbon steric layer on HAp particles produced by EG molecules also induced less agglomeration of the particles in the pastes. These two surfactants helped to achieve good flowability of the water-based HAp pastes. However, high viscosity and hydrophobicity of PDMS caused poor mixing and increased viscosity of the pastes. It was also found that the custom-developed syringe-based practical viscometer can be conveniently used to quickly assess flowability of pastes of very high viscosity. i i TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iii LIST OF TABLES vii LIST OF FIGURES ix ACKNOWLEDGMENTS xiii 1 INTRODUCTION 1 2 LITERATURE REVIEW 5 2.1 Bioceramics 5 2.1.1 Hydroxy apatite 9 2.1.2 Hydroxyapatite/Polymer Composites 10 2.2 Bioceramic Pastes for Medical Applications 12 2.2.1 Calcium Phosphate Cement ....13 2.2.2 Hydroxyapatite Pastes 14 2.2.3 Calcium Hydroxide Pastes 15 2.3 Rheology of Ceramic Pastes , 15 2.3.1 Viscosity of Ceramic Pastes 16 2.3.2 Pastes Injectability Measurement 20 2.3.3 Factors Influencing Rheological Properties of Ceramic Pastes 22 i i i 2.4 Additives in Bioceramic Pastes 25 2.4.1 Tri-sodium Citrate Dispersant 26 2.4.2 Ethylene Glycol 28 2.4.3 Poly(dimethyl siloxane) 29 3 SCOPE AND OBJECTIVES 31 3.1 Scope of the Investigationo 31 3.2 Objectives 32 4 EXPERIMENTAL METHODOLOGY 34 4.1 Materials and Paste Preparation 34 4.2 Injectability Evaluation 38 4.2.1 Syringe Viscometer Calibration 38 4.2.2 Injectability Tests 41 4.3 Characterization of HAp-based Bioceramic Pastes 42 4.3.1 Microstructural Characterization 42 4.3.2 Particle Size and Size Distribution 42 4.3.3 Zeta Potential Measurement 42 4.4 Optimization of EG-TSC-HAp Pastes 43 5 RESULTS AND DISCUSSION 44 5.1. Syringe Viscometer Calibration 44 iv 5.2. Injectability Evaluation and Characterization of EG-HAp Pastes 46 5.2.1 Effects of Powder-to-Liquid (P/L) Ratio on Injectability of EG-HAp Pastes....46 5.2.2 Effects of EG on Injectability of EG-HAp Pastes 54 5.2.3 Microstructural Characterization of EG-HAp Pastes 59 5.2.4 Particle Size Distribution in EG-HAp Pastes 60 5.2.5 Zeta Potential Measurement for EG-HAp Pastes 62 5.3 Injectability Evaluation and Characterization of TSC-HAp Pastes 63 5.3.1 Effects of P/L Ratio on Injectability of TSC-HAp Pastes 63 5.3.2 Effects of TSC on Injectability of TSC-HAp Pastes 69 5.3.3 Microstructural Characterization of TSC-HAp Pastes 71 5.3.4 Particle Size Distribution in TSC-HAp Pastes 73 5.3.5 Zeta Potential Measurements for TSC-HAp Pastes 74 5.4 Injectability Evaluation and Optimization of EG-TSC-HAp Pastes 75 5.4.1 Experimental Design of EG-TSC-HAp Pastes 75 5.4.2 Injectability Evaluation and Characterization of EG-TSC-HAp Pastes 77 5.4.3 Optimization of EG-TSC-HAp Paste - Experimental Verification 85 5.5 Injectability Evaluation of PDMS-HAp Pastes 87 6 CONCLUSIONS 91 7 RECOMMENDATIONS FOR FUTURE WORK. 94 v APPENDIX - DETAILS OF RESPONSE SURFACE OPTIMIZATION METHOD LIST OF TABLES Table 2.1-1. Desired properties of implantable bioceramics 5 Table 2.1-2. Mechanical properties of bioceramics and human bone 6 Table 2.1-3. Acronyms, formulas, Ca/P ratios and solubility products of representative calcium phosphates 8 Table 2.3-1. Common deflocculants used in polar liquids 23 Table 4.4-1. Compositions of the EG-HAp, TSC-HAp and PDMS-HAp pastes studied 37 Table 5.1-1. Viscometer calibration results using Brookfield Standard Viscosity Fluids of 5.00 and 12.5 Pa.s 44 Table 5.2-1. P/L ratio range and injectability test results of EG-HAp pastes 46 Table 5.3-1. P/L ratio range and injectability test results of TSC-HAp pastes 63 Table 5.4-1. The experimental design actual factor levels for EG-TSC-HAp pastes 76 Table 5.4-2. The experimental design layout and results in actual values 77 Table 5.4-3. Median diameter of HAp particles and volume fraction of D < 1.0 and D > 4.0 in 1.2 g/ml EG-TSC-HAp pastes 85 Table 5.4-4. Optimziation criteria settings for EG-TSC-HAp paste composition for 1.30 g/ml and 9.00 Pa.s 86 Table 5.4-5. A DX suggested composition of the optimized EG-TSC-HAp paste with P/L=\.3 g/ml and the predicted viscosity of 9.00 Pa.s; experimental verification of the paste viscosity is included 87 Table 5.5-1. Injectability tests results for PDMS-HAp pastes 88 Table A l . "Sequential Model Sum of Squares" summary 95 Table A2. "Lack-of-fit" tests summary 96 Table A3. "Model Summary Statistics" summary 96 vii Table A4. ANOVA for the linear model adopted in this work 97 Table A5. Coefficients for the linear model adopted for process optimization in this study...97 LIST OF FIGURES Figure 2.3-1. Shear stress vs. shear rate curves of Newtonian and non-Newtonian liquids; (a) Bingham, (b) shear-thining (pseudo-plastic), (c) shear-thickening (dilatant) and (d) Newtonian models 17 Figure 2.3-2. Schematic diagram of basic tool geometries for the rotational viscometers (a) concentric cylinder, (b) con and plate and (c) parallel plate ;..19 Figure 2.3-3. Schemical diagram of capillary viscometer 19 Figure 2.3-4. Potential energy of interation between two particles with electrical double layers 25 Figure 2.4-1. Distribution of citrate species (cit3~, Hcit2", I-I^ cit", and F^cit) at 25°C as a function of pH 27 Figure 2.4-2. Molecular structure of (a) citric acid, (b) tri-sodium citrate and (c) tri-potassium citrate 28 Figure 2.4-3. Molecular structure of poly(dimethyl siloxane) 30 Figure 4.1-1. XRD spectrum of HAp powder from Sigma Aldrich 34 Figure 4.1-2. SEM micrograph of HAp powder 35 Figure 4.1-3. Particle size and size distribution of HAp powder 35 Figure 4.1-4. (a) HAp powder in an alumina jar, (b) Pulverisette® 5 Planetary Mill, (c) HAp paste after milling and (d) HAp paste in a Diapex syringe 38 Figure 4.2-1. Schematic illustration of (a) a general capillary viscometer used in previous research and (b) syringe viscometer evaluated in this work 39 Figure 4.2-2. Syringe viscometer installed on Instron 3360 series Universal Testing System... 41 ix Figure 5.1-1. Plunger resistance (a) and injection force profile for Brookfield Standard Viscosity Fluids of (b) 5.00 Pa.s and (c) 12.5 Pa.s viscosity 45 Figure 5.2-1. Injection force profiles for E G (0 vol%)-HAp pastes 47 Figure 5.2-2. Injection force profiles for E G (20 vol%)-HAp pastes 48 Figure 5.2-3. Injection force profiles for E G (40 vol%)-HAp pastes 49 Figure 5.2-4. Injection force profiles for E G (60 vol%)-HAp pastes 50 Figure 5.2-5. Injection force profiles for E G (80 vol%)-HAp pastes 51 Figure 5.2-6. Injection force profiles for E G (100 vol%)-HAp pastes 52 Figure 5.2-7. Apparent viscosity change vs. P/L ratio of EG-HAp pastes for a constant E G concentration 54 Figure 5.2-8. Injection force profiles for (a) 0.8 g/ml EG-HAp pastes and (b) 0.9 g/ml EG-HAp pastes with different E G concentrations 55 Figure 5.2-9. Injection profiles for (a) 1.0 g/ml EG-HAp pastes with different E G concentrations 56 Figure 5.2-10. Injection profiles for (a) 1.1 g/ml EG-HAp pastes and (b) 1.2 g/ml EG-HAp pastes with different E G concentrations 57 Figure 5.2-11. Apparent viscosity change vs. E G vol% in EG-HAp pastes for a constant P/L 58 Figure 5.2-12. S E M micrographs of 1.2 g/ml EG-HAp pastes, for (a) 60 vol%, (b) 80 vol% and (c) 100 vol% E G content 60 Figure 5.2-13. Particle size distribution analysis for 1.2 g/ml EG-HAp pastes 61 Figure 5.2-14. Zeta potential of HAp particles in diluted 1.2 g/ml EG-HAp pastes in the pH range of 4.0 to 11 62 Figure 5.3-1. Injection force profiles for TSC (0.0 wt%)-HAp pastes 64 x Figure 5.3-2. Injection force profiles for TSC (0.2 wt%)-HAp pastes 65 Figure 5.3-3. Injection force profiles for TSC (0.4 wt%)-HAp pastes 66 Figure 5.3-4. Injection force profiles for TSC (0.6 wt%)-HAp pastes 67 Figure 5.3-5. Apparent viscosity change vs. P/L ratio of TSC-HAp pastes for a constant TSC concentration 68 Figure 5.3-6. Injection force profiles for TSC-HAp pastes with different TSC concentrations 70 Figure 5.3-7. Apparent viscosity vs. TSC content (wt%) in TSC-HAp pastes for a constant P/L 71 Figure 5.3-8. SEM micrographs of 1.2 g/ml TSC-HAp pastes (a) 0.4 wt% and (b) 0.6 wt% 72 Figure 5.3-9. Particle size and size distribution analysis of 1.2 g/ml TSC-HAp pastes 73 Figure 5.3-9. Zeta potential of HAp particles in 1.2 g/ml TSC-HAp pastes in the pH range of 4.0 and 11 74 Figure 5.4-1. Experimental factorial design layout represented in coded values 76 Figure 5.4-2. The effect of TSC and EG admixture on apparent viscosity of EG-TSC-HAp pastes at: (a) P/L = 1.06, (b) P/L = 1.15 and (c) P/L = 1.24 78 Figure 5.4-3. The effect of P/L and TSC on apparent viscosity of EG-TSC-HAp pastes at: (a) EG = 48.11, (b) EG = 60.00 and (c) EG = 71.89 79 Figure 5.4-4. The effect of P/L and EG on apparent viscosity of EG-TSC-HAp Pastes at: (a) TSC = 0.44, (b) TSC = 0.50 and (c) TSC = 0.56 80 xi Figure 5.4-5. SEM micrographs of 1.24 g/ml EG-TSC-HAp Pastes; (a) EG(48.11 vol%)-TSC(0.44 wt%), (b) EG(71.89 vol%)-TSC(0.44 wt%), (c) EG(48.11 vol%)-TSC(0.56 wt%) and (d) EG(71.89 vol%)-TSC(0.56 vol%) 82 Figure 5.4-6. Zeta potential of 1.2 g/ml EG-TSC-HAp Pastes; (a) EG(48.11 vol%)-TSC(0.44 wt%), (b) EG(71.89 vol%)-TSC(0.44 wt%), (c) EG(48.11 vol%)-TSC(0.56 wt%) and (d) EG(71.89 vol%)-TSC(0.56 vol%) 83 Figure 5.4-7. Particle size distribution for 1.2 g/ml EG-TSC-HAp Pastes; (a) EG(48.11vol%) TSC(0.44 wt%), (b) EG(71.89 vol%)-TSC(0.44 wt%), (c) EG(48.11 vol%)-TSC (0.56 wt%) and (d) EG(71.89 vol%)-TSC(0.56 vol%) 84 Figure 5.4-8. Injection force profile of the optimized composition EG-TSC-HAp paste, at P/L= 1.30 g/ml, 0.45 wt% TSC and 44.3 vol% EG 87 Figure 5.5-1. Injection force profiles for PDMS-HAp pastes 88 Figure 5.5-2. Apparent viscosity change of PDMS-HAp pastes as a function of P/L ratio 89 xii ACKNOWLEDGMENTS I would like first to express my sincere gratitude to my supervisor, Dr. Tom Troczynski for his comprehensive scientific guidance and advice during the graduate program. I wish to extend my appreciation to Dr. Quanzu Yang for his valuable suggestions for this project. I also owe thanks to other graduate students, faculty and staffs in the Department of Materials Engineering for all their help. The last but not least, I would like to express my gratitude to the Korea Science and Engineering Foundation for providing the financial support (Grant No. M06-2004-000-10286). xiii 1 INTRODUCTION The demand for tissue treatments to address trauma or disease has increased with the extended life-span of humans. Autograft supply is limited and requires additional surgeries, while allografts have the disadvantages of possible disease transmission from the donor and anti-immunization response after implant" . Therefore, intensive research into synthetic materials with sufficient biological and mechanical properties comparable to natural tissues has been undertaken. Biomaterials are such synthetic materials designed to replace partially or fully living tissues. For the materials point of view, they are generally categorized into metallic, polymeric, ceramic and composite biomaterials7. Bioceramics have received particular attention in hard tissue replacement due to their positive biological response22"27. Bioinert AI2O3 and ZrO*2 with superior mechanical properties (compressive strength up to ~4GPa, fracture toughness of about 5 MPaVm for alumina, up to 8 MPaVm for zirconia) have been utilized for load bearing applications such as hip/knee joint replacements. Bioactive glasses, although significantly more brittle (K c < 1 MPaVm) stimulate a specific biological reaction at the interface with hard tissues, resulting in a strong interface bonding. Calcium phosphates, especially hydroxyapatite (HAp, Caio(PO"4)6(OH)2), are primary candidates for hard tissue reconstruction as well as regeneration. Having similar chemical composition to the mineral component of natural bone, HAp is not only bioactive and bioresorbable, but also more biocompatible than any other ceramic. The osteoconductivity of HAp allows ingrowth of new autogenous bone, making it the most feasible bone substitute4'24'25. Different forms of HAp such as blocks, coatings, granules and injectable pastes have been used in orthopedic and dental surgeries5"8. Among them, injectable pastes have been 1 considered to possess great potential as bone or dental defects filler due to the following characteristics: 1. Damaged tissues with complex geometry can be filled with self-flowable pastes. 2. No massive surgical operation is required, allowing a minimally invasive procedure 3. Drugs or genes for the prevention of bacterial infection and faster healing process can be delivered together with the paste. 4. It was shown that use of pastes can reduce time and cost for treatments10. Previously studied group of bioceramic pastes are calcium phosphate cements (CPC), which transform into HAp (and/or other calcium phosphates) through a dissolution-precipitation mechanism during setting in-vivo9. Recent studies show that injection of CPC has been adopted for treating calcaneal bone cyst and pathological fracture cases", tibial plateau fracture12 and bone defects around oral implants13. However, injectable HAp pastes are attracting more attention because they do not need the setting conversion reaction into HAp, and can be directly injected into defects from a pre-mixed ready-to-use syringe. Furthermore, the pre-mixed paste system provides clinicians with an easy-to-handle procedure for paste injection, and there is no need to purchase paste mixing/injection tools. Recently, Ostim (Heraeous Kulzer, Germany) introduced the first commercial HAp paste being resorbed and replaced by autogenous bone15. Researchers from the FAP Dental Institute in Japan have developed a modified HAp paste which can repair tooth decay in enamel16. For injectable bioceramic pastes, rheological behavior is the key characteristic that must be taken into account for maximizing its clinical efficiency. For example, in order to prevent wash-out by body fluids penetration, pastes should contain as much solid powder as 2 possible to create volumetric stability, but high ratio of powder to liquid deteriorates the injectability of pastes. Furthermore, it causes "filter pressing effect" which leads to phase separation between solid and liquid. Bad flowability of pastes causes poor injection leading to incomplete filling of the complex edges of reconstruction sites. Therefore, optimum flowability and consistency of paste must be maintained and viscosity changes must be carefully observed over time due to the self-hardening effect of the pre-mixed and syringe contained pastes17"19. Different methods have been used for assessing flowability of injectable pastes. Kharoun et al.1 7 used a syringe to measure injectability simply by the wt % of a paste extruded from the syringe. Leroux et al. designed a manometer attached syringe and compared injectability of calcium phosphate cements by indicated pressures. However, these methods could not perform quantificational precise process due to the inner diameter of the syringe needles being far too wide, thus sensitivity for detecting viscosity changes was too low and the tests were not conclusive. The apparatus designed by Baroud et al.21 was a capillary rheometer barrel-type which can calculate viscosity from measured injection force, but the inability to resume the experiment after one test is completed due to cleaning and drying of the apparatus obstructs the procedure. In the present study, a custom-developed syringe-based viscometer was designed for assessing flowability of injectable pastes, and a processing technology was developed for improving injectability of HAp-based bioceramic pastes through surface modification of HAp. The pastes were prepared by mixing HAp powder, distilled water and surfactants (ethylene glycol and tri-sodium citrates) in a planetary ball mill. Poly(dimethyl siloxane) was used as 3 another liquid medium for comparison. The effect of each surfactant on injectability was characterized by the syringe viscometer. The pastes were further characterized with a zeta potential analyzer, particle size and distribution analyzer and scanning electron microscope. 4 2 LITERATURE REVIEW 2.1 Bioceramics T h e w o r d " c e r a m i c s " comes f r o m the Greek w o r d " K e r a m o s " m e a n i n g "Pottery," or "Potter 's C l a y . " C e r a m i c s can be defined as products made f r o m inorganic materials h a v i n g non-metal l ic propert ies 4 4 . C e r a m i c s have relat ively h i g h elastic m o d u l u s , m e l t i n g temperature and c h e m i c a l resistance compared to metals and p o l y m e r s due to the prevalen i o n i c and covalent b o n d i n g . O n the other hand, l o w fracture toughness is the unavoidable disadvantage o f ceramics. T h e imperfect ions such as micro-pores or cracks create stress concentration during l o a d i n g leading to b r e a k - d o w n o f ceramic b o d y rather than undergoing plastic d e f o r m a t i o n 2 2 . B i o c e r a m i c s are general ly defined as ceramic materials used for the repair, reconstruction and replacement o f diseased or damaged parts o f hard tissues. B i o c e r a m i c s must meet the property requirements i n Table 2.1-1 to be used as implants . N o biomater ia l implants i n contact w i t h l i v i n g tissues can be complete ly inert. T h e natural reaction o f the fibrous tissue encapsulates and isolates the implants f r o m the host to a v o i d t o x i c material release. R e g a r d i n g the degree o f such implant-tissue interaction, b i o c e r a m i c s are broadly Table 2.1-1. D e s i r e d properties o f implantable b ioceramics . 