NOTE TO USERS

The original manuscript received by UMI contains pages with indistinct print. Pages were microfilmed as received.

This reproduction is the best copy available

UMI

Synthesis and Characterization of Bone Analogue Materials

Ivana Soten

A thesis submitted in confonnity wîth the requirements for the degree of Master of Science

Department of Chemistry University of Toronto

@Copyright by Ivana Soten 1998

National Library Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services seMces bibliographiques

395 Wellington Street 395, nie Weilingtm Oroiwa ON K I A ON4 OttawaON K 1 A W Canada Canada

The author has granted a non- L'auteur a accordé une licence non exclusive licence dowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, 10- distncbute or sell reproduire, prêter, distriibuer ou copies of this thesis in microfoxm, vendre des copies de cette thès- @ sous paper or electronic formats. la forme de microfiche/film, de

reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts ftom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.

Synthetic analogues of bone are king actively pumied as materials for biomedical

applications in the field of bone replacement, augmentation and repair. Numerous stringent

criteria have to be met for a biomateriai to be considered as an acceptable bone implant,

including the ability to integrate into bone and not cause any deleterious side effects.

Different approaches to a bone implant materials are introduced.

A materials chemistry approach to synthesizing a new type of bone analogue material

is described. The strategy uivolves the spontaneous growtfi, under aqueous physiological pH

conditions, of an oriented hydroxyapatite film with micron dimension porosity, on the

surface of a layer of TiO, that has been sputter deposited on Ti metal. This procedure creates -

desirable a>-crystallized phases of hydroxyapatite (OHM) and octacalcium phosphate (OCP)

with prefened orientation respectively dong the [ûûl] and [101] directions. In an attempt to

mimic the hierarchical organic-inorganic composite construction of bone, a calcium

dodecylphosphate ester mesolamellar phase has been integrated into these oriented porous

films to create a CaDDP-OH.-OCP-Ti0,-Ti multilayer architechue. The nucleation and - growth processes, together with the characterization of these films, are investigated using a

multi-analytical approach based upon PXRD, SEM, TEM, SAED, XPS, FT-IR microscopy

and profiIornetry. The relationship of the synthetic calcium phosphate composite materials to

boae as well as other bone analogue materials is discussed.

1 wish to thank my supe~sor, Professor Geoffrey k Ozin for his encouragement,

enthusiasm, and belief in my abilities. 1 am t d y gratefui for his scientific guidance as weli as

his Wendship and support with al i the problems 1 had.

1 am thankful for technical assistance and stimdating discussions with Dr. Neil

Coombs, Dr. Douglas D. Perovic, Dr. Aieksandra Perovic and Dr. John Davies with various

aspects of the elecwn microscopy and electron diffraction midies, as well as the bone

biology relevant to this materials chemistry approach to the bone implant problem. The

expert advice of Dr. Srebri Petrov with the powder X-ray ciiffiaction measurements is also

gratefully acknowledged. The technical assistance of Ms. Sue Mamiche-Afara with the

sputter coating is deeply appreciated. 1 also wish to th& Dr. h e r Dag for FT-IR

microscopy, optical proflomehy results and most vaiuable discussions, and Dr. Deepa

Khushalani for the help with prwfreading the thesis. The expert direction of Dr. Natasha

Varaksa in the early stages of this projea proved to be extremely helpful.

Finally, 1 wish to thank my parents and my husband for their love and support.

TABLE OF CONTENTS

Abstract

Acknowieùgments

Table of Contents

List of Fi ires

C-R 1: General introduction

1 .O Thesis Org-tion

1.1 General Principles of BiomineralUation

1.2 Molecular Structure of Bone

1.2.1 Organic Bone Phase

1.2.2 Inorganic Bone Phase

1 -3. Biochemical Process of Bone Mineralization

1.4. Biomimetic Approach to Matenais Chemistry

1.4.1 Bone implant Materials

1 S. References

CHAPTER 2: Experimental Methods

2.1. Synthesis

2.1.1 Preparation of Calcium Phosphate on TiO, Surface

2.1.2. Preparation of CaDDPICaP Composite Material on the Surface

of TiO,

2.2 Instrumentation

2.2.1 Spuner Coating

2.22 Powder X-ray Diffraction

2.2.3 S d n g Electron Microscopy

2.2.4 X-ray Photoelectron Spectroscopy

2.2.5 Surface Roughness tester

2.2.6 Ion-Beam Milling

22.7 Transmission Eiectron Microscopy

2.2.8 FT-IR Microscopy

CHAPTER 3: Calcium Phosphate Based Bone Analogue Materials

3.1 Preamble

3.2 Background

3.2.1 Surfactant Science

3.3 Chemistry of Octacalcium Phosphate

3.4 Synthesis

35 Results and Discussion

35.1 Calcium Phosphate Growth on TiO, - Surface

35.2 CaDDP-CaP Composite on TiO, - Surface

3.6 Conclusion

3.7 References

Figure 1.1. Gibbs free energy of a cluster in relation to its size

Figure 1.2. SEM of the red abalone shell

Figure 13. Drawing of the hierarchical structure of bone

Figure 1.4. Structure of marine algae Emilianea Huxleyi

Fire 1.5. Drawing of the collagen sûucture

mre 1.6. Hydroxyapatite and fluoroapatite crystal structure

Figure 1.7. Schematic representation of boue mioerdhtion

Fire 2.1. Thickness of the calcium phosphate film obtained by RST technique

Fi ire 2.2. Illustration of an ion-beam milling apparatus

Figure 3.1. Critical micelle concentration, solubility and Krafft-point definition

F i i r e 3.2. Models of lyotropic liquid crystalline phases

Figure 33. Crystal structure of octacalcium phosphate

F i i r e 3.4. SEM images of samples after 7,22,30,48 and 72 hours growth tirne

Figure 35. TEM of sample after 22 hours growth t h e

Figure 3.6. PXRD patterns of sample on and scraped from the subsnate

Figure 3.7. Graph of thichess of the calcium phosphate film versus growth time

Figure 3.8. FT-IR microscopy spectra of samples after various growth time

Figure 3.9. Selected area diEraction patterns of calcium phosphate Nm

Figure 3.10. Representation of clean and hydroxylated surfaces of rutile

Figure 3.11. Oxygen 1s W S spectra of substrate prior and after growth periods

Figure 3.12. Calcium 2p XPS spectra of samples &ter different growth periods

Figure 3.13. PXRD patterns of Ti-Ti0,-OHAp - and Ti-Ti0;OHAp-CaDDP - F i 3.14. SEM of Ti-Ti0,-OHApCaDDP - F i r e 3.15. TEM of Ti-Ti0,-OHAp-CaDDP - F i 3.16. Illustration of the structure of Ti-Ti0,-OHAp-CaDDP -

Chapter One General Introduction

1.0 Thesis Organization

In this work a materiais chemistry approach to synthesizhg a new type of boae

analogue material is described. Formation of a synthetically prepared calcium phosphate-

calcium dodecylphosphate ester composite material is discussed as weU as the relationship of

the composite material to the hierarchical structure of bone.

Chapter 1 focuses on the principles of the biomineraluation process with an emphasis

on bone structure and bone formation. A materials chemistry approach to biomirnicry and

different bone implant materials is also discussed.

Chapter 2 presents the experimental details of the work presented as well as al1 the

characterization techniques employed.

Chapter 3 discusses the growth process of hydroxyapatite on a TiOl surface, the role

of the precursor phase octacalcium phosphate, substrate and solution conditions. The

formation of a synthetic bone analogue material is presented and the mode1 of interfaciai

mmplementarity between the inorganic and organic phase in the composite material is

proposed.

1.1 General Principles of Biominembation

Biomineralization is a process by which organisms form inorganic minerals via

intriate cellular activities. These activities provide the necessary conditions for mineral

nucleation and crystal growth.' The majority of the biominerals have calcium and iron as the

most common cations while phosphates are the most comrnon anions, followed by oxides

and carbonates, Table 1.1. The principal skeletal biominerals are calcium carbonates, calcium

1-2

Table 1.1. Types and functions of the main inorganic solids found in biologicai systemsl

phosphates and silica. The strongest carbonate and phosphate skeletons are well ordered

composites of organic polymer intenpersed with a mineral phase. A second and significant

hinction of biomineral deposits is that they act as a storage system from which ions may be

withdrawn during periods of physiological demand. It should be noted that crystais h a h g

unusual morphologies (fomed by living organisms) have been impossible to mimic in the

labarotory. Recently, however, several reports of synthetically prepared unusuai, nature-like

morphologies have emerged. Four main stages of biomineralization are most often

iatroduced in the literature'"