1. N o n t o x i c 2. N o n c a r c i n o g e n i c 3. N o n a l e r g i c 4. noninf lammatory 5. b i o c o m p a t i b l e 6. biofunct ional for its l i fet ime i n the host 5 categorized into three groups; nearly bioinert, bioactive, and bioresorbable23. Nearly bioinert ceramics undergo little or no chemical change during long-term exposure to the physiological environment and develop a few urn thick fibrous layers at the interface22. The application of biomaterials in hard tissue repair started with bioinert ceramics. These materials are applied mostly in permanent implants. High-density and purity (> 99.7%) alumina has been used in load-bearing hip and knee joint replacement and dental implant because of its combination of excellent corrosion resistance, good biocompatibility, high wear resistance and high compressive strength (Table 2.1-2). Zirconia is also exceptionally inert in the physiological environment and has advantages over alumina of higher fracture toughness and flextural strength and lower Young's modulus. Tetragonal zirconia polycrystals (TZP) stabilized with yittria, and magnesium oxide partially stabilized zirconia (PSZ) have been suggested for surgical implants for load-bearing applications such as a femoral head for total hip replacement22'23. The most recent development in this field is zirconia film coated zirconium alloy hip implant introduced by Smith and Nephew86. Table 2.1-2. Mechanical properties of bioceramics and human bone. Elastic modulus (GPa) Compressive strength (MPa) Tensile strength (MPa) Fracture toughness (MPam 1 / 2 ) Ref. Alumina (>99.7) 400 4250 5 23 Zirconia TZP 150 2000 7 23 Mg-PSZ 208 1850 8 23 Hydroxyapatite Dense 40-117 <400 <50 ~ 1.0 24,77 Porous(82-86%) 0.83-1.6xl0"3 0.21-0.41 25 Bioglass® 35 -500 42 0.5-1 26 AAV glass- 118 1080 215 2.0 27 ceramics Cortical bone 12-18 130-180 50-151 6.8 28 Cancellous bone 0.1-0.5 4-12 28,29 6 Bioactive ceramics stimulate a specific biological reaction at the interface, which results in the formation of bond between the implant and the soft or hard tissues22. With such chemically enhanced interfacial bonding, fracture may occur either in the biomaterial or the bone rather than the interface. This interface has been referred to as the 'bonding zone' consisting of mineralized organic meshwork24. Bioglass® composed of Si02-CaO-Na20-P205 was the first bioactive glass discovered by L. L. Hench and his co-workers in 196926. Later, Cervital® glass-ceramics and Apatite/Wollastonite(A/W) glass-ceramics were developed to increase long-term stability in the physiological conditions and fracture toughness respectively27. Bioactive glasses have been successfully used in the replacement of osscicles in the middle ear to heal conductive hearing loss, and endosseous ridge maintenance implant to keep the thickness and width of the jawbone after extracting teeth. In the past, reconstruction of the vertebral column extensively damaged by tumors and trauma was carried out with autograft or allograft using the combination of metal, poly(methylmethacrylate) bone cement or alumnia. However, the autograft and allograft had certain limits such as the quantity and long-term durability. The results were not always satisfactory due to loosening and dislocation after surgery27. In order to make a strong bone bonding vertebral implant, A/W glass-ceramic was used first in 1983. Its success led to other load bearing areas such as intercalary replacement of a segment of the long bone and reconstruction of iliac crests . Hydroxyapatite is also not only a representative bioactive but also a bioresorble ceramic depending on the crystallinity, porosity, Ca/P ratio and etc. HAp will be further discussed in Section 2.1.1. After implantation into a body, a bioresorbable ceramic degrades chemically by body fluids and macrophages. The resorbed area is replaced by surrounding endogenous tissues. 7 Theses characteristics make bioresorbable ceramics a prime candidate for bone filler, tissue engineering scaffolds hybridized with biodegradable polymers and drug delivery applications ' ' . Calcium phosphates are resorbable and the degree of degradation generally increases as Ca/P ratio decreases (Table 2.1-3). Tricalcium phosphate (TCP) has higher biodegradation rate than HAp, which is more favorable regarding ingrowth of new bone and releasing drugs incorporated into bone cement31. It was reported that calcium sulfate is one of the fastest resorbing ceramics. Rauschmann. et al.3 2 investigated the biphasic material of nanocrystalline HAp and calcium sulfate for local delivery of antibiotics in bone infections. Nano-HAp was incorporated to enhance biocompatibility and reduce biodegradation rate of calcium sulfate. Excessive resorption rate shortens drug release period. Moreover, the fast rate of resorption deteriorates stability and strength of the implant before new-grown tissue could take over the load. For successful usage of bioresorbable ceramics, the byproducts released from the degradation must be non toxic and can be easily disposed or extracted without damage to cells and tissues. Table 2.1-3. Acronyms, formulas, Ca/P ratios and solubility products of representative calcium phosphates. Acronym Formula Ca/P K s p Tetracalcium phosphate TTCP Ca 4 P 2 0 9 2.0 -Hydroxyapatite HAp Ca10(PO4)6(OH)2 1.67 2.34 x 10"59 Amorphous calcium phosphate ACP Ca10.xH2x(PO4)6(OH)2 - -Tricalcium phosphate(a,P,y) TCP Ca 3(P0 4) 2 1.50 2.83 x 10"30 Octacalcium phosphate OCP Ca 8H 2(P0 4) 6- 5H 20 1.33 2.00 x 10"49 Dicalcium phosphate dihydrate DCPD CaHP0 4-2H 20 1.0 2.32 x 10"7 8 2.1.1 Hydro xyapatite Human bone and teeth contain calcium and phosphorus mostly in the form of crystalline carbonated hydroxyapatite and fluorapatite which are members of family of 33 minerals named Apatites . Hydroxyapatite (HAp) ceramic has the chemical formula Caio(P04)6(OH)2. After Levitt et al.3 4 suggested a method of preparing a ceramic apatite from mineral fluorapatite in 1969, there have been many effort to synthesize and commercialize HAp. Nowadays, HAp powder is synthesized in four different ways; (1) precipitation, (2) hydrolysis, (3) solid-state reaction and (4) hydro thermal reaction33. Details of the crystal structure of both the biomineral and the synthetic HAp were determined by Beevers et al.3 5 and Kay et al.3 6, respectively. HAp belongs to the hexagonal group with lattice constants a=0.942 nm and c=0.688 nm. The atomic ratio of Ca/P is 1.67 and the density is 3.163 g/cm3. HAp can be classified both as a bioactive and bioresorbable ceramic. Inherited by analogous chemical composition to bone mineral, synthetic HAp is more biocompatible than any metal, polymer and even other ceramic. With such bio-chemical properties, its osteoconductivity makes it the most feasible bone substitute. It allows ingrowth of new bone by serving as a substrate. It promotes the adhesion of matrix-producing cells and organic molecules, forming a 3 to 5 um thick amorphous zone with high concentration of phosphate and calcium ions. This is the area where bone mineral crystals form and as this region matures, the bonding zone between the HAp implant and natural bone tissue shrinks to 50-200 nm thickness providing the high interfacial strength24. The crystallinity, porosity, and composition of bone adjust to the biological and biomechanical environment. This makes the mechanical properties of bone vary from location 9 to location. The mechanical properties of HAp ceramics can be also varied by altering the sintering temperature, crystallinity, grain size, porosity and etc. In general, the mechanical properties decrease significantly with increasing amount of mircoporosity and grain size. High crystallinity, low porosity and small grain size tend to provide higher elastic modulus, strength and toughness of the ceramic. It has been reported that the flextural strength and fracture toughness of dense HAp are much lower in dry conditions than wet37. Comparing the properties of synthetic HAp ceramics with human bone (Table 2.1-2), dense HAp has a reasonably good compressive strength but significantly lower fracture toughness than cortical bone, and the overall properties of porous HAp are clearly poorer than cancellous bone. Hence, HAp ceramics can not be used as a stand alone product for load bearing applications despite its good biological and chemical properties. HAp ceramics have been used in orthopedics and dentistry over two decades in the forms of powder, coatings, dense or porous solid blocks, and injectable pastes. The application areas are usually non-load bearing due to the lack of sufficient strength and toughness22. Examples include tooth root replacement, augmentation of alveolar ridge, maxillofacial reconstructions, coating for dental and orthopedic metal implants, and pulp capping. 2.1.2 Hydroxy apatite/Polymer Composites The difference in elastic modulus between natural bone and implanted metallic devices creates a stress shielding effect, i.e. most of the load is carried by the implant. This tends to cause bone-remodeling function less active, and the surrounding tissues resorb 1 0 thereby resulting in loosening of the devices at the interface. Polymers are ductile and less stiff and ceramics are not only more brittle but also stiffer than human hard tissues, so they are not suitable for load-bearing sites. Biomineralized tissues such as bones, teeth and shells are ideal structural composites created by biological and physio-chemical processes wherein inorganic reinforcing phase grows within/on the organic matrix, providing both strength and toughness28. The synergistic properties of two or more types of materials can be combined to suit mechanical and physiological demands of the host tissues. HAp is the commonly used ceramic reinforcement within the polymer matrix with the aim of increasing mechanical properties as well as bioactivity and osteoconductivity of the composite. Bonefield et al.3 9 proposed to make a high density polyethylene/HAp composite by injection molding and commercialized it for middle ear implants under the name of HAPEX®. Abu Baker et al.4 0 used thermal sprayed HAp particles to reinforce poly(etheretherketon)(PEEK) for orthopedic implants and achieved stiffness range of 2.8-16.0 GPa and tensile strength of 45.5-69 MPa, the lower bound of mechanical properties of the human bone (7-30 GPa and 50-150 MPa, respectively). Ever since biodegradable aliphatic polyesters such as poly(lactic acid)(PLA), poly(glycolic acid)(PGA) and poly(lactic-co-glycolic acid)(PLGA) were approved for human clinical uses, active research continued on biodegradable composites with bioactive ceramics. Those biodegradable materials avoid the need for surgical operation for their removal, and can reduce the pain of the patients and the total cost of treatments41,42. More recently, this concept of biodegradable composites has been introduced in manufacturing of macroporous scaffolds with interconnected pores (diameter of pores > 150 um) for tissue engineering. The construct is then implanted into the desired site in the patient's body and degraded over time 1 1 by de-esterification and physiological factors. The process continues as the osteoblast cells proliferate, and vascularization occurs through the macroporous structure for new autogeneous bone formation. In addition, developing the scaffolds as a mode for drug delivery function has been studied to locally release growth factors and antibiotics for enhancing bone ingrowth, treating bone defects and even supporting wound healing7'43. Hybridization of HAp with polymer has also been applied into injectable systems. Poly(methyl methacrylate)(PMMA) bone cement is currently the most widely used polymer in injectable pastes, but its lack of bioactivity is the major drawback. To solve such problem, Moursi et al:46 added 20 wt% HAp into a PMMA cement. After 8 days of osteoblast cell culture, proliferation on the composite was significantly higher than on pure PMMA cement. Fujishiro et al 4 5 developed a CPC/gelatin gel composite. The addition of the gel gave the composite stability and increased the compressive strength with up to 5 % of gelatin gel addition. In most cases, the fabrication of injectable ceramic composites with polymer incorporation improved mechanical properties, but these achievements were limited by the amount of polymer that can be added to the paste. For instance, Mickiewick et al.4 7 reported that after a critical concentration, polymers started forming a thick coating on the ceramic crystal clusters. This prevented them from interlocking, inducing plastic flow and as a consequence, decreasing mechanical properties. Fujishiro et al.4 5 showed that over-loaded gel decreased the strength due to the formation of pores caused by leaching of gelatin in solution. 2.2 Bioceramic Pastes for Medical Applications Bone failure occurs mainly due to diseases, external impacts, aging or malnutrition and must be cured in order to preserve body system. Autografts have been used for 1 2 reconstructing bone tissue. However, limited availability and high costs of two surgeries make 1 3 them less attractive as a long term soluton " . Bone graft substitutes have increasingly become accepted as a good alternative for hard tissue replacement. Injectable bone pastes have been preferred where defects or holes must be fixed immediately and complex shape defects are difficult to reach or simply because of the easy handling. Also, by using bone paste it is possible to prevent microcracks often created by fastening plates, screws and pins ' . 2.2.1 Calcium Phosphate Cement Developed in the late 1980s, calcium phosphate cement (CPC) has been one of the most studied bioceramic pastes for orthopedic applications. CPC powder is mixed with water or aqueous solution and sets at body temperature. Precipitation (setting) reactions may take several hours and form entangled crystals of one or more calcium phosphates. Examples for the cement formulations include powder mixtures of tetracalcium phosphate (TTCP) and dicalcium phosphate anhydrate (DCPA) or dicalcium phosphate dehydrate (DCPD) or P-tricalcium phosphate (P-TCP)48. Although being an excellent alternative to conventional poly(methyl methacrylate) (PMMA) cement in terms of osteoconductivity and thermal damage, CPC are sensitive to excessive liquid intake and wash-out due to the penetration of body fluids before setting. This may result in decrease of mechanical properties of the CPC. Fast setting CPC (i.e. setting in 10 to 30 minutes at 37°C) with addition of viscous polymeric additives49"51'84 into a liquid component was suggested to solve the wash-out problem. In human medical applications, Constantz et al.5 2 have achieved a remarkable success in the minimally invasive treatment of acute fracture through percutaneous administration of CPC. Recent studies showed that CPC injection has been adopted for treating calcaneal bone cyst and pathological fracture cases53; for tibial plateau fracture54; and in filling the bone defects 13 around oral implants35. BoneSource® (Striker, USA) CPC was approved for commercial use C O in 1997 for the repairing and healing of cranial defects. Stelnicki et al. investigated using BoneSource® as an onlay implant for supraorbital and malar augmentation, and the paste showed excellent interaction with soft tissue and steady-fast adherence to the adjust bone. CPC also has been used in dental applications. For example, PROSPEC™ MI paste is a mixture of casein phosphopeptide (CPP) and amorphous calcium phosphate (ACP). This delivers calcium and phosphate ions into the tooth enamel under acidic conditions, resulting in sealing of demineralized dentin tubules. 2.2.2 Hydroxyapati te Pastes Hydroxyapatite pastes recently receive a lot of attention as they do not require the precipitation reaction like CPC in physiological conditions. Furthermore, pre-mixed HAp paste can be applied directly from a ready-to-use syringe into bone defects, and so time-consuming mixing of components before injection is not required. Ostim (Heraeus Kulzer, Germany) is a commercially available HAp paste mostly used in maxillofacial surgeries such as filling bone defects after cystectomy, apicectomy or dental implant inserting. Researchers from the FAP Dental Institute in Japan have developed a HAp paste that can be applied to tooth which works as a synthetic tooth enamel. It can reconstruct enamel without the need of prior excavation. The process not only repairs early caries lesions but can also help to prevent their re-occurrence by strengthening the natural enamel16. 2.2.3 Calcium Hydroxide Pastes The introduction of calcium hydroxide, CaOHz, as an endodontic agent by Hermann in 1920s has since proved calcium hydroxide to possess a pronounced antimicrobial activity against most of the bacterial species found in root canal infections. It also has been used as an intracanal medication in endodontic therapy such as temporary or permanent root canal filling, root canal dressing, and direct pulp capping. The success of endodontic treatment depends on the reduction or elimination of bacteria present in the root canal system. The antimicrobial effect of calcium hydroxide relates directly to its high pH of about 12.5 in the root canal environment. Calcium hydroxide powder has been mixed with various liquid vehicles such as distilled water, saline, anesthetic solution, Ringer's solution, camphorated monochlorophenol, cresatin, and glycerin to form a paste57. Sealapex (Sybron/Kerr, USA) and DT Temporary dressing (Dental Therapeutics AB, Sweden) is commercialized aqueous Ca(OH)2 pastes, and Vitapex (Neo Dental Chemical Products Co., Ltd., Japan) and Diapex (Diadent, Korea) are silicon oil (PDMS)-based Ca(OH)2 pastes56. 2.3 Rheology of Ceramic Pastes One of the key features of bioceramic pastes for clinical applications is its direct delivery into bone or tooth defects without a surgical operation required. However, requirements of delivering the pastes through a rather small diameter needle (typically ~ 500 um) often arise in actual treatments. More specifically, procedures like in-situ fracture fixation in orthopedics, filling root canals or filling gaps caused by dental implants and sealing furcation perforation in endodontics are the examples19. During injection of the ceramic pastes, two competitive phenomena occur through a small diameter needle. One is 1 5 paste flow and the other is paste filtering. While liquid carries particles, they may pack to higher density inside a needle inducing phase separation. This phenomenon is called "filter pressing" and occurs if the pressure required to filter the liquid through the particles is lower than the pressure required to eject the paste59. "Filter pressing" of ceramic pastes happens due to lack of particle carrier (mostly liquid medium), inhomogeneous mixing (large agglomeration of particles) and solidification of the pastes. Hence, injectability of bioceramic pastes must be controlled so that the "filter pressing" can be prevented. Understanding of paste's rheological properties is a key in maintaining the control. 2.3.1 Viscosity of Ceramic Pastes Rheology is the science of the flow and deformation of matter. For fluids, the study of rheology involves measurement of viscosity, i.e. the fluid's resistance to flow. However, behaviour of highly concentrated ceramic suspensions such as slurries and pastes is rather complex, e.g. their rheological properties depend on temperature, shear rate, and concentration of additives and fillers60. Since it is impossible to measure the viscosity of every microscopic volume of a fluid, apparent (global) viscosity determined by a viscometer is generally used with the SI unit of pascal second (Pa.s) or millipascal second (mPa.s) equal to centipoises (cP), eg. 1 Pa.s = 1000 mPa.s = 1000 cP: Shear stress (x) Viscosity (n) = [Pa.s] Eq. 2.3-1 Shear rate (y') Kinetic viscosity, i.e. the ratio of the viscosity to the density of a fluid, is another way to express a fluid's resistance to flow. The S.I unit of kinetic viscosity is meter squared per 1 6 second (m Is), but centistroke (CST) equivalent to mm Is is often used . Newtonian fluids, frequently referred to as "true" liquids, obey the viscosity definition that the shear stress is linearly proportional to the shear rate (i.e. the viscosity coefficient, the line slope, is constant). However, concentrated suspensions do not show the simple proportionality between shear stress and shear rate and they are sometimes called non-Newtonian liquids. Figure 2.3-1 illustrates shear stress vs. shear rate curves of Newtonian and abnormal liquids. Shear-thining (psuedo-plastic) suspension has relatively high viscosity when sheared slowly, but low viscosity when sheared quickly. This is preferred behavior for pastes, as the injection force decreases for faster flow. At the opposite extreme, shear-thickening (dilatant) suspension has low viscosity when sheared slowly, but high viscosity when sheared quickly. In Bingham flow, the materials behave like a Newtonian liquid after the yield stress61. Shear rate Figure 2.3-1. Shear stress vs. shear rate curves of Newtonian and non-Newtonian liquids; (a) Bingham, (b) shear-thining (pseudo-plastic), (c) shear-thickening (dilatant) and (d) Newtonian models. 1 7 For a very dilute suspension of noninteracting spheres in a Newtonian flud, the viscosity for laminar flow is described by the Einstein equation60. The viscosity of a suspension (r]s) is greater than the viscosity of the liquid medium (nL), and the ratio is referred to as the relative viscosity (nr): T1r = r| s/r|L= 1 +2.5/p Eq. 2.3-2 fp: volume fraction of dispersed spheres At a concentration above about 5-10 vol%, interaction between particles during flow causes the viscosity to increase as the volume fraction of particles increases. Viscosity of suspension of uniform spherical colloidal particles is approximated by the Dougherty-Kreiger equation60: T l r = [ l " / p / / c r ] Eq. 2.3-3 ~~fcr'- packing factor at which flow is blocked K\\\ apparent hydrodynamic shape factor of the particles Dynamic viscometers, such as rotational and capillary viscometers, can measure both apparent viscosity and observe rheological properties. On the other hand, a kinematic viscometer can determine kinematic viscosity by measuring time it takes for a fluid to flow through an orifice due to gravity alone. Rotational viscometers shown in Figure 2.3-2 measure the forces on rotating tools while they are immersed in containers of a suspension. Typically, Brookfield rotational viscometer is used for low shear rate measurement (at less than 250 rpm). If high shear rate measurements are required, concentric-cylinder or cone-and-plate viscometers should be used. Capillary viscometer illustrated in Figure 2.3-3 measures 1 8 pressure drop over the length of a capillary tube as suspension is pumped through the tube. The measured flow rate, tube dimension and pressure drop are then converted to viscosity by Hagen-Poseuille equation. However, if high flow rate causes the test suspension to enter the dilatant regime, Fig. 2.3-1, dilatant flow blockage can occur. Figure 2.3-2. Schematic diagram of basic tool geometries for the rotational viscometers (a) concentric cylinder, (b) con and plate and (c) parallel plate. Figure 2.3-3. Schematic diagram of capillary viscometer. Plastic forming bodies have much higher solids content than particle/fluid suspensions and pastes. Plastic body rheometers can measure extremely high viscosities of plastic forming bodies. For example, torque rheometer records energy or torque required to mix the bodies in 1 9 a measuring chamber. Piston extruder method monitors extrusion pressure required to extrude 62 bodies at several flow rates through a standard die . 2.3.2 Pastes Injectability Measurement During clinical treatment, surgeons often report poor injectability of some bioceramic pastes. Such drawbacks were caused by the lack of manufacturer's knowledge of rheological properties on the injectable pastes63. Many researchers have tried to develop simple, practical methods to evaluate injectability of pastes to address this issue. Kharoun et al.17 used a syringe with opening of 2.0 mm diameter to measure injectability of calcium phosphate cements by wt% of the paste extruded from the syringe, as defined in Eq. 2.3-4: Weight of the paste extruded from a syringe x 100 Injectability (%) = Weight of the paste in a syringe Eq. 2.3-4 Applied force for injection was increased up to 100 N, and any sample requiring over 100 N was declared as non-injectable paste. Leroux et al.2 0 designed a syringe connected with a catheter (diameter 3 mm) and a tee connector leading to a manometer. The pressures indicated by the manometer were used to compare injectability of calcium phosphate cement, Eq. 2.3-5. 1 Injectability co : Eq. 2.3-5 Applied pressure for injection The higher pressure applied, the worse the injectability. The methods by Kharoun et al, and Leroux et al., however, were only semi-quantitative, i.e. only useful to show the degree or 2 0 ease of paste injection. Baroud et al.21 designed capillary rheometer type apparatus modifying ASTM standard specification (#F 451-99a) for acrylic bone cement injection. The injection force measured by Universal Testing Machine was calculated into viscosity by Hagen-Poseuille equation: A / ? 7 i r 4 Fr4t Viscosity (n) = = [Pa.s] Eq. 2.3-6 SLQ 8R2LV ^p: injection pressure [Pa] F: injection force [N] r. radius of capillary tube [m] R: radius of barrel [m] L: length of capillary die [m] Q: flow rate [m3/s] V: volume extruded [m3] ^r = extrusion time [s] Bioceramic pastes for dental and maxillofacial applications sometimes must be injected through very small diameter needle (~ 500 um), and flowability of such pastes must be characterized very carefully. However, the opening dimensions of the needle in designs 17 20 21 previously studied ' ' for injection tests were too wide, leading to too low sensitivity for detecting viscosity changes. When it comes to pre-mixed pastes for one-step injection-ready, observation of the viscosity change should be carefully monitored because partial solidification of the pastes occurs with time. The deterioration of injectability which leads to the blockage of a syringe needle in service might then happen during a critical stage of surgical procedure. Therefore, a novel method to evaluate rheological properties of pre-mixed pastes contained in a syringe needs to be developed, defining a rationale for the present work. 2 1 2.3.3 Factors Influencing Rheological Properties of Ceramic Pastes Knowledge, hence control, of rheological properties of ceramic pastes relates ceramic colloidal science. In ceramic suspensions, a number of physical and chemical factors control rheological properties64 • Physical factors S Solid content S Particle size and size distribution V Packing ability • Chemical factors S Chemical additives • pH Solids fractional content, expressed as a ratio of the weight of powder to volume of liquid (P/L), has a major effect on rheological properties. In wet ceramic bodies, fluids mainly serve two purposes: (i) fill pores between packed particles and (ii) help moving the particles through flow. Viscosity of the paste increases with P/L ratio because less accessible liquid for particle flowing exists. Even though P/L ratios are same in different pastes, viscosities can vary depending on the particle size distribution (PSD) and the packing ability of the particles in the pastes. Generally, the broader PSD is, the better the particles can be packed, hence less fluid is required to fill pores, resulting in better flowability of the mix. When particles pack well (i.e. for broad PSD), a broad range of solids contents can be used to produce suspensions with a wide range of viscosities and rheological properties from shear-thining to dilatant64. When particles do not pack well (narrow PSD), even low solid contents can result in high viscosity and rehology tends to be dilatant. 2 2 The pH of ceramic suspension is critical in control particles dispersion. Each particular ceramic in a solvent has an isoelectric point (IEP) which is the pH point where electrostatic charge on the surface is zero. Particles typically flocculate at their IEPs and deflocculate as the pH increases or decreases away from that point. Flocculation of ceramic suspensions causes viscosity to increase, whereas deflocculation lets the particles homogeneously disperse in a suspension and decreases viscosity65. Various chemical additives inducing flocculation or deflocculation are commercially available. Deflocculating chemicals in Table 2.3-1 enhance electrostatic surface charge, provide lubricity during particle/particle collisions, and/or provide steric layers that help to prevent flocculation. When organic deflocculants are added to aqueous suspensions, the polymers adsorb onto the particles by a hydrophobic effect, i.e. water forces organics out of the solution to an interface Table 2.3-1. Common deflocculants used in polar liquids. Inorganic Organic Sodium carbonate Sodium polyacrylate Sodium silicate Ammonium polyacrylate Sodium borate Sodium citrate Tetrasodium pyrophosphate Sodium succinate Sodium tartrate Sodium polysulfonate Ammonium citrate such as the particle surface. This occurs because the carbon-containing backbone of organic additives is usually hydrophobic. When an additive is electrostatically negative and the particles are also electostatically negative, the additives can still adsorb onto and coat particles due to the hydrophobicity of the additives stronger than the electrostatic forces. On the other hand, when anionic deflocculants adsorb onto electrostatically positive surfaces, the net 2 3 positive surface charges are cancelled and the surface becomes electrostatically negative. Many flocculants are soluble inorganic salts of divalent cations such as Ca 2 + and Mg 2 + . These ions help to flocculate suspension by canceling negative electrostatic surface charges and allowing the weak Van der Walls forces of attraction to dominate66. Zeta potential is the effective electrostatic potential at certain distance from suspended particle surface. It measures the electrostatic potential at the shear plane, Eq 2.3-7, which is a "double layer" thickness away from the actual particle surface. Suspended particles interact with each other in response to their zata pontential. Like-charged particles repel, opposite-charged particles fHrjve ¥<; = [mV] Eq. 2.3-7 £ r £ 0 E Yr: Zeta potential fn'. Henry's constant n: viscosity of electrolyte E: imposed electric filed ^ve: electrophoretic mobility. attract and zero surface charge allows van der Walls forces of attraction to pull particles together. When absolute values of zeta potentials are high, generally above 30 mV, suspended particles will be deflocculated. When zeta potentials are low approaching zero, suspended particles will be flocculated. Figure 2.3-4 shows electrostatic potentials as a function of distance from the surface of a particle. 2 4 Repulsion energy VDW Attraction energy Particle m 0 Net potential energy Distance Figure 2.3-4. Potential energy of interaction between two particles with electrical double layers. 2.4 Additives in Bioceramic Ceramic Pastes Flowability of injectable bioceramic pastes can be improved by incorporating additives such as particle dispersants (organic/inorganic deflocculants), polymers and extra non-aqueous fluids. Addition of biocompatible organic dispersants has shown to be very effective in decreasing the viscosity of pastes. Sarda et al.6 8 found that the injectability of a-TCP-water mixtures was increased by 50% with the addition 1.5 wt% citric acid. Ginebra et al.5 0 added a polymeric drug (poly(4-HMA)) in a-tricalcium phosphate-aqueous sodium hydrogen phosphate mixture. The hydrolysis of the TCP into calcium-deficient hydroxyapatite proceeded at a lower rate because of the addition of the polymeric drug. As a consequence, the cement hardening was slower, which resulted in the increased of injectability of the cement. Andrianjatovo and Lemaitre67 observed that the injectability of a brushite CPC was strongly improved by adding polysaccharides. Leroux et al.2 0 used 2 5 sodium glycerophosphate, lactic acid, glycerol and chitosan to improve the injectability of TCP+TTCP cement. The hygroscopic compound of sodium glycerophosphate ( C H 2 O H -CH(0-Na2-P03)-CH20H) quickly adsorbed water, then slowly released it in the whole medium. This improved injectability of the cement, but also increasing the setting time. 4.0 wt% of lactic acid (CH3-CHOH-COOH) addition increased injectability by around 30 %. Viscous and hygroscopic compound glycerol (CH2OH-CHOH-CH2-OH) seems to allow brushite plates (the precipitate resulting from cement hydration) to slide over other, making the paste flow. The use of a small amount of chitosan (< 0.5 %) led to a good fluidity and injectability. However, a larger amount (> 0.5 %) decreased the injectability due to agglomeration of particles. 2.4.1 T r i - s o d i u m C i t r a t e D i spe rsan t Citrate ion is a minority bone component and plays an important role in crystal modification, which prompted studies into interaction of citrate ions and hydroxyapatite surface. Lopez-Macipe et al. 6 9 studied the role of pH in the adsorption of citrate ions on HAp surface. D.N. Misa70 discovered the interaction mechanism between citrate ions from alkali metal citrates and HAp surface. According to the above authors, it was found that an ion-exchange between phosphates and citrate ion species of cit , Hcit , H2cit , and H3cit (cit : citrate ion C3H50(COO)33~, Hcit2-, H2cif: intermediate citrate ions and H3cit: citric acid) at the solid-solution interface results in negatively surface charged HAp particles. The adsorption of citrate ions is caused by a higher affinity of citrate ions than phosphate species for the Ca sites on HAp surface. Therefore, HAp particles become more stable and well-dispersed in polar liquid environment containing citrates. Different species of various surface charges arise depending on the pH of the solution. At pH < 6, the main citrate and phosphate 2 6 species for the interaction are Hcit2 and H 2 P0 4 , but cit3 and H2PO4 are at pH > 6. Figure 2.4-1 shows distribution of citrate species at 25 °C as a function of pH. This figure was removed because of copy right restriction. The figure removed can be found at the Journal of Colloide and Interface Science, 200, L.M. Anabel, G.M. Jaime, and R.C. Rafael. The role of pH in the adsorption of citrate ions on Hydroxyapatite, 114-120 (1998). Figure 2.4-1. Distribution of citrate species (cit3", Hcit2", H2cit", and Hscit) at 25°C as a function of pH. Citric acid (CA) and metal alkali citrates such as tri-sodium citrate (TSC) and tri-potassium citrate (TPC) of which molecular structure is shown in Figure 2.4-2 are commonly used water-soluble particle dispersants in ceramic processing. They prevent spontaneous agglomeration of ceramic particles in polar liquids by creating repelling potential energy higher than attraction (Van der Waals) energy for adjacent particles. Uwe Gbureck et al.31 reported 30 % increase of injectablity of the calcium phosphate cement (CPC) consisting of tetracalcium phosphate (TTCP) and dicalcium phosphate anhydrate (DCPA) by adding 500 mM TSC solution compared to the pure-water based cement. The high zeta potentials caused by adsorption of the multiple charge citrate ions enabled the effective dispersion of the cement particles with decreasing viscosity. Also, they enhanced strength of the set cement by decreasing micro-pore size resulted from the orderly packed particles under the effect of the citrate ions. 2 7 (a) H ( b ) H ( c ) H 1 1 Na + 1 K + H— C — COOH H - C - C O O - H— C — C O O -^ I Na + I K + H O - C — COOH H O — C - C O O - H O — C — C O O -^ ^ Na + ^ K + H— C — COOH H - C - C O O " H— C — C O O -I I I H H H C 6 H 8 0 7 C 6 H 5 Na 3 0 7 C 6 H 5 K 3 0 7 Figure 2.4-2. Molecular structure of (a) citric acid, (b) tri-sodium citrate and (c) tri-potassium citrate. 