1. Preorganiration of organic components

2. InterfaCid molecular recognition

3. Vectorial (chernical) regdation

4. Cellular processing

1. Preorganization of organic components

This stage is considered to always ocnir f k t in ail known organisms. Preorganized

environment serves to control the formation of inorganic materials. Nature uses self-

assembly of enclosed protein cages and lipid vesicles or extended protein-polysaccharide

networks for confinement and physical shaping of the mineralization zone. VesicIes and

membranes are predominantly used in intracellular biornineralization while extended

networks dominate in extracellular biomineralization. The process is often confined to the

nanoscale but could be at the micron level as well, for example, in the preorganization of

polymeric collagen (elaborated in 1.2.1 .)

2. Interfacial moleeular recognition

The second stage involves controlled nucleation of inorganic clusters. The

aforementioned preorganized organic phase has specinc sites on the surface that serve as

blueprints for site-directed inorganic nucleation. Assembly of mineral nuclei is governed by

electrostatic, structwal and stereochemical complementarity at the organic-inorganic

interface. The role of the organic surface is lowering of the activation energy of nucleation,

AG*. In addition to lowering the activation energy through the presence of heterogeneous

surfaces in the medium, organisms mntrol the local supersaturation levels. Nuclei will grow

into stable structures only if the energy released through the formation of bonds in the solid

Figurel.1. Gibbs fiee energy of a cluster in relation to ia size (r). a is the surface free energy, AG,, is the free energy change per mole associated with liquid-solid phase change, V, is the cluster molar volume, AG, is the free energy change of the cluster and r' is the critical radius of the cluster.'

Figure 1.2. SEM of the red abalone sheil exhibithg both the prismatic @) and nacre (n) section fomed from calcite and aragonite, re~pectively.~

state is greater than the energy required to maintain newly formed solid-liquid interface,

Figure 1.1.' Process of nucleation of inorganic crystals is considered as a transition state with

nuclei of different structure and orientation comprising a series of activated clusters that

differ in AG'. The fiequency by which they transform into stable entities depends on the

reaction trajecîory which is determined by molecular recognition process. Specific crystal

faces c m be preferentiaily nucleated by interfacial organic-inorganic complementarity thus

stabilizing the transition state.

3. Vectorid (chernical) regalation

This process is associated with crystal growth and termination. In the absence of

cellular intervention, crystals would grow within an organic host accordhg to the laws of

crystalluation. The resulting particles would be constrained in size but 4 t h nomal

morphology. As mentioned above, nature is able to produce various kinds of unique

morphologies. For instance, in the red abalone shell, calcite and aragonite polyrnorphs are

deposited at Mereut locations within the same material, Figure 1.2.' Chernical regulation

involves transport anaor reaction-rnediated processes. Transport mechanisms regulate levels

of supersaturation through facilitated ion flux, complexation-decumplexation switches, local

redox and pH modifications.' Reaction mechanisms influence the kinetics of surface-

mediated processes, such as cluster formation and expression of crystal habit. Control of

biomineralization at this level results in structural, compositional and morphological

specificity in the biomineral phase.

The f i a l stage in the process of biomineralization is the production of higher order

architectures, for example as found in bone, Figure 1.3: Bone architecture is built by starting

with the assembly of collagen and hydroxyapatite that further assemble into microfibrils and

then into osteons. At this level of hierarchy, cells and blood vessels are incorporated into

structure and by association of osteons the final structure of bone is obtained. It should be

noted that all of the stages of biomineralization mentioned are oontrolled at the cellular level,

that is, cells produce organic matrices initially, which assembles into various foms and exert

control over the growth of inorganic materials. Fuither cellular activities are responsible for

the production of mmplex, berarchical structures with unique shapes and functions.