2.4.2 Ethylene Glycol Ethylene glycol (EG) is a colorless and odorless dialcohol (HO-CH 2-CH 2-OH). The viscous and hygroscopic liquid EG is slightly heavier than water, i.e. the specific gravity is 1.11, with a melting point of-13 °C and boiling point of 198 °C 7 1 . Since EG has been a major industrial chemical, toxicity and poisoning of EG has been studied extensively on mice, rats, and rabbits72'73. In vivo, EG undergoes rapid biodegradation into four organic acids: glycoaldehyde, glycolic acid, glyoxylic acid and oxalic acid. Among these metabolites, glycolic acid was reported to cause acidosis. Approximately 1.4 ml/kg of 100% EG is the estimated as a lethal dose72. One of the interesting physico-chemical properties of EG is that it is miscible with water. EG has capability of forming intermolecular and intramolecular hydrogen bonding. In the study of water-EG binary mixtures conducted by de Oliveria et al.74, the structural parameters observed for EG molecules in the binary solution were found to be very similar to the ones found in pure EG. Therefore, it is believed that there is the similarity between the driving force of the molecular bonding in EG and water, and this leads to small changes on the molecular structures in the binary solutions, and thus to miscibility with each 2 8 other. The viscosity of aqueous ethylene glycol solutions have been investigated by many scientists, e.g. Jeonome, Dunstam and Hayduk75. Viscosity of ethylene glycol and water binary solutions increases almost linearly with mole fraction of ethylene glycol; ethylene glycol is getting less viscous as a temperature increases in the range of 0 to 50 °C. These properties of EG make it useful in various applications, e.g. as a component of: • plastics manufacturing • permeable cryoprotectant in cryobiology • a freeze storage of many microorganisms Furthermore, hydrophobic hydrocarbon chain of EG plays a role of surface active agent by lowering surface tension. The EG mixed solvent lets the particles in aqueous suspension stabilized by forming steric layer on the surfaces. Poly(ethylene glycol)(PEG) is considered non toxic and is used in a variety of health products, e.g. skin creams and lubricants, frequently combined with glycerin. When attached to various protein medications, PEG allows a slowed removal of the protein from the blood. It also has been shown that PEG can improve healing of spinal injuries in dogs85. 2.4.3 Poly(dimethyl siloxane) Polysiloxane polymers have been widely used in medial fields as polymeric biomaterials due to their favorable material properties. Poly(dimethyl siloxane) (PDMS) is one of the most commonly used, containing long backbone chain of siloxane (-Si-O-Si-) with attached dimethyl (CH3-CH3) functional groups to each Si atom (Figure 2.4-3). Combination of the inorganic chain and organic substituents on silicon resulted in low surface energy, high elasticity, dimensional stability, hydrophobiciy, high oxygen permeability and resistance to 2 9 aggressive chemicals. Especially, low surface energy of PDMS (PDMS: 21.2, epoxy: 40-50, and water: 73 dyne/cm) reduces the possibility of thrombogencity to occur. When other polymeric biomaterials with higher surface energy directly contact with blood, blood cells and proteins adsorb onto the surfaces, platelets adhere and eventually coagulation happens . Medical grade PDMS (Dow Corning Co., USA) has been used mainly for the treatment of facial deformities such as Romberg's disease, facial lipodystropy, first and second branchial arch deformities, and some posttraumatic and postopertative facial deformities. The material has also been used to correct secondary nasal fractures, postradiation facial deformity, 77 receding chin, and oral and perioral deformities such as alveolar ridge resoprtion and rhytids . H H H H I l l i H-CH HC-H H-C-H H-C-H I l l i — S i — O — S i — O — S i — O — Si — I l l i H-CH H-CH H-C-H H-C-H I l l i H H H H Figure 2.4-3. Molecular structure of poly(dimethyl siloxane). Moreover, PDMS has been used with inorganic materials to make inorganic/oganic hydrid composites in the form of injectable paste, membrane and solid block for bone regeneration. Neand Ignjatovic et al.7 8 synthesized injectable PDMS/HAp composite cement by cross-70 linking PDMS matrix where HAp was incorporated as filler. Maeda et al. prepared poly(L-lactic acid)-PDMS-calcium carbonate hydrid membranes for guided bone regeneration. Chen et al.8 0 precipitated apatite on PDMS-modified CaO-Si02-Ti02 hybrids by soaking the composite in a simulated body fluid solution. 3 0 3 SCOPE AND OBJECTIVES 3.1 Scope of the Investigation HAp pastes have shown successful performance in reconstruction of hard tissue defects, mostly in non-structural applications such as craniofacial, maxillofacial and endodontic areas. The biological properties of HAp (biocompatibility, osteoconductivity and bioactivity) and the advantages of an injectable pastes (non-invasive surgery treatment, easy handling and applicability into complex geometry by self-flowing) allow HAp pastes to have unique advantages over other forms of biomaterials. Pre-mixed HAp pastes can be injected directly from a ready-to-use syringe without time-consuming mixing of components, such as required for calcium phosphate cements and polymeric cements. The optimum paste should be easy to handle and inject, conforming well to the details of the cavity while resistant to slumping and washout once in place. This qualitative description was not however, so far, followed with quantification of the "injectability" parameters, although some limited studies were reported in literature (refer to Chapter 2). However, surgeons frequently complain about difficult pastes injection caused e.g. by improper mixing of solid and liquid medium, unstable (or too high / low) viscosity, washout sensitivity, localized solidification and so on. Also, increased solids loading (necessary for volumetric stability after injection) deteriorates rheological properties of non-optimized pastes. The primary scope of the present work is the study and quantification of the flow and injectability of HAp-based pastes, using a simple syringe-based device. Although there were previous studies of pastes injectability, the resulting parameters describing injectability were 3 1 qualitative and scientifically vague. The sensitivity of detecting change of injectability was insufficient. The experimental models resulting from previous studies could not be used to assess injectability of the pastes through a small diameter needle (~ 500 um). The current work addresses this lack of fundamental knowledge of the pastes behavior assessment. It is hypothesized that stable viscosity in the range 8-10 Pa.s for the paste with large volume fraction of solids (> 40 vol%) constitutes a desirable combination of quantitative properties of pastes. However, the general goal is to better understand and improve the rheological properties of HAp pastes, so that stably injectable HAp pastes with high solid loading can be produced. In order to address these issues, a simple syringe-based viscometer which can sensitively evaluate injectability of the pastes has been studied in the thesis. The effects of chemical surfactant additives on the injectability of the HAp pastes were investigated. The uniqueness of the present study is the application of a simple, low-cost, syringe-based viscometer for correlating viscosity with injectability of HAp-based bioceramic pastes, in a wide range of viscosities of 1-15 Pa.s. The system can be easily reproduced by any lab involved in rheological research, with minimal expense. 3.2 Objectives In view of the above scope of the investigation, the followings are the detailed objectives of this work; 1. To design, calibrate and evaluate a simple and practical viscosity measuring device and method using a commercial syringe (Diadent, Korea) for assessing injectability of HAp-based bioceramic pastes. 3 2 2. To study the effects of surfactants on rheological properties of HAp-based bioceramic pastes, including: • Ethlyene glycol (EG) • Tri-sodium citrate (TSC) • Ethylene glycol + Tri-sodium citrates (EG + TSC) • Poly(dimthyl siloxane) (PDMS) 3. To correlate the effects of the surfactants on the pastes injectability by investigating: • HAp particle size and size distribution • Surface charge of HAp • HAp particle dispersion in the pastes 4. Based on the outcome of the Objectives 1, 2, and 3, the composition of the EG + TSC additives containing water-based HAp pastes was optimized to achieve stably injectable pastes with maximum HAp solids content. 3 3 4 EXPERIMENTAL METHODOLOGY 4.1 Materials and Paste Preparation The phase and crystallinity of the HAp powder (Caio(P04)6(OH)2, catalog # 04238, Sigma Aldrich, USA was confirmed by an X-ray diffractometer (XRD, Rigaku MultiFlex, Japan). All peaks were identified as HAp peaks, as shown in Figure 4.1-1. The morphology of the HAp powder was verified by a scanning electron microscopy (SEM, S3000N, Hitachi, Japan). Figure 4.1-2 shows that HAp particles have almost round shape, with high degree of agglomeration. Particle size distribution of the HAp was measured by particle size analyzer (PSSD, Horiba CAPA-700, Japan), in distilled water. Although some particles may be as large as 20 um as seen in Figure 4.1-2, the median size of the HAp particle was 2.79 um, and the particle size ranged from 0.1 to 7.0 um, as shown in Figure 4.1-3 (F(%): Mass percent, CMPF(%): Cumulative Mass Percent Finer). 1200 -i 1 20 30 40 50 60 70 2 Theta (degree) Figure 4.1-1. XRD spectrum of HAp powder from Sigma Aldrich. 3 4 Figure 4.1-2. SEM micrograph of HAp powder. 25 20 -0-1 0-2 0.3 0.4 0.5 0.6 0.7 0.8 0.91.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.010.0 Particle size (um) Figure 4.1-3. Particle size and size distribution of HAp powder. 3 5 Four types of HAp-based bioceramic pastes were produced by mixing the HAp powder with different liquid media (de-ionized water DI, ethylene glycol EG, poly(dimethyl siloxane) PDMS and the particle dispersant (tri-sodium citrate, TSC). Ethylene glycol ( H O C H 2 C H 2 O H , ) was purchased from Fisher Scientific, USA and tri-sodium citrate dihydrate (C6H5Na307-2H20) from EM Science, USA. The pastes are referred to as follows: • EG-HAp paste: HAp + DI water + Ethylene glycol • TSC-HAp paste: HAp + DI water + Tri-sodium citrates • EG-TSC-HAp paste: HAp + DI water + Ethylene glycol + Tri-sodium citrates • PDMS-HAp paste: HAp + poly(dimethyl siloxane) The paste composition is expressed as powder to liquid ratio (P/L, g/ml), vol% of EG of a liquid medium and wt% of TSC of a paste weight. All compositions of the EG-HAp, TSC-HAp pastes and PDMS-HAp pastes studied in this work are compiled in Table 4.1-1. Compositions of the EG-TSC-HAp pastes prepared for compositional optimization using the Design Expert 5.0.7 software program for a statistical analysis, are displayed in Table 5.4-1 in Section 5.4. An intensive planetary ball milling was applied for homogenization mixing of the pastes as pictured in Figure 4.1-4 (a), (b) and (c). The HAp powder was loaded into an alumina jar (Volume = 250 ml, 99.7 % A1203) with alumina balls (D = 20 mm, 99.7% A1203) and milled at 300 rpm for 30 min. (4.0 to 5.0 wt% AI2O3 contamination by the friction of the AI2O3 jar and balls was detected by energy dispersive spectroscopy installed in the SEM.) with Pulverisette® 5 Planetary Mill (Fritsch, Germany). 1.0 ml of the mixed paste was carefully inserted into a Diapex syringe, and a commercial needle (BD PrecisonGlide™ 20G 1, USA) was attached to the end of the syringe tip for an injection test, as pictured in Fig. 4.1-4 (d). The disposable Diapex syringes used for tooth root canal filling product (Diapex) were generously donated by Diadent Inc., Korea. 3 6 Table 4.1-1. Compositions of the EG-HAp, TSC-HAp and PDMS-HAp pastes studied. EG-HAp pastes TSC-HAp pastes P/L E G (vol%) DI water (vol%) Injection test* P /L T S C (wt%) DI water (vol%) Injection test* 0.7 0 100 0 0.9 0.0 100 0 20 80 - 0.2 100 0 40 60 - 0.4 100 X 60 40 - 0.6 100 X 80 20 - 1.0 0.0 100 X 100 0 - 0.2 100 0 0.8 0 100 0 0.4 100 0 20 80 0 0.6 100 -40 60 0 1.1 0.0 100 X 60 40 X 0.2 100 X 80 20 X 0.4 100 0 100 0 X 0.6 100 0 0.9 0 100 0 1.2 0.0 100 X 20 80 0 0.2 100 X 40 60 0 0.4 100 0 60 40 0 0.6 100 0 80 20 - 1.3 . 0.0 100 X 100 0 - 0.2 100 X 1.0 0 100 X 0.4 100 X 20 80 0 0.6 100 0 40 60 0 1.4 0.0 100 X 60 40 0 0.2 100 X 80 20 0 0.4 100 X 100 0 0 0.6 100 X 1.1 0 20 40 100 80 60 X X X 60 40 0 PDMS-HAp pastes 80 20 0 P/L P D M S (vol %) Injection test* 100 0 0 0.1 100 -1.2 0 100 X 0.2 100 0 20 80 X 0.3 100 0 40 60 X 0.4 100 0 60 40 0 0.5 100 0 80 20 0 0.6 100 X 100 0 0 * - : no injection test performed due to excessive fluidity of paste, x: no injection test performed due to excessive viscosity of paste and o: injection test performed. 3 7 Figure 4.1-4. (a) HAp powder in an alumina jar, (b) Pulverisette® 5 Planetary Mill, (c) HAp paste after milling and (d) HAp paste in a Diapex syringe. 4.2 In jec tab i l i t y E v a l u a t i o n 4.2.1 Sy r i nge V i s c o m e t e r C a l i b r a t i o n A capillary viscometer schematically illustrated in Figure 4.2-1 (a) uses the Hagen-Poiseuille's equation (Eq. 2.3-6) to measure viscosity of Newtonian fluid penetrating through a capillary tube, and has been studied before81. It is important that the capillary dimensions R and r are measured with high precision due to the extreme sensitivity of viscosity coefficient calculation to the capillary dimensions, i.e. viscosity is proportional to r4/R2 in Eq. 2.3-4. The length to diameter ratio (L/2r) should be normally between 20 and 40. Larger ratios and ratios O I less than suggested require applying large corrections to the data . In addition, the ratio of the reservoir diameter to capillary diameter (R/r) should be between 3 and 15. In this work we have closely reproduced a capillary viscometer geometry using a commercial syringe. The 3 8 Figure 4.2-1 Schematic illustration of (a) a general capillary viscometer used in previous research and (b) syringe viscometer evaluated in this work. followings are the dimensions of a Diapex syringe and a BD PrecisonGlide™ 20G 1 needle schematically illustrated in Figure 4.2-1 (b): • Radius of BD needle (r): 0.3 2 x 10"3 m • Length of BD needle (I): 29.0 x 10"3 m • Radius of Diapex Syringe (Ri): 2.4 x 10"3 m • Radius of Diapex Syringe tip (R2): 1.0 x 10"3 m The dimensional ratios of L/2r = 45.3 and R;/r = 7.5 meet approximately the requirements indicated above for proper capillary viscometer design (the ratio of L/2r is slightly larger than the suggested value of 20-40). Viscosity is calculated from Hagen Poiseuille's equation: Apnr4 Fr4t Viscosity (n) = = [Pa.s] Eq. 2.3-6 %LQ %R2LV 3 9 p: injection pressure [Pa]; p=F/KR F: injection force [N] r. radius of capillary tube [m] R: radius of barrel [m] L: length of capillary die [m] Q: flow rate [m /s] V: volume extruded [m3] ^ t = extrusion time [s] Hagen Poiseuille's equation assumes that the material flows through a capillary tube in a laminar flow, and is a Newtonian fluid (i.e. viscosity is independent of the flow rate). However, since HAp-based bioceramic pastes are non-Newtonian fluids, the data acquired from injection tests using the syringe viscometer must be corrected through calibration tests with fluids of well known viscosity. The expected factors causing measurement errors include plunger friction, plunger back pressure, paste compressibility, capillary entrance and end effects, turbulence and drainage. To use the equation 2.3-4 for the Diapex syringe of geometry illustrated in Figure 4.2-1 (b), the decreased pressure and the increased flow velocity in the syringe tip due to R\ and R2 should be compensated through calibration, Brookfield Standard Viscosity Fluids of 5.0 Pa.s and 12.5 Pa.s were used for viscometer calibration because the range of the two viscosities is close to that of Diapex calcium hydroxide paste. After the two standard viscosity fluids were filled in the syringe, injection forces were measured as described in Section 4.2.2 in order to calculate experimental viscosities and then compare the values with the known viscosities of the fluids. 4.2.2 Injectability Tests Figure 4.2-2 shows an installation of the syringe viscometer. Applied injection force was measured by Instron 3360 series Universal Testing System (Instron Inc., USA). HAp-based bioceramic paste loaded syringe was securely inserted into the syringe holder, and the compression plate of the Instron 3360 series UTS was slowly moved down to contact the syringe plunger end. 1.0 ml of the paste was ejected at crosshead speed of 1.2 x 10" m/s for 46 s. S-9 program (Instron Inc., USA) automatically plotted an injection force vs. diplacement Figure 4.2-2. Syringe viscometer installed on Instron 3360 series Universal Testing System. graph, and the 5 highest peaks and the 5 lowest troughs between 10 to 40 mm of the displacement were chosen for calculation of an average injection force. The average injection force was used to calculate viscosity of the paste using the calibrated equation (Eq.5.1-1). The injection test was repeated 3 times on each paste. 4 1 4.3 Characterization of HAp-based Bioceramic Pastes 4.3.1 Micrpstructural Characterization Morphology of dried pastes and HAp particle dispersion was observed by scanning electron microscopy (SEM, S3000N, Hitachi, Japan). The paste extruded from the syringe was dried at room temperature, gold-coated and viewed under 5.0 kV beam. 4.3.2 Particle Size and Size Distribution HAp particle size distribution of the pastes was analyzed by a particle size analyzer (PSSD, Horiba CAPA-700, Japan). After paste was dispersed in distilled water with magnetic stirring and ultra-sonification, the suspension was loaded into a measurement cell and centrifuged at 960 rpm for 6 min. 11 sec. for a light scattering measurement on sedimenting particles. 