Flgure 13. Drawing of the hierarchical structure of bone'

One of Natures many extraordinary structures that displays the different stages of the

biornineralizatioa process is the marine alga Emilianea Huxleyi, Figure 1.4.' The single-

celled organism is surrounded by a close-packed arrangement of calcite (CaCO,) scales

(coccoliths) organized dong the outer ce11 wall, Figure 1.4a. Individual units of the coccolith

have a characteristic species-specinc shape, Figure 1.4b. Electron diffraction patterns

recorded at different regions of individual units indicate that the complex shape is exhibited

by a well-ordered single crystal of calcite oriented dong specific crystallographic directions,

Figure 1.4b. The morphology of individual scales at the early stages of growth is tabular with

no curved edges and in addition they are oriented and aligned the same way as a mature

crystals, Figure 1.4~. Enclosed spaces provide control over the extent and location of

nucleation while ion pumps in membranes control the physicochemical conditions. In

particular, the organic surface influences the orientation of a particular face, namely [1210]

by means of epitaxial matching at the organic-inorganic interface. Fully formed coccoliths,

comprised of approximately 25 individual scales, are exocytosed nom the cells and assemble

F i i r e 1.4. Marine algae Emilianea Huxleyi: (a) SEM of the organism with calcite scales (coccoliths) organized dong the outer celi wall; (b) shape and crystallopphic directions of an individual scale: (c) SEM of the scales at the initd stage of development'

on the cell surface. Calcite crystals presumably carry on their surfaces organic moieties by

which they are recognized by functionalities in the outer rnembranous surface.

1.2 MoIecular Structure of Bone

12.1 Organic Bone Phase

Bone coasists of roughly 70% mineral embedded in an organic matrix consisting

largely of collagen. The collagen is the basic structural fiber in comective tissues and a l l of

tbem have similar structure at the fibrillar level. Differentiation in the hierarchical structure

takes place when these fibrils are arranged in a particular architecture, thus forming a tissue

for a unique function. The collagen matrix of mineralized bone tissues is formed almost

entirely of type 1 collagen which represents 70% of the organic constituents and 90% of

protein in bone. The characteristic feature of collagen is that of a chah which is wound into

triple helix to form rope like structure, roughly 300 MI by 1.5 nm, Figure 1.5.'

The collagen molecule has two unusual features. The a chah is so tightly coiled that

every third amino acid is glycine, the simplest amino acid and the only one without a c'de

chah that could fit into this helix. Also, collagen contains hydroxyproline and hydroxylysine

that are rarely found in other proteins. These hyckoxyl groups help stabilize the triple helix.

Once the collagen fibea are secreted from cells they become strengthened by cross-linking.

Lysine and hydroxylysine residues are deaminated to yield reactive aldehyde groups which

f o m covalent bonds between molecules. Five of these triple helices align longitudinally with

an overlap of approxirnately one quatter of the moiecule to fom a microfibril that is 4 nm in

diameter. Collagen fibril has a 67 nanometer pattern that arises fiom a "staggef' in the

assembly of adjacent molecules. M e r every fourth spacing there is only 32 nm of molecule

left to fit the next stagger. There are, therefore, 35 nm gaps between heads and tails of the

molecule. Electron micrographs of mineralized wmective tissue have shown that bone

crystallites ocnir at regular intervals dong the collagen fibers with the long c-axis oriented

paraIlel to fibriis. Crystals of hydroxyapatite occur in the gaps as weil as around the collagen

molecules."" It is assumed that coilagen promotes nucleation, however, collagen in some

tissues does not calcify. It is possible that some other mammolecules attach to mllagen and

convert it to the mineralized fonn.

Figure 15. Drawing of the coliagen str~cture.~

Table 13. Major noncollagenous proteins in bone'

Table 1 .2~ summarizes the major noncollagenous proteins to be found in bone.

Glycoproteins are providing calcium andor phosphate binding sites while s e d g as a

structural component as well. Osteopontin is being secreted by the cells in the early stages of

bone formation. This protein, together with bone proteoglycans, is highly sulfated. The d e

of osteopontin is prirnarily in the attachment of osteoblastic cells. Although it has calcium

binding sites, the role is primarily in the growth of the inorganic phase rather than in

nucleation which is assigned to bone sialoprotein.

13.2 Inorganic Bone Phase

Mineralized bone consists of approxirnately 70% hydroxyapatite. This is one of

several calcium phosphate phases listed in Table 1.3." The chernical formula of

hydroxyapatite (OHAp) is Ca,o(P04)6(OH)2 and it can crystallize in stoichiometric (s-OHAp,

Ca/P= 1.67) or nonstoichiometric (ns-OHAp, Cap* 1.67) form. Nonstoichiometric apatites

can be calcium nch (Cm1.67) or calcium deficient (CaP4.67). Aside nom

hydroxyapatite, bone consists of carbonated OHAp and fluoroapatite (FAp) in which OH'

ions are exchanged for CO? and Fions, respectively.

Table 13. Different calcium phosphate phasesZS

F o i r e 1.6. Hydroxyapatite crystal structure. (a) oxygen coordination of columnar ca2+ (a) and linking of columns via PO, tetrahedra @). (b) projection of the fluoroapatite structure on to the basal plane (0001)'

The space group for OHAp is P& and for FAp P6Jm. The structure consists of

columns of Ca2+ ions spaced by one-haif of the c-axis parameter dong the three-fold axis.

Each of these CaZ* ions is comected to its neighboring CaZ' ions above and below by three

shared oxygen atoms that lie in the mirror plane; on one side there are three O(1) atoms and

on the other side three O(2) atorns at a distance of -2.4 Each calcium atom is also

coordinated by three M e r oxygens O(3) at a greater distance. Thus, the columnar calcium

is 9-fold coordinated by oxygen atoms, Figure 1.6a: Columns of c ~ Z ' ions with their

coordinathg oxygen atoms are iinked together by PO: tetrahedra in which three oxygen

atoms corne from one column and the fourth from the adjacent column, Figure 1.6a. The

result is a three dimensional network of PO,$ tetrahedra with enmeshed wlumnar Ca2+ ions

and channels passing through it, Figure 1.6b: The remaining ions, OH- and their adjacent

~ a " ions that are required to complete the structure are located in the channels. Calcium ions

fomi triangles rotated by 60" from each other about the c-axes at whose centers OH* (or F)

ions are located. In the FAp structure, fluorine ions are three fold coordinated by ~ a ' ' and al1

four atoms are in the same plane while in OHAp, the OH- groups lie above the plane. The

solubility of FAp is lower than that of OHAp and FAp can be formed by exchange of 08

ions or by dissolution of OHAp followed by direct precipitation of FAp or mixed F-OHAp.

Aside from being the main part of the bone structure, hydroxyapatite forms the enamel of the

teeth.

1 3 Biochemid frocess of Bone Mineralization

Bone is a dynamic tissue that undergoes remodeling throughout life. In the

remodeling process, new bone is laid down on the resorbed surface of the old one. Therefore,

bone tissue creates an interface with itself on a continual basis. There are two major types of

bone cells that are responsible for bone resorption and growth throughout Me. Osteoblasts

are cells that produce new bone and osteoclasts are cells that resorb the old one. In vitro

shidies of bone formation can be perfonned using a substrate immersed in ce11 culture

(assuming that the substrate can successfully play the role of the resorbed bone surface):

The sequence of bone formation starts with secretion of a collagen fiber-free organic

matrix which is cailed the cement line, of which the major components are osteopontin,

calcium, phosphorou and bone sialoprotein. The width of the cernent line is reported to Vary

from 0.2 - 5 p. Minerakation of this matrix occurs by the seeding of nanocrystalline

calcium phosphate. The generai appearance of the calcified globular accretion laid d o m on

the substrate by osteogenic cells are shown in Figure 1.7am6 Proteins in this organic matrix are

believed to provide binding sites for nucleation of calcium phosphate. It is important to

differentiate berneen this cell-mediated and protein dependent biological mineralization

phenomenon and that of spontaneously forming calcium phosphate layes. Early crystals are

1-2 nm in size as determined by field-emission transmission electron microscopy (FETEM).

They grow and form globular ametions approximately 1 pm in size, Figure 1.7b. The

mineralized interfacial layer is succeeded by a collagen fiber assembly where the collagen is

stiil uncalcified, Figure 1.7~. In the extracellular space, hy droxy lated p rocdagen is

assembled into fibrils by the introduction of cross-links. Collagen fiber assembly becomes

evident in the interfacial mineralized zone. This zone acts as an anchoring surface for the

collagen that becomes mineralized by two methods, narnely fiber mineralization and

extracollagenous matrix mineralization. The latter is described by random seeding of

crystallites in the noncollagenous proteins that are continually secreted by osteogenic cells.

These proteins rnay or may not be aîtached to coilagen fibers. Thus, whereas secretion of

coilagen continues, enveloprnent of the collagen fibers by the underlying mineralized matrk

ocnirs, Figure 1.7d.