4.3.3 Zeta Potential Measurement Zeta potential of HAp particles in the pastes was measured by a zeta potential analyzer (ZetaProbe, Colloidal Dynamics, Canada). 7.0 wt% of paste suspension concentration was prepared by diluting pre-weighed amount of paste with DI water, and homogeneously dispersed by a magnetic stirrer for 24 hrs. The suspension was titrated to pH 11 with 2.0 M NaOH solution, and zeta potential of the suspension particles was measured while 2.0 M HC1 solution was added until pH 4.0. 4.4 Optimization of E G - T S C - H A p Pastes The combined effect of EG and TSC admixtures on the rheological properties of EG-TSC-water-HAp pastes was studied using designed experiments, in particular the response surface central composite design (RS-CCD). The experimental design was based on the study of EG and TSC on HAp paste injectability. The design and the statistical analysis of the reological data of EG-TSC-HAp pastes were performed using the Design-Expert 5.0.7 software (Stat-Ease Inc., USA). The following optimization steps were performed. The program suggests the best experimental outline in terms of the pastes compositions, and the viscosity of all compositions is determined (3 repetitions for each), Table 5.4-1. The program creates a least-squares model of viscosity, as a function of the amount of the components. The model is then used to identify the compositions resulting in the lowest viscosity for the highest solids loading. The suggested compositions are verified through experiments. Further procedural details are included in Appendix, and reference text83. 4 3 5 RESULTS AND DISCUSSION 5.1 Syringe Viscometer Calibration The injection force for the Diapex Syringe + needle, containing no fluid to determine the plunger resistance, and containing the standard 5.00 Pa.s and 12.5 Pa.s viscosity fluids were measured by the syringe viscometer, using the same displacement rate of 1.2 mm/sec as for the experiments with pastes. The injection force profile of the fluids and the resistance (three tests for each) are shown in Figure 5.1-1. By subtracting the average plunger resistance of 3.02 N, the average of the absolute injection forces required to eject the calibration fluids were obtained and are listed in Table 5.1-1. For calculation of the "theoretical" viscosity of the 5.00 Pa.s and 12.5 Pa.s standard fluids, the forces and other test parameters (i.e. syringe geometry) were inserted into the Hagen Poiseulle's equation (Eq. 2.3-6). It was found that the equation overestimates the viscosity, i.e. + 6.18 % and + 4.76 % error for the each fluid was generated as compared with the viscosity of standard fluids. To compensate for the calculation errors, (1 - K) correction factor was included, as shown in Eq. 5.1-1. Repeated calibration with the standard fluids (with inclusion of the correction factor in the equation for viscosity) reduced errors to less than ± 1.0 % (Table 5.1-1). Hence, the equation of 5.1-1 has been used for viscosity calculation of the HAp-based bioceramic pastes, with the estimated accuracy of Table 5.1-1. Viscometer calibration results using Brookfield Standard Viscosity Fluids of 5.00 and 12.5 Pa.s. Averege Viscosity Error Viscosity Error Fluids injection force (BC*) (BC*) (AC&) (AC&) 5.00 Pa.s 17.7 N 5.31 Pa.s + 6.18% 5.02 Pa.s + 0.450 % 12.5 Pa.s 39.3 N 13.1 Pa.s + 4.76 % 12.4 Pa.s - 0.897 % BC: Before calibration AC: After calibration 4 4 (F-Fry4t Viscosity (rjAC) = (1 - K) [Pa.s] E q . 5 . 1 - 1 SR2LV r~ T]AC- Viscosity after calibration K: Constant of 0 . 0 5 4 (the average of + 6 . 1 0 % and + 4 . 6 5 % errors) Fr: Resistance of the plunger r, L, V: geometrical parameters (refer to E q . 2 . 3 - 6 ) measurements better than 1 % . However, it is expected that the complex interactions of the HAp particles would occur during the flow of the pastes. Hence, viscosity of the pastes measured by the syringe viscometer should be considered as the empirical value rather than the scientific standard meaning of viscosity. Therefore, this "empirical value" is called "apparent viscosity" throughout this thesis. 20 30 Displacement (mm) TT 50 60 Figure 5 . 1 - 1 . Plunger resistance (a) and injection force profile for Brookfield Standard Viscosity Fluids of (b) 5 . 0 0 Pa.s and (c) 1 2 . 5 Pa.s viscosity. 4 5 5.2 Injectability Evaluation and Characterization of E G - H A p Pastes 5.2.1 Effect of Powder-to-Liquid (P/L) Ratio on Injectability of E G - H A p Pastes The solids loading amount in an injectable paste was found to be an influential factor controlling the flowability of paste. The ratio of powder to liquid, P/L (g/ml), has been used to represent the solid amount contained in paste. In the present study, the effects of a P/L ratio on the EG-HAp pastes flowability was investigated by measuring injection force of the pastes with increasing P/L ratio by 0.1 g/ml. Table 5.2-1 shows the range of the P/L ratio and the Table 5.2-1. P/L ratio range and injectability test results of EG-HAp pastes. EG (vol %) P/L (g/ml) Vol%ofHAp Injection force (N) Apparent Viscosity (Pa.s) 0 0.7 18.1 5.79 ±0.263 0.95 ± 0.0898 0.8 20.2 16.7 ± 1.06 4.65 ±0.363 0.9 22.2 48.5 ±1.85 15.5 ±0.632 20 0.8 20.2 15.4 ±0.487 4.22 ±0.166 0.9 22.2 46.8 ± 1.68 14.9 ±0.575 1.0 24.0 64.6 ± 1.31 21.4 ±0.449 40 0.8 20.2 11.3, ±0.389 2.83 ±0.133 0.9 22.2 32.9 ± 1.14 10.2 ±0.389 1.0 24.0 52.1 ± 1.50 16.8 ± 0.512 60 0.9 22.2 3.53 ±0.294 0.172 ± 0.100 1.0 24.0 4.55 ±0.209 0.521 ±0.0712 1.1 25.8 8.07 ± 0.447 1.72 ±0.153 1.2 27.5 29.6 ± 1.16 9.09 ±0.397 80 1.0 24.0 4.19 ± 0.210 0.398 ±0.0720 1.1 25.8 6.28 ±0.541 1.11 ±0.185 1.2 27.5 16.8 ±0.793 4.70 ±0.271 100 1.0 24.0 3.19 ± 0.117 0.0566 ±0.0400 1.1 25.8 15.4 ±0.480 4.22 ±0.164 1.2 27.5 43.6 ±0.817 13.9 ±0.279 4 6 100 80 (a) 0.7 g/ml EG(0 vol%)-HAp paste < 18.2 vol%HAp> u 60 u 0) 40 20 10 20 30 40 50 60 100 80 (b) 0.8 g/ml EG(0 vol%)-HAp paste < 20.2 vol% HAp > o o o 60 c o o 40 H 20 10 20 30 40 50 60 100 80 (c) 0.9 g/ml EG(0 vol%)-HAp paste < 22.2 vol% HAp > p 60 c o •5 4 ° u 20 10 20 30 40 Displacement (mm) 50 60 Figure 5.2-1. Injection force profiles for EG (0 vol%)-HAp pastes. 4 7 100 80 H 8 60 o ~ 40 o u 20 (a) 0.8 g/ml EG(20 vol%)-HAp paste < 20.2 vol% HAp > 100 80 (b) 0.9 g/ml EG(20 vol%)-HAp paste < 22.2 vol% HAp > 100 10 20 30 40 50 60 10 20 30 40 50 60 Temporary blockage 10 20 30 40 50 60 Displacement (mm) Figure 5.2-2. Injection force profiles for EG (20 vol%)-HAp pastes. 4 8 100 0) u 80 60 (a) 0.8 g/ml EG(40 vol%)-HAp paste < 20.2 vol% HAp > ~ 40 u 0) 20 10 20 30 40 50 100 80 (b) 0.9 g/ml EG(40 vol%)-HAp paste < 22.2 vol% HAp > 60 20 30 40 50 60 100 80 (c) 1.0 g/ml EG(40 vol%)-HAp paste < 24.0 vol% HAp > 10 20 30 40 Displacement (mm) Temporary blockage Figure 5.2-3. Injection force profiles for EG (40 vol%)-HAp pastes. 4 9 100 80 1 c o 40 o E" (a) 0.9 g/ml EG(60 vol%)-HAp paste < 22.2 vol% HAp > 20 1 10 20 30 40 D i s p l a c e m e n t (mm) 100 80 i O DU £ c o ••8 4 0 0) (c) 1.1 g/ml EG(60 vol%)-HAp paste < 25.8 vol% HAp > 20 10 20 30 40 D i s p l a c e m e n t (mm) 100 80 60 i c o = 40 o a 20 (b) 1.0 g/ml EG(60 vol%)-HAp paste < 24.0 vol% HAp > 50 60 10 20 30 40 D i s p l a c e m e n t (mm) 50 60 100 80 60 c •22, 20 0 4 (d) 1.2 g/ml EG(60 vol%)-HAp paste < 27.5 vol% HAp > 50 61 10 20 30 40 D i s p l a c e m e n t (mm) 50 60 Figure 5.2-4. Injection force profiles for EG (60 vol%)-HAp pastes. injectability test results for the EG-HAp pastes. The profiles of the injection force measurements of EG (0 vol%)-HAp Pastes are shown in Figure 5.2-1. For EG-free HAp pastes, 0.9 g/ml was the maximum value for paste still to be ejected through the needle attached to the syringe viscometer. Comparing to reproducible injection profiles of the 0.7 g/ml paste, the injection tests for the 0.9 g/ml paste produced significant increase but also fluctuation of the injection forces. As will be further discussed in Section 5.2.3, it appears that such deviation was caused by unhomogeneously mixed components of the 0.9 g/ml paste due 5 0 100 80 Seo c '•§40 H •Si. 'c1 (a) 1.0 g/ml EG(80 vol%)-HAp paste < 24.0 vol% HAp > 20 10 20 30 40 100 80 OJ 60 o J 40 u d) (b) 1.1 g/ml EG(80 vol%)-HAp paste < 25.8 vol% HAp > 20 50 60 10 20 30 40 50 60 100 80 (c) 1.2 g/ml EG(80 vol%)-HAp paste < 27.5 vo% HAp > S 60 £ c o ~ 40 20 10 20 30 40 Displacement (mm) 50 60 Figure 5.2-5. Injection force profiles for EG (80 vol%)-HAp pastes. 5 1 100 10 20 30 40 50 60 100 80 P 60 c o •5 40 a> 20 (b) 1.1 g/ml EG(100 vol%)-HAp paste < 25.8 vol% HAp > 10 20 30 40 50 60 100 80 S 60 i-.o (c) 1.2 g/,ml EG(100 vol%)-HAp paste < 27.5 vol% HAp > 20 30 40 Displacement (mm) 50 60 ;ure 5.2-6. Injection force profiles for EG (100 vol%)-HAp pastes. to increased HAp powder loading. The P/L change of 0.2 from 0.7 to 0.9 g/ml made the viscosity of EG(0 vol%)-HAp paste increase >16 times. Deterioration of EG-HAp pastes injectability due to P/L ratio increase was more clearly observed for 1.0 g/ml EG (20 vol%) and EG (40 vol%)-HAp pastes. As seen in Figure 5.2-2 (c) and Figure 5.2-3 (c), fluctuation of up to about ±10 N in injection force was observed for both 1.0 g/ml pastes. The fluctuations are believed to be caused by temporary blockage of the flow in the well known friction science "stick-slip" fashion82. Relative to the EG (0, 20 and 40 vol%)-HAp pastes, the injection forces of the EG (60 and 80 vol%)-HAp pastes slightly increased with the P/L ratio. The average forces for 0.9, 1.0 and 1.1 g/ml EG (60 vol%)-HAp pastes were 3.53 ± 0.294, 4.55 ± 0.209 and 8.07 ± 0.447 N, respectively. The average forces of 1.0, 1.1 and 1.2 g/ml EG (80 vol%)-HAp pastes were 4.19 ± 0.210, 6.28 ± 0.541 and 16.8 ± 0.793 N respectively. The injection force for EG (100 vol%)-HAp pastes also increased with the P/L,ratio from 1.0 to 1.2 g/ml. However, the average injection force for each of the 1.1 and 1.2 g/ml EG (100 vol%)-HAp pastes was higher than those for EG (60 and 80 vol%)-HAp pastes with the same P/L ratio. More detailed discussion of this phenomenon will be provided later in Sections through 5.2.2 and 5.2-4. Analysis of the EG-HAp pastes injection profiles indicates that the temporary blockage of the needle ("stick-slip") appeared only for the pastes which required over 50 N of an injection force. Increase of P/L ratio made also an injection force to gradually increase with an injection time, which indicates deterioration of paste flowability. Figure 5.2-7 compiles the apparent viscosity data obtained from the injection force graphs. The viscosity increase of the EG-HAp pastes is obviously strongly dependent on P/L ratio, but also on EG content. This is further discussed in details in the following Section 5.2.2. 5 3 100 re o. $0.1 o 0 vol % EG —v — 20 vol % EG LT- — 40 vol % EG 60 vol % EG A 80 vol % EG - -o-— 100 vol % EG 0.01 - r 0.7 —r~ 0.8 —I 1— 0.9 1.0 P/L ratio (g/ml) —c~ 1.1 1.2 Figure 5.2-7. Apparent viscosity change vs. P/L ratio of EG-HAp pastes for a constant EG concentration. 5.2.2 Effects of EG on Injectability of EG-HAp Pastes Surfactants are surface active agents that lower surface tension of the solvent in which they dissolve. They also act as stabilizing agents, helping particles dispersion. A disadvantage of water as a liquid medium for paste is its relatively high surface tension of YLV =73 dyne/cm, which makes wetting of the particles in the paste more difficult (this is a direct consequence of the well known equation for surface tension equilibrium at Liquid-Vapor-Solid triple point, i.e. YSV=YSL+7LVCOS9 , where 9 is the wetting angle; when YLV increases, the wetting angle also increases). It is hypothesized in the present study that water soluble EG, with the lower surface tension of YLV=47.7 dyne/cm, would be effective in decreasing the overall surface 5 4 tension of the water-EG solution, thus improving wettability of the liquid on HAp solid, thus breaking down the agglomerated HAp particles during ball milling for producing better homogeneity of the EG-HAp paste, with greater flowability. 100 80 8 60 o "fj 40 CD (a) 0.8 g/ml EG-HAp pastes < 20.2 vol% HAp > 0 vol% EG 20 vol% EG 40 vol% EG 20 30 40 Displacement (mm) Figure 5.2-8. Injection force profiles for (a) 0.8 g/ml EG-HAp pastes and (b) 0.9 g/ml EG-HAp pastes with different EG concentrations. 5 5 100 20 30 40 Displacement (mm) Figure 5.2-9. Injection profiles for (a) 1.0 g/ml EG-HAp pastes with different EG concentrations. The effects of EG on the injectability of EG-HAp pastes were investigated by measuring injection force of the pastes having different EG concentration from 0 to 100 vol%. For the 0.9 g/ml pastes in Figure 5.2-8 (b), the average injection forces for the 0, 20, 40 and 60 vol% pastes were decreasing, i.e. were on average 48.5 ± 1.85, 46.8 ± 1.68, 32.9 ±1 .14 and 3.53 ± 0.294 N, respectively. By adding 40 vol% EG into the 1.0 g/ml paste, the fluctuation of the injection force was reduced as compared to the EG (20 vol%)-HAp pastes in Figure 5.2-9. The higher concentration of 60, 80 and 100 vol% EG in the 1.0 g/ml pastes caused the average injection force to decrease to 4.46 ± 0.209, 4.19 ± 0.210, 3.19 ± 0.117 N, respectively without any needle blockage ("stick-slip"). On the other hand, the injectability of the 1.1 and 1.2 g/ml pastes with 100 vol % EG became worse than that of the 60 and 80 vol% EG pastes as shown in Figure 5.2-10. The average injection force of the 1.1 g/ml EG (100 vol 100 80 S 60 c o "JM 40 u 0 (a) 1.1 g/ml EG-HAp pastes < 25.8 vo% HAp > 60 v o l % EG SO voi% EG 100 vol"/, EG 100 80 i (b) 1.2 g/ml EG-HAp pastes < 27.5 vol% HAp > 60 voi% EG 80 vol% EG 100 vol% EG 20 30 40 50 Displacement (mm) 60 Figure 5.2-10. Injection profiles for (a) 1.1 g/ml EG-HAp pastes and (b) 1.2 g/ml EG-HAp pastes with different E G concentrations. %)-HAp paste (15.4 ± 0.480 N) was ~2.5x higher than that of the EG (80 vol%)-HAp paste (6.28 ± 0.541 N) and twice higher than that of the EG (60 vol%)-HAp paste (8.07 ± 0.447 N). The average injection force of the 1.2 g/ml EG (60 vol%)-HAp paste (29.7 ± 1.16 N) was 5 7 lower than the 80 vol% EG paste of 16.8 ± 0.793 N. However, the average injection force of the 1.2 g/ml EG (100 vol%)-HAp paste was increased to 43.6 ± 0.817 N. The injection force hikes marked by the arrows in Figure 5.2-10 (b) were more often observed in the 100 vol% EG pastes indicating its deteriorated flowability as compared to the 60 and 80 vol% EG pastes. 100 T T re Q . o o c £2 «J Q. Q. < 10 V 16 0.1 0.01 \ o 0.7 g/ml — v — 0.8 g/ml — • 0.9 g/ml — O " — 1.0 g/ml A 1.1 g/ml 1.2 g/ml \ \ \ \ \ >> \ b —o 20 40 60 80 EG concentration (vol%) 100 Figure 5.2-11. Apparent viscosity change vs. EG vol% in EG-HAp pastes for a constant P/L. Although EG additive generally decreases injection forces of the EG-HAp pastes, there is- a critical concentration of EG beyond which there is no further improvement of flowability of the pastes (and finally the effect reverses for the P/L > 1 g/ml and EG at 100 vol%). EG content up to 80vol % was found to be effective to produce good injectability for the 0.8, 0.9, 1.0, 1.1 and 1.2 g/ml EG-HAp pastes. Inhomogenously mixed pastes (i.e. those 5 8 falling in the region above the dash line, of apparent viscosity larger than 9.00 Pa.s, in Fig. 5.2-11) additionally experienced significant fluctuation of the injection forces and the needle temporarily blocked through "stick-slip" effects. 5.2.3 Microstructural Characterization of E G - H A p Pastes Microstructure of dried EG-HAp pastes was characterized via scanning electron microscope to observe the morphology of HAp particles. Figure 5.2-12 (a) is the microstructure of 1.2 g/ml EG (60 vol%)-HAp paste showing regions of heavily agglomerated HAp particles. It is believed that these crumbs were generated by insufficient amount of EG to make homogeneously mixed paste during the intensive ball milling. On the other hand, the Figure 5.2-12 (b) of the 80 vol% EG paste shows smooth and homogenous surface without HAp crumbs. 10 times higher magnification micrographs of the surface confirm that the HAp particles in the paste are homogenously distributed. Such well dispersed HAp particles in the 80 vol% EG paste seem to let the viscosity of the paste twice lower than that of the 60 vol% EG paste. Figure 5.2-12 (c) shows that the 100 vol% EG paste has somewhat rougher surface, likely caused by agglomerated HAp particles. It is believed that the HAp agglomerates cause increase of this paste's viscosity, i.e. > 3x of the 80 vol% EG paste. However, it should be noted that the microstructure of dried pastes in these SEM pictures does not directly represent the dispersion state of HAp particles in the wet pastes due to the possibility of morphology change during the drying stage. Figure 5.2-12. S E M micrographs of 1.2 g/ml EG-HAp pastes, for (a) 60 vol%, (b) 80 vol% and (c) 100 vol% E G content. 5.2.4 Particle Size Distribution in EG-HAp Pastes Figure 5.2-13 illustrates the results of the particle size distribution measurement for the 1.2 g/ml EG-HAp pastes. The particles of the pastes were broadly ranged from 0.3 to 10 urn. Noticeable difference of the volume fraction and size distribution was not found in the 6 0 particle size range of less than 2.0 urn. The volume fraction between 0.3 and 1.0 um of the 60, 80 and 100 vol% EG pastes was 4.76, 4.66 and 3.90 %, respectively. In the particle size range of 6.0 to 10.0 um, however, the volume fraction of the 60 vol% EG paste was 23.9 vol % which is more than twice of the EG (80 vol%). This higher portion of the coarse particles of the 60 vol % EG paste seems to be generated by unhomogenous mixing and increase of its viscosity. Unlike the other pastes, the 100 vol% EG paste has 2.70 vol% of 10.0 um size. It is considered that the highly coarse particles of 10.0 um are related to the HAp agglomerates seen in the SEM micrograph of Figure 5.2-12 (c). 