Initiai organic matrix

Seeding of calcium phosphate crystallites

Clystal growth and collagen assembly - -

I . - p.--

Cdlagen calcification and matrix mineralization

Figure 1.7. Schematic representation of bone mineralization. (a) secretion of noncollagenous matrix. (b) calcium phosphate nucleation at protein active sites. (c) crystal growth and coliagen secretion and assembly. (d) coiiagen and rnatrix minemüzation6

1.4 Biomimetic Approach to Materials Chemistrg

Over the y e m , a huge interest has emerged in mimicking the chernical processes in

Narure. The complex processes observed in biomineralization are characterized by the ability

known as "synthesis with constnictiony' or 4?nolecular tectonics" that serve as inspiration for

innovations in chemistry, in particular inorganic materials che~nistry-'~ Biomimcry could

lead to materials with better mechanical performance and "smart" materials that respond to

changes in their environment.' Biological materials are capable of remodeling and structural

redesign in response to environmental stresses. For example, in bone, more materiai is

deposited in the areas of greater load. This is due to the fact that collagen is piezoelectric so

the osteocytes may be able io detea whether the bone is being bent or being loaded off-

center as might happen if a broken bone does not join properly in line. Interestingly, the very

opposite occurs in a passive inorganic system, for example a loaded spring wiU corrode much

more quickly then a non-loaded one, hence the development of self-correcthg structures is

very desirable.'

The drive for novel adhesives based on proteins used by mussels is driven by the

lack of adhesives that will cure in a wet environment, the poisonous nature of solvent

systems and the need for medical adhesived Soft materials are also of interest in

biomimicry, for medical use (artifïcid skin and artery) and robotics applications. Biological

ceramics, such as bone and nacre are some of the toughest materials known cornpared to

synthetic ceramic materials which are an order of magnitude lower in toughness. The main

problem at present seems to be control of the orientation of the ceramic phase. Therefore,

material scientists are constantly looking at the ways that Nature is able to exert control over

production of such materials.

1.4.1. Bone Implant Materials

Different approaches to the preparation of possible bone implant rnaterials have

been studied extensiveiy."" In certain studies metals or metal ailoys are used directiy as

implants,130r these are coated with bioactive materials, such as hydroxyapatite, to aid in

adhesion to bone." Although metals or meral alloys meet biomechanid requirements of

implants and some achieve measures of biocompatibility, they exhibit poor interfacial

bonding between the metallic surface and the surroundhg bone. Good biocompatibility of

certain metals, for example titanium, is thought to occur as a result of spontaneous formation

of a thin titanium oxide layer on the surface. This in tuni is responsible for the low corrosion

rate. in addition, titanium and titanium based alloys have been shown to be bioactive as

calcium phosphate is readily able to nucleate and form a thin film on these substrate~.'~

However, the fiims obtained in this way are not uniform and grow to a thickness of only ca.

3-4 nm even after a month of aging, thereby reducing their efficacy as bioactive materials. In

vivo experirnents have established that bone does not bind as weU to titanium metal as it does

to hydroxyapatite coated rnetal.'"l6 It has been suggested that if a titanium implant is initially

coated with a layer of hydroxyapatite then the layer has to be at least one millimeter thick

and also exhibit bulk properties in order for the bone to properly adhere to this surface.

Different methods for preparing surface coatings for implants have been explored

such as dip coating, chernical vapor deposition (CVD), electrophoretic deposition, and

plasma spray-coating being the most widely used. However, coating of the intemal cavities

of cornplex-shaped implant materials is not feasible and the stoichiornetry of the coated

phase is hard to contr01.~ Furthemore, porous bioactive coatings exhibit better bone ingrowth

properties and most of the above mentioned techniques are able to produce only dense,

ceramic materials. Lack of strong adhesion between coatings and metal substrates is an

additional problem.

Coilagen molecules as well as synthetic polyrnea have been extensively studied for

their potentiai as bone implants. Most promising materials included coliagen -

glycosaminoglycan copolymer comp~site,'~ poly (lactic-a-glycolic acid) biodegradable

foamdg and bone cements? Bone cements include poly (methyl methacrylate) or a

copolyrner of methyl methacrylate and styrene that are polymerized in situ. In addition,

various biomacromoledar controlled dmg delivery systems were investigated such as geiled

bicontinuous cubic phasesa and hydrogels."