1.2 g/ml EG-HAp pastes — 60 vol% E G EZZ3 80 vol% E G E 2 100 vol% E G 3.0 25 20 1.5 1.0 0.5 0.0 F(%):D<1.0 60 vol% E G : 4.76 % 80 vol% EG: 4.66 % 100 vol% E G : 3.90 % r< 0 . 2 , - - ' ' ' 0.3 0.4 0.5 0.6 0.7 0,80.91.0 D (um) 2.0 I L L I 3.0 4 .P - - ' 5 .0 6.0 7.0 8 .09.00JO 0.8 0.9 Figure 5.2-13. Particle size distribution analysis for 1.2 g/ml EG-HAp pastes. 6 1 5.2.5 Zeta Potential Measurement for EG-HAp Pastes Zeta potential of the 1.2 g/ml EG-HAp Pastes was measured to investigate the EG content effect on the surface charge change and the isoelectric point (IEP) of the HAp particles in the pastes. Figure 5.2-14 shows that the zeta potential at pH between 6.3 and 6.4 (the pH range of the water-diluted HAp suspension of the pastes) becomes a little more Figure 5.2-14. Zeta potential of HAp particles in diluted 1.2 g/ml EG-HAp pastes in the pH range of 4.0 to 11. positive and the IEP of the HAp particles shifts only marginally from 7.62 to 7.48 as the EG concentration of the pastes increases. It appears therefore that the better dispersion of E G -containing pastes is not related to change of surface potential of HAp particles in the paste. 5.5 Injectability Evaluation and Characterization of TSC-HAp Pastes 5.4.1 Effects of P/L Ratio on Injectability of TSC-HAp Pastes The effects of a P/L ratio on the TSC-HAp pastes flowability were studied with the same method as the EG-HAp pastes in Section 5.2-1. The range of the P/L ratio and the injectability test results of the TSC-HAp pastes are shown in Table 5.3-1. Figure 5.3-2 shows the injection force profiles for the TSC (0.2 wt%)-HAp paste. By increasing the P/L ratio from 0.9 to 1.0 g/ml, the average injection force was increased from 9.73 ± 0.649 to 36.3 ± 0.280 N, and the flowability of the 1.0 g/ml paste became worse with an injection time as Table 5.3-1. P/L ratio range and injectability test results of TSC-HAp pastes. TSC (wt %) P/L (g/ml) HAp content (vol %) Injection force (N) Apparent viscosity (Pa.s) 0.0* 0.7 18.1 5.79 ±0.263 0.945 ±0.0898 0.8 20.2 16.7 ± 1.06 4.65 ±0.363 0.9 22.2 48.5 ± 1.85 15.5 ±0.632 0.2 0.9 22.2 9.73 ± 0.649 2.29 ± 0.222 1.0 24.04 36.3 ±0.280 11.4 ±0.0960 0.4 1.0 24.0 4.98 ±0.563 0.668 ±0.192 1.1 25.8 5.90 ±0.0712 0.983 ± 0.0243 1.2 27.5 63.1 ±2.09 20.5 ±0.714 0.6 1.2 27.5 3.44 ±0.103 0.142 ± 0.0352 1.3 29.1 6.23 ± 0.420 1.09 ±0.143 * The TSC (0.0 wt%)-TSC HAp Pastes are the same as the EG (0 vol%)-HAp Pastes. 100 80 a> o c o •22, 60 20 (a) 0.7 g/ml TSC(0.0 wt%)-HAp paste < 18.2 vol%HAp> 10 20 30 40 50 60 100 80 u o 60 *— c o u .-a> 40 20 (b) 0.8 g/ml TSC(0.0 wt%)-HAp paste < 20.2 vol% HAp > ~ 10 20 30 40 50 60 100 80 0) o 60 c o t ! 40 a> 20 (c) 0.9 g/ml TSC(0.0 wt%)-HAp paste < 22.2 vol% HAp > 10 20 30 40 Displacement (mm) 50 60 Figure 5.3-1. Injection force profiles for TSC (0.0 wt%)-HAp pastes. 6 4 100 80 8 60 £ 40 u 0) 20 (a) 0.9 g/ml TSC(0.2 wt%)-HAp paste < 22.2 vol% HAp > fr 100 80 8 60 (b) 1.0 g/ml TSC(0.2 wt%)-HAp paste < 24.0vol% HAp > 10 20 30 40 50 60 10 20 30 40 50 60 Displacement (mm) Figure 5.3-2. Injection force profiles for TSC (0.2 wt%)-HAp pastes. evidenced by the positive slope formation of Injection force/Displacement graph. In case of the 0.4 wt% TSC pastes in Figure 5.3-3, the irregular injection force fluctuation with the deviation of 22.3 N was observed at the P/L ratio of 1.2 g/ml, which reflects an incompletely mixed paste state. 6 5 100 80 o 60 c o .Si, 'j? 20 (a) 1.0 g/ml TSC(0.4 wt%)-HAp paste < 24.0 vol% HAp > 100 80 8 60 o •5 40 20 (b) 1.1 g/ml TSC(0.4 wt%)-HAp paste < 25.8 vol% HAp > 10 20 30 40 50 60 10 20 30 40 50 60 100 80 o c = 40 u o 20 (c) 1.2 g/ml TSC(0.4 wt%)-HAp paste < 27.5 vol% HAp > 10 20 30 40 Displacement (mm) ^max ^min - 22.3 N 50 60 Figure 5.3-3. Injection force profiles for TSC (0.4 wt%)-HAp pastes. 6 6 100 80 8 60 c o ~ 40 o o 20 (a) 1.2 g/ml TSC(0.6 wt%)-HAp paste < 27.5 vol% HAp > kr i t j i i . . .L.u 100 80 CD 8 60 £ c o S 40 20 (b) 1.3 g/ml TSC(0.6 wt%)-HAp paste < 29.2 vol% HAp > f&A——T- m — i - , —..^..fs.-^—..^^ 10 20 30 40 50 Displacement (mm) 10 20 30 40 50 60 60 Figure 5.3-4. Injection force profiles for TSC (0.6 wt%)-HAp pastes. For 0.6 wt% TSC shown in Figure 5.3-3, the 1.2 and 1.3 g/ml pastes produced very low average injection forces of 3.44 ±0.103 and 6.23 ± 0.420 N, respectively, but homogenous paste formation was not possible at the P/L ratio of 1.4 g/ml. 6 7 The apparent viscosity change of the TSC-HAp pastes vs P/L is shown in Figure 5.3-5. Like the behavior of the EG-HAp pastes, the viscosity of the TSC-HAp paste containing the same TSC concentration increased with a P/L ratio. All the injection profiles considered, the apparent viscosity below 9.00 Pa.s (dash line in Fig. 5.3-5) determines the P/L ratio limitation for homogeneously mixed TSC-HAp Pastes with good flowability. For the TSC (0.0 wt%)-HAp Pastes, 0.8 g/ml is the maximum P/L ratio satisfying this requirement. Similarly, 0.9, 1.1 and 1.3 g/ml are the P/L ratios for 0.2, 0.4 and 0.6 wt% TSC pastes, respectively. Figure 5.3-5. Apparent viscosity change vs. P/L ratio of TSC-HAp pastes for a constant TSC concentration. 6 8 5.3.2 Effects of TSC on Injectability of TSC-HAp Pastes In this study, the TSC was added by 0.2 wt% increments into HAp paste mixed with distilled water (Table 5.3-1). Figure 5.3-6 shows the injection force profiles for the TSC-HAp pastes. In the Figure (a), the flowability improvement of the 0.9 g/ml paste with 0.2 wt% TSC is easily recognized by the decreased injection force of 38.8 N and the disappearance of the injection force fluctuation, compared to the pure HAp paste. By adding 0.4 wt% TSC into 1.0 g/ml paste, a stable injection force of -5.0 N was achieved, i.e. no injection force increase with time like for the TSC (0.2 wt%)-HAp Paste in the Figure (b). The Figure (c) shows that 22.3 N of the injection force fluctuation was generated in the TSC (0.4 wt%)- HAp pastes but another 0.2 wt% more addition of TSC produced very homogenously mixed paste with the low average injection force of 3.43 ±0.103 N. The powerful effect of TSC on decreasing apparent viscosity of HAp paste is also shown in Figure 5.3-7 illustrating the apparent viscosity change of the TSC-HAp pastes as a function of TSC concentration. 0.4 wt% TSC made it possible to process high-solids content, i.e. HAp (1.1 g/ml) paste with low apparent viscosity (0.983 ± 0.0243 Pa.s), comparable to the low solids content 0.7 g/ml TSC(0.0 wt%)-HAp paste (0.945 ± 0.0898 Pa.s). 0.6 wt% TSC addition resulted in only slightly higher apparent viscosity (1.09 ± 0.143 Pa.s) of the 1.3 g/ml paste with 60.7 % HAp loading increase. The required minimum concentration of TSC for producing stably flowable HAp pastes can be determined by again setting the maximum apparent viscosity of the paste at about 9.00 Pa.s (the dash line in Figure 5.3-7), on the basis of the all injection profiles of the TSC-HAp pastes. It appears that 0.4 wt% TSC is the proper concentration for 1.0 and 6 9 100 o o c o u 60 40 20 H (a) 0.9 g/ml TSC-HAp pastes < 22.2 vol% HAp > 10 20 30 40 50 60 100 80 H o 60 c o •g 40 0) 20 (b) 1.0 g/ml TSC-HAp pastes < 24.0 vol% HAp > 0.2 wt% TSC paste _ — 0.4wt%TSC paste 10 20 30 40 50 60 100 80 ^ u 60 S 40 20 (c) 1.2 g/ml TSC-HAp pastes < 27.5 vol% HAp > 0.4 wt% TSC paste 0.6 wt% TSC paste ^ m a x ^ m i n - 22.3 N 10 20 30 40 Displacement (mm) 50 60 gure 5.3-6.Injection force prof i les for T S C - H A p pastes w i t h different T S C concentrations. 7 0 100 -r T (A O U (0 Q_ C o > O- 0.7 g/ml - V - 0.8 g/ml - • - 0.9 g/ml -O 1.0 g/ml -A— 1.1 g/m O 1.2 g/ml - • - 1.3 g/ml 0.01 +-0.0 0.2 0.4 0.6 TSC concentration (wt%) Figure 5.3-7. Apparent viscosity vs. TSC content (wt%) in TSC-HAp pastes for a constant P/L. 1.1 g/ml HAp pastes and 0.6 wt% TSC is adequate for 1.2 and 1.3 g/ml pastes. These values of TSC will be later utilized to decide the TSC range of EG-TSC-HAp pastes for the surface response central composite designed experiments, discussed in Section 5.4.4 5.3.3 Microst ructura l Character izat ion of T S C - H A p Pastes Microstructure of the 1.2 g/ml TSC (0.4 wt%)-HAp and TSC(0.6 wt%)-HAp Pastes was observed through scanning electron microscope. In Figure 5.3-8 (a), the surface of the 0.4 wt% TSC paste shows irregularly distributed HAp particle clusters and its magnification (the right micrograph of the Figure) confirms that the clustered structure is composed of the highly agglomerated HAp particles. On the other hand, the 0.6 wt% TSC paste of the Figure (b) 7 1 shows homogeneously distributed HAp particles, resulting in smooth paste surface. From the SEM observations, it was therefore confirmed that the dispersion and agglomeration of the HAp particles in paste must be controlled for the optimization of flowability. TSC additive provides such control, especially at 0.6 wt% level. However, it should be noted that the microstructure of dried pastes in these SEM pictures does not directly represent the dispersion state of HAp particles in the wet pastes due to the possibility of the evolution of morphology change during the drying stage. — 10 um 'mn. 5 rOQkV xi Ok • SOum.' 1 um Figure 5.3-8. SEM micrographs of 1.2 g/ml TSC-HAp pastes (a) 0.4 wt% and (b) 0.6 wt%. 7 2 5.3.4 Particle Size Distribution in TSC-HAp Pastes The HAp particle size distributions for the 1.2 g/ml TSC (0.4 wt%)-HAp and TSC (0.6 wt%)-HAp pastes are shown in Figure 5.3-9. Unlike the 1.2 g/ml EG-HAp pastes shown in Figure 5.2-13, they displayed a narrower size distribution from 0.2 to 5.0 um, i.e. no HAp particles larger than 5.0 um were detected. It was also found that the 0.6 wt% TSC paste has twice higher volume fraction of the particles < 1.0 urn, as compared to the 0.4 wt% TSC paste, and 6x higher volume fraction of sub-micron particles than the 1.2 g/ml EG-HAp pastes. It is thus independently confirmed that the TSC in the pastes acts as a powerful particle dispersant and effectively prevents HAp particles from agglomeration. 20 10 H 1.2 g/ml TSC-HAp pastes 0.4 wt% TSC 0.6 wt% TSC 0.1 0.3 a. 0.4 0.5 0.6 0.7 O.80A1.O D (um) 2.0 rx 8 3.0 4.0 5.0 6.0 7.0 8.09.00.0 F(%): D<1.0 0.4 wt% TSC: 12.0 % 0.6 wt% TSC: 28.4 % 60 vol% EG: 4.76 % 80 vol% EG: 4.66 % 100vol%EG:3.90% 0.8 0.9 Figure 5.3-9. Particle size and size distribution analysis of 1.2 g/ml TSC-HAp pastes. 5.3.5 Zeta Potential Measurements for TSC-HAp Pastes The surface charge effect of TSC was investigated by measuring the zeta potential of the HAp particles in TSC-HAp paste. Figure5.3-10 shows the measurement result for the 1.2 g/ml TSC (0.4 wt% and 0.6 wt%)-HAp pastes and the 0.8 g/ml TSC (0.0 wt%)-HAp paste indicating the negative shift of zeta potential with increase of TSC content. At pH between 7.4 and 7.6 (i.e. the pH of the diluted TSC-HAp pastes), the zeta potential of TSC (0.4 and 0.6 1.2 g/ml TSC-HAp pastes — ® — (a)0.4wt%TSC (b)0.6wt%TSC — ••— (c) 0.8 g/ml paste without TSC (a) (b) (c) IEP 6.22 5.70 7.57 -30 Figure 5.3-9. Zeta potential of HAp particles in 1.2 g/ml TSC-HAp pastes in the pH range of 4.0 and 11. 7 4 wt%)-HAp pastes was -16.5 mV and -21.4 mV, respectively. The addition of TSC into HAp pastes generated increasingly negative surface charge on the HAp, compared to ~ 0.39 mV of the TSC-free 0.8 g/ml paste, and shift of IEP into more acidic region. Such increased zeta potential of the HAp is believed to arise by an ion-exchange between the phosphate species of H 2 P 0 4 - and the citrate species of cit3 - according to the previous studies of the citrate ion reactions with HAp 7 9 ' 8 0 . The negative surface charge of the HAp particles is believed to make them well dispersed in TSC-HAp paste by the increase of the repulsion energy between the surface modified particles. 5.4 Injectability Evaluation and Optimization of EG-TSC-HAp Pastes 5.4.1 Experimental Design of EG-TSC-HAp Pastes Based on the injectability evaluation of the EG-HAp pastes and the EG-TSC pastes in Section 5.2 and 5.3, the combined effect of EG and TSC admixture on EG-TSC-HAp pastes was studied using designed experiments. A two level factorial response surface central composite design (RS-CCD) used the parameters levels as described in Table 5.4-1 and Figure. 5.4-1 (statistical justification of such choice of the parameters levels and general outline of the design may be found in the reference83). The experimental design and the statistical analysis of injection force and apparent viscosity data were performed using Design-Expert (DX) 5.0.7 software produced by Stat-Ease Inc., Minneapolis, USA. Figure 5.4-1 illustrates the experimental design layout for hypothetical two parameters A, B in coded values (i.e. -1 = low level, 0= mid level and +1 =high level). The resulting response surfaces were obtained through least square fit to experimental data. The plots displayed in Section 5.4 were generated by the DX program. The full experimental design layout and the results Table 5.4-1. The experimental design actual factor levels for EG-TSC-HAp pastes. Variables - a level -1 level + 1 level + a level P/L (g/ml) 1.0 1.06 1.24 1.3 E G (vol %) 40 48.11 71.89 80 T S C (wt %) 0.4 0.44 0.56 0.6 CQ 1 h. O u £ a> = 0 al > TJ 0) TJ 3 -1 • (-1,1) • (0,1.682) (1,1) • (-1.682,0) • (0,0) • (1.682,0) • • • H.-1) (0,-1.682) • (1>-1) 1 -2 -1 1 0 1 2 g Center points # Two level factorial points if Star points Coded value factor A Figure 5.4-1. Experimental factorial design layout represented in coded values. (actual values) are shown in Table 5.4-2. The experimentally determined apparent viscosity values were input into the DX program to analyze the correlation of P/L, TSC concentration and EG concentration with apparent viscosity of the EG-TSC-HAp pastes. In order to produce least-squares fit of the viscosity data as a function of the three independent variables, a natural log transformation was applied because the response (i.e. viscosity) varies by more than two orders of magnitude (from 0.0143 Pa.s to 8.73 Pa.s). That means, a least-squares model of ln(apparent viscosity) was developed. The DX program provides several useful statistical tables that can be used to identify which design model to choose for in-depth statistical study. These statistical details are provided in the Appendix. The final result of this part of the 7 6 work was an empirical model of the variation of viscosity as a function of the three independent variables as follows: LnfApparent viscosity [Pa.s])= -10.2 + 12.8P/I[g/ml] - 9.66TSC [wt%] + 0.0032EG [vol%] Table 5.4-2. The experimental design layout and results in actual values. Std # Run # Block # P/L (g/ml) TSC (wt%) E G (vol%) Injection force (N) Apparent viscosity (Pa.s) 4 1 Block 1 1.24 0.56 48.11 6.51 1.19 3 2 Block 1 1.06 0.56 71.89 3.07 0.0143 6 3 Block 1 1.15 0.50 60.00 5.40 0.810 2 4 • Block 1 1.24 0.44 71.89 19.2 5.52 1 5 Block 1 1.06 0.44 48.11 5.64 0.894 5 6 Block 1 1.15 0.50 60.00 5.68 0.905 7 7 Block 2 1.06 0.44 71.89 4.97 0.666 10 8 Block 2 1.24 0.56 71.89 17.90 5.08 11 9 Block 2 1.15 0.50 60.00 5.85 0.964 12 10 Block 2 1.15 0.50 60.00 5.35 0.794 9 11 Block 2 1.06 0.56 48.11 3.35 0.110 8 12 Block 2 1.24 0.44 48.11 8.09 1.73 17 13 Block 3 1.15 0.50 40.00 5.76 0.934 20 14 Block 3 1.15 0.50 60.00 6.11 1.05 16 15 Block 3 1.15 0.60 60.00 5.22 0.750 13 16 Block 3 1.00 0.50 60.00 4.75 0.588 15 17 Block 3 1.15 0.40 60.00 8.22 1.78 14 18 Block 3 1.30 0.50 60.00 28.6 8.73 19 19 Block 3 1.15 0.50 60.00 5.60 0.880 18 20 Block 3 1.15 0.50 80.00 6.19 1.08 5.4.2 Injectability Evaluation and Characterization of EG-TSC-HAp Pastes Figures 5.4-2,3,4 explore the above empirical model of viscosity. As shown in Figure 5.4-2, the profiles of apparent viscosity variation with P/L ratio are similar. At a constant P/L ratio, apparent viscosity decreased sharply with the TSC concentration but increased minimally with the EG concentration. No significant interaction between TSC and EG content is detected, i.e. the profiles of viscosity vs. TSC are similar for various EG contents. Figure 7 7 5.4-3 shows that the profiles of apparent viscosity variation with EG concentration are similar. At a constant EG, apparent viscosity increased steeply with P/L at the lower level TSC, but slightly at the higher level TSC, indicating the interactive effects between these two variables. The effect of TSC on reducing apparent viscosity was more powerful in the region of the high level P/L rather than the low level P/L. (c) Actual constant: P/L = 1.24 0.56 48.11 Figure 5.4-2. The effect of TSC and EG admixture on apparent viscosity of EG-TSC-HAp pastes at: (a) P/L = 1.06, (b) P/L = 1.15 and (c) P/L = 1.24 7 8 (a) Actual constant: EG = 48.11 (b) Actual constant: EG = 60.00 (c) Actual constant: EG = 71.89 Figure 5.4-3. The effect of P/L and TSC on apparent viscosity of EG-TSC-HAp pastes at: (a) EG = 48.11, (b) EG = 60.00 and (c) EG = 71.89 (a) Actual constant: TSC = 0.44 (b) Actual constant: TSC = 0.50 (c) Actual constant: TSC = 0.56 M ro CL 1.591 ' i 1.232 * in o o 0.873 > 0.514 c 0.155 £ ro Q. Q. < E G (vol%) 1.24 1.19 P/L (g/ml) Figure 5.4-4. The effect of P / L and EG on apparent viscosity of EG-TSC-HAp pastes at: (a) TSC = 0.44, (b) TSC = 0.50 and (c) TSC = 0.56. The profiles of apparent viscosity variation with TSC are similar as shown in Figure 5.4-4, although the values of apparent viscosity are significantly different at different levels of TSC. At a constant TSC, apparent viscosity increased slightly with EG in the overall P/L range, but steeply increased with P / L through the EG concentration. There is no significant interaction between EG and P/L. 8 0 Figure 5.4-5 shows SEM micrographs of the 1.24 g/ml EG-TSC-HAp pastes listed in Table 5.4-2. The higher TSC-content (0.56 wt%) pastes of the Figure (c) and (d) show much better dispersed HAp particles resulting in viscosity decrease, as compared to the pastes of the Figure (a) and (b) respectively (TSC at 0.44 wt%). On the other hand, the higher-EG content paste of the Figure (b) shows agglomerated HAp, likely causing the increase of paste apparent viscosity. The uneven surface indicating HAp agglomerates in the Figure (d) also has more EG than the EG(48.11 vol%)-TSC(0.56 wt%) paste in the Figure (c). However, it should be noted that the microstructure of dried pastes in these SEM pictures does not directly represent the dispersion state of HAp particles in the wet pastes due to the possibility of morphology change during the drying stage. Zeta potential of HAp in higher EG-content diluted paste (at a constant TSC) was found to be less negative for pH between 7.6 and 7.8, Fig. 5.4-6. It is believed that the ethylene groups ( C H 2 - C H 2 ) of EG surrounding HAp particles counteract the ion exchange of the citrate species from TSC, resulting in the particle surface in EG-TSC-HAp paste less negatively charged. This "screening" phenomenon seems to lead more EG added paste to produce less homogeneously dispersed HAp (Figure 5.4-5), to increase its apparent viscosity (Table 5.4-2) and finally to deteriorate its flowability. Figure 5.4-5. SEM micrographs of 1.24 g/ml EG-TSC-HAp pastes; (a) EG(48.11 vol%)-TSC(0.44 wt%), (b) EG(71.89 vol%)-TSC(0.44 wt%), (c) EG(48.11 vol%)-TSC(0.56 wt%) and (d) EG(71.89 vol%)-TSC(0.56 vol%). 8 2 1.24 g/ml E G - T S C - H A p pastes 10 > E .SS o c O Q. ro a> N -10 4 -20 4 -30 J 5 0 -5 -10 -15 -20 • (a) EG(48.11 vol%)-TSC(0.44 wt%) © (b) EG(71.89 vol%)-TSC(0.44 wt%) T (c) EG(48.11 vol%)-TSC(0.56 wt%) 7 ( d ) EG(71.89 vol%)-TSC(0.56 wt%) ~ 1 — 10 (a) (b) (c) w IEP 6.16 6.28 5.82 5.83 II Figure 5.4-6. Zeta potential of 1.2 g/ml EG-TSC-HAp pastes; (a) EG(48.11 vol%)-TSC(0.44' wt%), (b) EG(71.89 vol%)-TSC(0.44 wt%), (c) EG(48.11 vol%)-TSC(0.56 wt%) and (d) EG(71.89 vol%)-TSC(0.56 vol%). Such hindrance of EG on HAp dispersion in EG-TSC-HAp pastes was also observed in particle size distribution analysis of Figure 5.4-7. The EG(71.89 vol%)-TSC(0.44 wt%) and the EG(71.89 vol%)-TSC(0.56 wt%) pastes contained higher volume fraction of coarse HAp 8 3 particles (D > 4.0 um) than the other 48.11 vol % EG pastes shown in Table 5.4-3. However, two to three times higher volume fraction of the fine particles (D < 1.0 um) was detected in the TSC(0.56 wt%) pastes than in the other TSC(0.44 wt%) pastes, representing the strong TSC effect on dispersing the HAp particles. 40 30 20 10 1.24 g/ml EG-TSC-HAp pastes I I (a) EG(48.11 vol%)-TSC(0.44 wt%) I 1 (b) EG(71.89 vol%)-TSC(0.44 wt%) P ^ l (c) EG(48.11 vol%)-TSC(0.56 wt%) f5TV1 (d) EG(71 89 vol%)-TSC(0.56 wt%) 0.1 7.5 5.0 2-5 0.0 0.2 0.3 0.4 0.5 0.6 0.7 0.80.91.0 „ . . . 2 . 0 3 . 0 , - - ^ 0 5.0 6.0 7.0 8.09.010.0 D ( u m ) 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 8 9 10 Figure 5.4-7. Particle size distribution for 1.2 g/ml EG-TSC-HAp pastes; (a) EG(48.11 vol%)-TSC(0.44 wt%), (b) EG(71.89 vol %)-TSC(0.44 wt%), (c) EG(48.11 vol%) TSC(0.56 wt%) and (d) EG(71.89 vol%)-TSC(0.56 vol%). 8 4 Table 5.4-3: Median diameter of HAp particles and volume fraction of D < 1.0 and D > 4.0 in 1.2 g/ml EG-TSC-HAp pastes. EG(vol%)/TSC(wt%) Median D (um) F(%): D < 1.0 F(%): D > 4.0 (a) 48.11/0.44 2 . 4 1 6 . 8 0 1 4 . 0 (b) 71.89/0.44 2 . 4 4 7 . 4 6 2 5 . 0 (c) 48.11/0.56 1.50 1 9 . 5 1 .00 (d) 71.89/0.56 1.89 1 5 . 6 1 8 . 6 5.4.3 Optimization of EG-TSC-HAp Paste - Experimental Verification To enhance clinical efficiency of injectable paste system, it must have good flowability so that the paste can fill complex defect geometry, and volumetric stability so that the injected paste into a body can resist gravitational deformation and washing-off by body fluids. However, opposite rheological characteristics are necessary to achieve the two factors. "Pseudoplastic" behavior of paste is favorable for good flowability since less fluctuation in viscosity is expected at high shear strain rates applied during paste injection through a needle. On the other hand, thixotropic behavior of paste is required for good volumetric stability since its viscosity increases by several orders of magnitude once in rest (zero shear strain rate). Moreover, high solids loading in paste is required for the volumetric stability, but increased solids content generally causes flowability and injectability of the paste to deteriorate. Hence, careful control of the flowability and the stability must be provided for injectable paste for successful clinical application. It is hypothesized in the present work that optimization of the content of surfactants in the paste (such as TSC and EG) may provide such control, and ultimately achieve the right combination of paste flowability and volumetric stability. Based on the analysis of the EG-TSC-HAp pastes viscosity (Section 5.4.2), the optimization of the pastes composition was • ' 8 5 performed through numerical optimization, using features of the DX software program, as described in Section 5.4.3 and the Appendix. The model has been used to search for the conditions (levels of the variables) satisfying the criterion for "optimum" process outcome. The quantitative optimization criterion for the EG-TSC-HAp pastes was set as dynamic (i.e. during injection) viscosity = 9.00 Pa.s and solids content = 1.30 g/ml. 9.00 Pa.s was considered as an absolute criterion because above that threshold injection force, fluctuation Table 5.4-4. Optimziation criteria settings for EG-TSC-HAp paste composition for 1.30 g/ml and 9.00 Pa.s. Name Goal Lower Limit Upper Limit Lower Weight Upper Weight P/L (g/ml) 1.30 1.29 1.31 6.00 6.00 TSC (wt %) 0.40-0.60 0.40 0.60 1.00 1.00 EG (vol %) 40.00-50.00 40.00 50.00 1.00 1.00 Viscosity (Pa.s) 9.00 8.00 10.00 6.00 6.00 and injection force increase with an injection time indicated bad / unstable flowability for both EG-HAp pastes and the TSC-HAp pastes. Table 5.4-4 shows the optimization criteria settings to find a desirable composition for 9.00 Pa.s and 1.30 g/ml EG-TSC-HAp paste. One solution for the searching conditions was suggested by the DX program as shown in Table 5.4-5. In order to verify the results of the empirical modeling and optimization, injection tests with the suggested paste composition, Tab. 5.4-5, were performed. Figure 5.4-9 shows the injection force profiles for 3 tests, indicating a relatively smooth flow behavior of the homogenously mixed paste. The experimental apparent viscosity of the paste was 9.38 ± 0.233 Pa.s, differing by a modest 4.24 % as compared to the model-predicted apparent viscosity of 9.00 Pa.s. It appears therefore that the empirical model predicts the pastes viscosity with reasonable accuracy in the 8-10 Pa.s range of interest. 8 6 Table 5.4-5. A DX suggested composition of the optimized EG-TSC-HAp paste with P/L=\3 g/ml and the predicted viscosity of 9.00 Pa.s; experimental verification of the paste viscosity is included. P/L (g/ml) TSC (wt %) EG (vol %) Experimental Apparent Viscosity (Pa.s) Difference Experiment vs. Model (%) Solution 1.30 0.45 44.3 9.38 ±0.233 4.24 Figure 5.4-8. Injection force profile of the optimized composition EG-TSC-HAp paste, at P/L= 1.30 g/ml, 0.45 wt% TSC and 44.3 vol% EG . 5.5 Injectability Evaluation of PDMS-HAp Pastes Poly(dimethyl siloxane) fluid has been used before as liquid medium for formation of non-setting clinical injectable pastes56,57. It was found that PDMS is efficient in suppressing solidification of paste because PDMS is non-volatile. One example of such paste is Diapex (Diadent, Korea), which is calcium hydroxide mixed with PDMS for temporary tooth root canal filling. This calcium hydroxide-PDMS paste is designated for patients who need certain 8 7 period of time before permanent root canal filling. After endodontic surgery, root end and canal are temporary filled with Diapex, which is later removed and finally another preferred material is placed into the root canal for permanent sealing. Table 5.5-1. Injectability tests results for PDMS-HAp pastes. P/L (g/ml) HAp content (vol%) Injection force (N) Apparent viscosity (Pa.s) 0.2 5.95 5.53 ±0.0529 0.855 ±0.018 0.3 8.67 12.4 ±0.0481 3.189 ± 0.016 0.4 11.2 20.0 ±0.364 5.792 ±0.124 0.5 13.7 48.4 ±0.960 15.49 ±0.328 100 80 8 60 c o = 40 a> (a) 0.2 g/ml PDMS-HAp paste < 5.95 vo l% HAp > 20 100 80 8 60 c o ~ 40 i (c) 0.4 g/ml PDMS-HAp paste < 11.2 vo l% HAp> 20 10 20 30 40 Displacement (mm) 100 80 a> 0 60 o c 1 40 -\ ai 20 (b) 0.3 g/ml PDMS-HAp paste < 8.67 vo l% HAp > r 10 20 30 40 50 60 Displacement (mm) 100 80 (d) 0.5 g/ml PDMS-HAp paste <13.7 vo l% HAp> 50 60 20 30 40 Displacement (mm) 10 20 30 40 50 60 Displacement (mm) 50 60 Figure 5.5-1. Injection force profiles for PDMS-HAp pastes. In this work, injectability of PDMS-HAp paste was studied to understand the effect of PDMS on its flow behavior, by measuring the injection force, as described in Chapter 4. As for the other HAp-based pastes described in the previous Sections, P/L ratio for PDMS-HAp pastes was increased by 0.1 g/ml step, as shown in Table 5.5-1. Figure 5.1-1 represents the injection force profiles for the PDMS-HAp pastes. The 0.2 g/ml paste and the 0.3 g/ml paste were ejected through the needle smoothly with the average injection force of 5.53 ± 0.0529 N and 12.36 ± 0.048 NI, respectively. The injection force fluctuation indicating unhomogenous mixing of the paste is observed the 0.4 g/ml, and more extensively for the 0.5 g/ml pastes. Figure 5.5-2 shows apparent viscosity change of the PDMS-HAp Pastes vs P/L ratio. The viscosity increase was directly proportional to the P/L ratio from 0.2 to 0.4 g/ml, but 0.5 g/ml induced the steep increase of viscosity. 0.2 0.3 0.4 P/L (g/ml) Figure 5.5-2. Apparent viscosity change of PDMS-HAp pastes as a function of P/L ratio. 8 9 The PDMS fluid-based HAp pastes are found to be far more viscous than the water-based pastes, so that solids loading in still injectable paste were relatively low compared to the other liquid medium-based HAp pastes. The maximum P/L ratio of injectable PDMS-HAp pastes was 0.5 g/ml (13.7 vol% of HAp) with 15.5 ± 0.328 Pa.s. In case of the pure DI water-based paste, 0.9 g/ml (22.2 vol% of HAp) was the maximum value with 15.5 ± 0.632 Pa.s. 1.2 g/ml (27.5 vol% of HAp) of the pure EG-based paste produced 13.9 ± 0.279 Pa.s. It seems that high viscosity of PDMS (350 CST) causes viscosity of the PDMS-HAp paste to increase. 9 0 6 C O N C L U S I O N S Based on the work carried out in this study, the following conclusions are drawn: 1. The custom-developed, simple syringe-based viscometer was successfully setup and calibrated to measure injection force and to evaluate flow behavior of injectable paste systems of rather high viscosities, in the range of 5-12.5 Pa.s. This low-cost, easy to setup practical injection testing device has the following advantages: • The flow behavior and injectability of paste can be evaluated by observing an injection force profile. Although in this work an Instron Universal Testing System including Series 9® software program was used, a much simpler system with load cell and data logger can also be used (as determined in early stages of this project). • The apparent viscosity of paste can be calculated with accuracy better than 1% using the modified Hagen-Poiseuill equation with small correction factor resulting from system calibration using standard viscosity fluids. • The syringe viscometer can be used to observe rheological properties change with aging time and therefore can be used to estimate shelf-life of pre-mixed paste contained in a syringe. 2. Ethylene glycol (EG) surfactant mixed with distilled water (DI water) was used as a liquid medium to investigate its effects on rheological properties of HAp Pastes. It was found that: • As the concentration of EG increases, the flowability of EG-HAp pastes generally improves (apparent viscosity decreases), resulting in reduced injection force and less of injection force fluctuation. At powder / liquid (P/L) ratios of 1.1 g/ml and 1.2 g/ml, 9 l however, EG(100 vol%)-HAp pastes have higher apparent viscosity than EG(60 vol%)-HAp pastes. • The critical concentration of EG effective enough to deliver good injectability (apparent viscosity lower than about 9.0 Pa.s) of the pastes is 40, 60, 60, 80 and 80 vol% for the 0.8, 0.9, 1.0, 1.1 and 1.2 g/ml EG-HAp Pastes, respectively. 3. The effect of tri-sodium citrate (TSC) dispersant on rheological properties of HAp paste was investigated, and it was found that: • The citrate species adsorption on HAp particles produces negatively charged surface which induces good dispersion and prevents the particles from agglomeration, resulting in improved flowability (lower apparent viscosity) of TSC-HAp paste. • 0.4 wt % TSC is the minimum concentration required for formation of homogeneously mixed 1.0 and 1.1 g/ml TSC-HAp paste, and 0.6 wt % TSC is for 1.1 and 1.2 g/ml TSC-HAp pastes. • 0.6 wt % TSC addition into 1.3 g/ml paste (of very high 29.2 vol% of HAp) reduces its apparent viscosity to 1.09 ± 0.143 Pa.s, similar to the viscosity of pure 0.7 g/ml paste (0.945 ± 0.0898 Pa.s). 4. The combined effect of EG and TSC admixtures on HAp pastes injectability was experimentally determined, and the composition optimization for the 1.3 g/ml paste was performed using the Response Surface Central Composite Design by the Design Expert software program. The following conclusions result from this part of the work: • Within the range of P/L: 1.0 ~ 1.3 g/ml, EG: 40 ~ 80 vol% and TSC: 0.4 ~ 0.6 wt %, it was observed that the P/L ratio and the TSC concentration have more significant 9 2 effect on the resulting viscosity than the EG concentration. • At high P/L (specifically 1.24 g/ml), increasing EG concentration induced poorer flowability (higher apparent viscosity) of EG-TSC-HAp pastes, likely through screening access of TSC to the surface of HAp. by its counter-effecting dispersing TSC function for the HAp surface modification. • The optimized composition suggested by the DX statistical optimization procedure for the highest solids content 1.3 g/ml EG-TSC-HAp paste with a maximum allowed apparent viscosity of 9.00 Pa.s was at 0.45 wt% TSC and 44.3 vol% EG. The experimentally verified viscosity of the 1.3 g/ml EG (44.3 vol%) and TSC (0.45 wt%) was 9.38 ± 0.233 Pa.s with 4.24 % deviation from the target (model predicted) viscosity of 9.00 Pa.s. 5. The PDMS fluid-based HAp pastes are found to be far more viscous than the other liquid medium-based pastes, so that solid loading for producing injectable paste was very low. The maximum P/L ratio was 0.5 g/ml (13.7 vol% of HAp) with the resulting very high apparent viscosity of 15.5 ± 0.328 Pa.s. 9 3 7 R E C O M M E N D A T I O N S FOR F U T U R E W O R K The following topics are suggested for further investigation based on the research results of this work: • Flow behavior of the HAp-based bioceramic pastes should be studied as a function of shear strain rate. The variation of viscosity as a function of strain rate is an important characteristic of the pastes, and a decrease of viscosity with strain rate ("Bingham behavior") is a desirable property of paste, as it allows application of very fine needles for paste injection into small cavities, e.g. dental root canals. • Investigation of rheological behavior change of the HAp-based bioceramic pastes with time should be explored. The flow and injectability of pastes is known to deteriorate with time, e.g. ageing due to localized solidification (or formation of large agglomerates), resulting in shortening of the paste's shelf-life. • The HAp particle size affects resorption rate of the pastes. Nano-crystalline HAp and biodegradable HAp/ Poly(lactic acid) nanocomposite powders may replace the micron-commercial HAp powder, and the study of these pastes' rheology and ageing behavior should be repeated. • Cell response study to the HAp-based bioceramic pastes should be initiated to determine biological effects of HAp surface modification techniques investigated in this work. 9 4 APPENDIX - DETAILS OF RESPONSE SURFACE OPTIMIZATION METHOD The DX program provides several useful statistical tables that can be used to identify which design model to choose for in-depth statistical study. Table A1 shows the "Sequential Model Sum of Squares" summary which provides a sequential comparison of models showing the statistical significance of adding the model terms to those terms already in the model. A small p-value (Prob > F) indicates that "adding the source terms has improved the model. Table A2 compiles the "Lack-of-fit" tests which diagnose how well each of the full models fit the data. A lack-of-fit error significantly larger than the pure error indicates that something remains in the residuals that can be removed by a more appropriate model. The "Model Summary Statistics" in Table A3 lists other statistics used to compare models. The "Root MSE" estimates the standard deviation of the error in the design. Smaller is better. Both the "R-Squared" and related "Adjusted R-Square" statistics should be close to one for very good models. A value of 1.0 represents the ideal case at which 100 percent of the variation in the observed values can be explained by the chosen model. The "Predicted R-Table A l . "Sequential Model Sum of Squares" summary. Source Sum of Squares DF Mean Square F Value Prob > F Mean 0.134517 1 0.13452 Blocks 1.67282 2 0.83641 Linear 22.2602 3 7.42007 9.9561 0.0009 Quadratic 7.83303 6 1.30551 4.01559 0.0372 Cubic 2.2373 4 0.55933 6.15362 0.0532 Residual 0.363574 4 0.09089 Total 34.5015 20 1.72507 9 5 Table A2. "Lack-of-fit" tests summary. Source Sum of Squares DF Mean Square F Value Prob > F Linear 10.3936 11 0.94487 70.285 0.0025 Quadratic 2.56054 5 0.512108 38.0936 0.0065 Cubic 0.323244 1 0.323244 24.0448 0.0162 Pure Error 0.0403302 3 0.013443 Square" estimates the amount of variation in new data explained by the model (the closer to one the predicted R-Square is, the better the model). The "PRESS" statistics is the "predicted residual sum of squares" which indicates how well the model fits the data. The PRESS for the chosen model should be small relative to the other models under consideration. The DX program uses an arbitrary score to select a design model. "Sequential Model Sum of Squares" score (M) is given as follows: M = 1, if p < 0.05; M = 0.5/p, if p > 0.05 and M = 0, if model is aliased. "Lack of Fit" score (L) is given as follows: L = 1, if p > 0.10 and L= p/0.10, if p < 0.10. The model with a maximum score is selected by using the formula: Score = (M)(L)(R predicted)- For the current series of optimization experiments, the linear model has the maximum score of 9.30E-06 as compared to the square and cubic models of 1.81E-06 and 0 respectively, and therefore it has been used for further statistical analysis and optimization process. Table A3. "Model Summary Statistics" summary. Source Root MSE R-Squared Adjusted R-Squared Predicted R-Squared PRESS Linear 0.863295 0.680863 0.612476 0.192931 26.3864 Quadratic 0.570183 0.920448 0.830953 -0.05283 34.4212 Cubic 0.301486 0.98888 0.952738 -6.84964 256.637 9 6 The DX program also provides an ANOVA (Analysis of Variance) for the selected model, the regression coefficient, standards errors and t-tests. The statistics for the linear model selected in this study follows in Table A4, and the least squares coefficients for the response surface are listed in Table A5. Finally, the model equations of the coded factors and also the actual variable levels are shown as below of Table A5. Table A4. ANOVA for the linear model adopted in this work. Source Sum of Squares DF Mean Square F Value Prob > F Block 1.67282 2 0.836412 Model 22.2602 3 7.42007 9.9561 0.0009 Residual 10.4339 14 0.745279 Lack of Fit 10.3936 11 0.94487 70.285 0.0025 Pure Error 0.0403302 3 0.013443 Cor Total 34.3669 19 Root MSE 0.863295 R-Squared 0.680863 Dep Mean -0.0820111 Adj R-Squared 0.612476 C V . -1052.66 Pred R-Squared 0.192931 PRESS 26.3864 Adeq Precision 9.07 Desire > 4 Table A5. Coefficients for the linear model adopted for process optimization in this study. Factor Coefficient Estimate DF Standard Error t for H 0 Coeff=0 Prob > |t| VIF Intercept -0.116854 1 0.194818 Block 1 -0.347825 2 Block 2 -0.0006016 Block 3 0.348427 A-P/L 1.13969 . <. 1. 0.233606 4.87868 0.0002 1 B-TSC -0.574131 1 0.233606 -2.45769 0.0276 1 C-EG 0.0380768 1 0.233606 0.162996 0.8729 1 9 7 Final Equation in terms of Coded Factors: : Final Equation in terms of Coded Factors: Ln(Apparent viscosity) = i | Ln(Apparent viscosity) = -0.12 j -10.18 +1.14* P/L ; i +12.78 * P/L -0.57* TSC | j -9.66 * TSC +0.038 * EG i ; +3.202E-03 * EG i 9 8 REFERENCES 1. M. Braddock, P. Houston, C. Campbell, P. Ashcroft. "Born again bone: tissue engineering for bone repair," News. Physiol. Sci., 16, 208-13 (2001). 2. M. Bonfiglio, W.S. Jeter. "Immunological responses to bone," Clin. Orthop., 87, 19-27 (1972). 3. A.S. Greenwald, S.D. Boden, V.M. Goldberg, Y. Khan, CT. Laurencin, R.N. Roster. "Bone-graft substitute: facts, fiction, and applications," J. Bone. Jt. Surg., 83, 98-103 (2001). 4. L.L. Hench, R.J. Splinter. "Bonding mechanism at the interface of ceramic prosthesis materials," J. Biomed. Mater. Res., 2[1], 117-141 (1971). 5. D.M. Liu, Q. Yang, T. Troczynski. "Sol-gel hydroxyapatite coatings on stainless steel substrates," Biomaterials, 23, 691-8 (2002). 6. D. Hakimimehr, D.M. Liu, T. "Troczynski. In situ preparation of poly (propylene fumarate)-hydroxyapatite composite," Biomaterials, 26, 7297-303 (2005). 7. J. F. Mano, R. A. Sousa, L. F. Boesel, N. M. Nerves, R. L. Reis. "Bioinert, biodegradable and injectable polymeric matrix composites for hard tissue replacement: state of the art and recent developments," Com. Sci. and Tech., 64, 789-817 (2004) 8. Joel M. White, DDS, Harold Goodis. "In vitro evaluation of a hydroxyapatite root canal system filling material," J. of Endo. 17[11], 561-6 (1991). 9. Brown WE, Chow LC. "A new calcium phosphate, water setting cement in cements research progress," In: Brown PW, editor. Westerville, OH: Amer. Ceram. Soc, 352-79 (1986). 10. M. Vallet-Regi, J. M. Gonalez-Calbet. "Calcium phosphates as substitution of bone tissues," Progress in Solid State Chemistry, in press (2004). 11. M. Csizy, R.E. Buckley, C. Fennell. "Benign calcaneal bone cyst and pathologic fracture-surgical treatment with injectable calcium-phosphate bone cement (Norian (R)): A case 9 9 report," Foot Ankle Int., 22:507-10 (2001). 12. P. Lobenhoffer, T. Gerich, F. Witte, H. Tscherne. "Use of an injectable calcium phosphate bone cement in the treatment of tibial plateau fractures: A prospective study of twenty-six cases with twenty-month mean follw-up," J. Orthop. Trauma, 16, 143-9 (2002) 13. L. Commuzzi, E. Ooms, J.A. Jansen. " Injectable calcium phosphate cement as a filler for bone defects around oral implants: an experimental study in goats," Clin. Oral Implants Res., 13,304-11 (2002). 14. E.J. Stelnicki, D.K. Ousterhout. "Hydroxyapatite pase (BoneSource) used as an onlay implant for supraorbital and malar augmentation," J. Craniofac. Surg. ;8(5):367-72.b (1997). 15. J.J. Chris Arts, N. Verdonshcot, B.W. Schreurs, P. Buma. "The use of a bioresorbable nano-crystalline hydroxyapatite paste in acetabular bone impaction grafting," Biomaterials, 27[7], 1110-8 (2006). 16. K. Yamagishi. "A synthetic enamel for rapid tooth repair," Nature, 433[7028], 819 (2005). 17. I. Khairoun, M.G. Boltong, F.C. M. Driessens, J.A. Planell. "Some factors controlling the injectability of calcium bone cements," J. of Mater. Sci. Mater. Med., 9, 425-8 (1998). 18 M. Markovic, S. Takagi, D. Kim, L.C. Chow, "Effects of particle size and cement liquid composition on calcium phosphate cement properties," J. Dental Res, 76. 2942 (1997). 19. M. Komath, H. K. Varma. "Development of a fully injectable calcium phosphate cement for orthopedic and dental applications," Bull. Mater. Sci., 26[4], 415-22 (2003). 20. L. Leroux, Z. Hatim, M. Freche, J.L. Lacout. "Effects of various adjuvants (lactic acid, glycerol, and chitosan) on the injectability of a calcium phosphate cement," Bone, 25[2], S31-4(1999). 21. G. Baroud, C. Matshushita, M. Samara, L. Beckman, Steffen T. "Influence of oscillatory mixing on the injectability of three acrylic and two calcium-phosphate bone cements for vertebroplasty," J. Biomed. Mater. Res., 68B[1], 105-11 (2004). 1 0 0 22. W.G. Billotte. "Ceramic biomaterials. In: J.B Park and J.D. Bronzino, editor, Biomaterials: principles and applications," CRC press, Boca Raton, (2002) 23. S.F. Hulbert. "The use of alumina and zirconia in surgical implants, In: L.L. Hench, Wilson J, editor, An introduction to bioceramics," Word Scientific, London, 25-40 (1999). 24. R.Z. LeGeros, J.P. LeGeros. "Dense hydroxyapatite, In: Hench LL, J. Wilson J, editor. "An introduction to bioceramics," Word Scientific, London, 139-80 (1999). 25. H.W. Kim, J.C. Knowles, H.E. Kim. "Hydroxyapatite porous scaffold engineered with biological polymer hybrid coating for antibiotic Vancomycin release," J. Mater. Med., 16, 189-95 (2005). 26. L.L. Hench. "Bioactive glass and glass-ceramic, In: J.F. Shackelford, editor, Bioceramics-applications of ceramics and glass materials in medicine," Trans Tech Publication, Switzerland, 37-64 (1999). 27. T. Kokubo "A/W glass-ceramic: processing and properties, In: L.L. Hench, J. Wilson, editor, An introduction to bioceramics," Word Scientific, London, 75-88 (1999). 28. T.M. Keaveny, W.C. Hayes. "Mechanical properties of cortical and trabecular bone, In: B.K Hall, editor. Bone growth," CRC Press, Boca Raton, 285-344 (1993). 29. E.B.W. Griesen, M. Ding, M. Dalstra, T.M.G.J. van Eijden. "Mechanical properties of cancellous bone in the human mandubular condyle are anisotropic," J. Biomech., 34, 799-803 (2001). 30. T. Kokubo, H.M. Kim, M. Kawashita. "Novel bioactive materials with different mechanical properties," Biomaterials, 24:2161-75 (2003). 31. U. Gbureck, J. E. Barralet, K. Spatz, L.M. Grover, R. Thull. "Ionic modification of calcium phosphate cement viscosity. Part I: hypodermic injection and strength improvement of apatitie cement," Biomaterials, 25, 2187-95 (2004). 32. M.A. Rauschmann, A. Thomas. VS. Wichelhaus, D. Elvira, Z. Ludwig, S. Reinhard, A. Volker. "Nanocrystalline hydroxyapatite and calcium sulphate as biodegradable comp 1 0 1 osite carrier material for localdelivery of antibiotics in bone infections," Biomaterials, 26[15], 2677-84 (2005). 33. T. S. B. Narasaraju, D. E. Phebe. "Some physico-chemical aspects of hydroxyapatite," J. of Mater. Sci., 31, 1-24 (1996). 34. G. E. Levit, P. H. Crayton, E. A. Monroe, R. A. Condate, "Forming methods for apatite prosthesis," J. Biomed. Mater. Res., 3, 683-5 (1969). 35. C. A. Beevers, D. B. Mclntyre. "The atomic structure of fluorapatite and its relation to that of tooth and bone mineral," Miner. Mag., 27, 254-9 (1956). 36. J. F. Kay. "Calcium phosphate coatings for dental implants," Dent. Clin. North. Amer., 36, 1-18(1992). 37. K. de Groot, C.P.A.T. Lein, J.D.C. Wolke, J.CM.A. de Bliek-Hogervost."Chemistry of Calcium phosphate bioceramics. In: T. Yamamuro, L.L. Hench, J. Wilson, editors. Handbook of bioactive ceramics, CRC Press, Boca Raton, 3-16 (1990). 38. V.A. Dubok. "Bioceramics-yesterday, today, tomorrow," Powder Metallurgy and metal ceramics, 39[7-8], 381-394 (2000). 39. M. Wang, N.H. Ladizeski, K.E. Tanner, I.M. Ward, W. Bonefiled. "Hydrostatically extruded HAPEX™"," J. Mat. Sci., 35, 1023-30 (2000). 40. M.S. Abu Baker, P. Cheang, K.A. Khor. "Tensile properties and microstructural analysis of spheroidized hydroxyapatite-poly(etheretherketone) biocomposites," Mat. Sci. Eng. A, 34S, 55-63 (2002). 41. Y.H. An, S.K. Woolf, R.J. Friedman. "Pre-clinic in vivo evaluation of orthopaedic bioabsorbable devices," Biomaterials, 21, 2635-52 (2000). 42. M. Vert. "Bioresorbable polymers for temporary therapeutic applications," Angrew Makromol Chem, 166, 155-68 (1989). 43. S.A. Gittens SA, H. Uludag. "Growth factor delivery for bone tissue engineering," J. Drug Target, 9, 407-29 (2001). 1 0 2 44. "Report of the Committee on Definition of the Term Ceramics," J. Amer. Ceram. Soc, 3[7], 526-42 (1920). 45. Y. Fujishiro, K. Takahashi, T. Sato. "Preparation and compressive strength of a-tricalcium phosphate/gelatin gel compositec cement," J. Biomed. Mater. Res., 54, 525-30 (2001). 46. A.M. Moursi, A.V. Winnard, P.L. Winnard, J.J. Lannutti, R.R. Seghi. "Enhanced osteoblast response to a polymethylmethacrylate-hydroxapatite composites," Biomaterials, 23, 133-44 (2002). 47. R.A. Mickiewiez, A.M. Mayes, D. Knaack. "Polymer-calcium phosphate cement composites for bone substitute," J. Biomed. Mater. Res., 61, 581-92 (2002). 48. S.M. Kenny, M. Buggy. "Bone cements and fillers: A review," J. Mater. Sci. Mater, in Medicine, 14, 923-38 (2003). 49. M. Takechi, Y. Miyamoto, K. Ishikawa, T. Toh, T. Yuasa, M. Nagayama, K. Suzuki. "Initial histological evaluation of anti-washout type fast-setting calcium phosphate cement following subcutaneous implantation," Biomaterials, 19:2057-63 (1998). 50. M.P. Ginebra, A. Rilliard, E. Fernandez, C. Elvira, S.J. Roman., J.A. Planell. "Mechanical and rhological improvement of a calcium phosphate cement by the addition of a polymeric drug," J. Biomed. Mater. Res., 57:13-118 (2001). 51. M. Takechi, K. Ishikawa, Y. Miyamoto, M. Nagayama, K. Suzuki. "Tissue response to anti-washout apatite cement using chitosan when implanted in the rat tibia," J. Mater. Sci.: Mater, in Medicine, 12597-602 (2001). 52. B.R. Constantz, I.C. Ison, M.T. Fulmer, R.D. Poser, S.T. Smith, M. Vanwagoner, J. Ross, S.A. Goldstein, J.B. Jupiter, D.I. Rosenthal. "Skeletal repair by in situ formation of the mineral phase of bone," Science, 267, 1796-99 (1995). 53. M. Csizy, R.E. Buckley, C. Fennell. "Benign calcaneal bone cyst and pathologic fracture-surgical treatment with injectable calcium-phosphate bone cement (Norian (R)): A case report," Foot Ankle Int., 22:507-10 (2001). 1 0 3 54. P. Lobenhoffer, T. Gerich, F. Witte, H. Tscherne. "Use of an injectable calcium phosphate bone cement in the treatment of tibial plateau fractures: A prospective study of twenty-six cases with twenty-month mean follw-up," J. Orthop. Trauma, 16, 143-9 (2002) 55. L. Commuzzi, E. Ooms, J.A. Jansen. " Injectable calcium phosphate cement as a filler for bone defects around oral implants: an experimental study in goats," Clin. Oral Implants Res., 13,304-11 (2002). 56. R. Holland, V. de Souza. "Ability of a new calcium hydroxide root canal filling material to induce hard tissue formation," J. Endodontics, 11[12], 535-43 (1985). 57. M.D. Oztan, A. Akman, D. Dalat, "Intracanal placement of calcium hydroxide: A comparison of two different mixtures and caririers," Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod., 94, 93-7 (2002). 58. E.J. Stelnicki, D.K. Ousterhout DK. "Hydroxyapatite paste (BoneSource) used as an onlay implant for supraorbital and malar augmentation," J. Craniofac. Surg., 8[5], 367-72 (1997). 59. M. Bohner and G. Baroud. "Injectability of calcium phosphate pastes," Biomaterials, 26, 1553-63 (2005). 60. J.S. Reed. "Chap. 15: Rheological behavior of slurries and pastes. In: Introduction to the principle of ceramic processing," Alfred University. 61. D.R. Dinger. "Chap. 4: Time-independent rheologies. In: Rheology for ceramists," Dinger Ceramic Consulting Services, Clemson, 2002. 62 D.R. Dinger. "Chap. 14: Viscometers and rheometers. In: Rheology for ceramists," Dinger Ceramic Consulting Services, Clemson, 2002. 63. K. Ishikawa. "Effects of spherical tetracalcium phosphate on injectability and basic properties of apatitic cement," Key Eng. Mater., 240-242, 369-72 (2003). 64. D.R. Dinger. "Chap. 15: Control of suspension rheology. In: Rheology for ceramists," Dinger Ceramic Consulting Services, Clemson, 2002. 1 0 4 65. J.S. Reed. "Chap. 10: Deflocculants and Coagulants. In: Introduction to the principle of ceramic processing," Alfred University. 66. D.R. Dinger. "Chap. 13: Zeta potential analysis. In: Characterization techniques for ceramists," Dinger Ceramic Consulting Services, Clemson, 2002. 67. H. Andrianjatovo, J. Lemaitre. "Effects of polysaccharides on the cement properties in the monocalcium phosphate/b-tricalcium phosphate system," Innov. Tech. Biol. Med. 16S[1], 140-7 (1995). 68. S. Sarda, E. Fernandez, M. Nilsson, M. Balcells, J.A. Planell. "Kinetic study of citric acid influence on calcium phosphate bone cements as water-reducing agent," J. Biomed. Mater. Res., 61[4], 653-9 (2002). 69. L.M. Anabel, G.M. Jaime, and R.C. Rafael. "The role of pH in the adsorption of citrate ions on Hydroxyapatite," J. Colloid, and Inter. Sci., 200, 114-20 (1998). 70. D.N. Misa. "Interaction of some alkali metal citrates with hydroxyapatite ion-exchange adsorption and role of charge balance," Colloids and Surfaces A, 141, 173-9 (1998). 71. A. Baudot and V. Odagescu. "Thermal properties of ethylene glycol aqueous solution," Cryobiology, 48, 283-94 (2004). 72. P.M. Leth, and M. Gregersen. "Ethylene glycol poisoning," Forensic Science International: articles in press, 2005. 73. M. Wusteman, S. Boylan, and D.E. Pegg. "The effect of cooling rate and temperature on the toxicity of ethylene glycol in the rabbit internal carotid artery," Cryobiology, 33, 423-9 (1996). 74. O.V. de Oliveira and L.C.G Freitas. "Molecular dynamics simulation of liquid ethylene glycol and its aqueous solution," J. Moleular Structure, 728, 179-187 (2005). 75. W. Hayduck and V.K. Malik. "Density, viscosity and carbon dioxide solubility and diffusivity in aqueous ethylene glycol solutions," J. Chem. Eng. Data, 16[2], 143-6 (1971). 10 5 76. J.H. Park, Y.H. Bae. " Huydrogels based on poly(ethylene oxide) and poly(tetramethylene oxide) or poly(dimthyl siloxane): synthesis, characterization, in vitro protein adsorption and platelet adhesion," Biomaterials, 23[8], 1797-1808 (2002). 77. J.B. Park and R.S. Lakes. "Biomaterials-An introduction, 2nd ED., Plenum Press, New York (1992). 78. N. Ignjatovic, J. Jovanovic, E. Suljovrujic and D. Uskokovic, "Injectable polydimethylsiloxane-hydroxyapatite composite cement," Bio-Med. Mater. Eng., 13, 401-10 (2003). 79. H. Maeda, T. Kasuga and L. L. Hench. "Preparation of poly(L-lactic)-polysiloxane-calcium carbonate hybrid membranes for guided bone regeneration," Biomaterials, In press. 80. Q. Chen, F. Miyaji, T. Kokubo, T. Nakaumra. "Apatite formation on PDMS-modified CaO-Si02-Ti02 hybrids prepared by sol-gel process," Biomaterial, 20, 1127-32 (1999). 81. American Society for Testing and Materials. "Standard test method for determination of properties of polymeric materials by means of a capillary rheometer," D 3835-02, 481-91. 82. R.A.L. "A historical perspective and review of elastomer stick-slip," Rubber Chem. Tech. 73[3], 486-503 (2000). 83. D.C. Montgomery. Chap. 16 Response Surface Methods and Designs. In: Design and analysis of experiments John Wiley&Sons Inc. (1991). 84. X. Wang, J. Ye, H. Wang. "Effects of additives on the rheological properties and injectability of a calcium phosphate one substitute material," J. Biomed. Mater. Res. B Appl. Biomater., 259-64 (2006) 85. J.Luo, R. Borgens, R.Y. Shi. "Polyethylenec glycol improves function and reduces oxidative stress in synaptosomal preparations following spinal cord injury," J. Neuro., 21 [8], 994-1007 (2004). 86. V. Good, M. Ries, R.L. Barrack, K. Widding, G. Hunter, D. Heuer. "Reduced wear with oxidized zirconium femoral heads," J. Bone Joint Surg., 85, 105-11 (2003). 10 6 

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.831.1-0078477/manifest

Comment

Related Items