One of the recent most promising approaches to bone implant materials is the use of

composite organic-inorganic materials since they can be designed to be cornpositional,

structurai and functional analogues of bone. Different combinations of materials have been

used, the most well known examples being polymer-hydroxyapatite composites.'l Moreover,

because organic and polymeric structures are used as targeted dmg delivery systems, it is

easy to imagine how composite organic-inorgaaic bone analogue materials could facilitate

bone repair and overcome infection. With this concept in mind it has been suggested that

bone mimicry could be accomplished using a composite of a calcium phosphate-based

mesophase and hydroxyapatite7?

1.5. References

1. K. Simkiss, K M. Wïbur, Biornineralization; Academic Press, hc.: San Diego, 1989

2. S. Mann, J. Webb, R. J. P. Williams (eds.), Biomineralization; VCH Publishers: New

York, 1989

3. M. Sarikaya, 1. A. Aksay (eds.), Biomûnetics; AIP Press: New York, 1995

4. J. Vinant, Strzictzual Biomateriuls; Princeton Univ. Press: New Jersey, 1990

5.1. C. Elliot, Structure and Chemistry of the Apatites and 0th Calcium Urthophospha~s;

Elsevier: Amsterdam, 1994

6. J. E. Davies, The A~torn ica l Record, 1996,245,426445

7. k k Campbell, G. E. Fryxell, J. C. Linehan, G. L. G d , J. Biomed Mater. Res, 1996,

32,111-118

8. P. Ducheyne, J. E Lemons, Bioceramicî: Material Characterizaiions Versus In V i o

Behaviour, Annals of the New York Academy of Science, New York, 1988

9. T. Kokubo, F. Miyaji, H. Kim, J. Am Ceram Soc., 1996,79,1127-1129

10. F. H. Silver, Biumt&is, Medical Daicer and Tissue Engineering; Chapman & HaU:

New York, 1994

1 1. T. Kokubo, Biomater, 1991,12,155-163

12. P. Ducheyne, D. Christiansen, Bioceramics 6, Oxford, U.K, 1993

13. R. Van Noort, J. Mater. Sei., 1987,22,3801-3811

14. K de Groot, R. G. T. Geesink, C. P. A. T. Klein, P. Serekian, J. Biomed Mater- &S.,

1987,21,1375-1387

15. T. Hanawa, M. Ota, Biomaterials, 1991,12,767-774

16. R. G. T. Geesnik, CZinical Orthopedics and Relnted Research, 1990,261,39-58

17. P. Ducheyne et al., J. Biomed Mater. Res., 1980, 14,225237

18. L. K. Louie, 1. V. Yannas, M. Spector, MRS Proceedings, 1993,331, 19-25

19. R. C. Thompson, M. J. Yaszemski, J. M. Powen, k G. Mikos, MRT Proceedings, 1993,

331,3341

20. J. R. de Wi, P. J. van Muilen, Biocompatibility of Clinicol Implant Moterîals; CRC

Press: Florida, 1981

21. S. Puwada, J. Naciri, B. R. Rama, MRS Proceedings, 1993,331,217-223

22. Shun-yan Wu, C. A. Steiner, MRli ProceedUigs, 1993,331,205-21 1

23. Tissue Engineering, M-RS. Bulletin, 1996,Z 1,18-65

24. G. k Ozin, N. Varaksa, N. Coombs, J. E. Davies, D. Perovic, M. Ziliox, J. Mater.

Ch., 1997,7(8), 1601- 1607

Chapter Two

Experimental Methods

2.1. Synthesis

2.1.1. Preparation of Calcium Phosphate on TiO, Surface

Titaaiwn foi1 was cut to a size of 10 x 8 mm, washed with ethanol and air dried.

The resdting clean substrates were sputter wated with titanium oxide using a Ti target in

a mixed oxygen-argon atrnosphere. For optimum results, sputtering was performed for six

houn and the thickness of the oxide layer produced was found to be CU. 900 A using a

quartz crystai microbalance.

The substrates prepared were placed on the bottom of polypropylene bottles in a

100 mi of solution with the following concentrations: 1.6 x 104 M CaCI2 and 7.6 x 10"

M KWO,. The pH of the resulting solution was close to physiological i.e. pH of 7.4, the

solutions were completely clear and no precipitation was observed throughout the

duration period of t&e experiment. The solutions were aged at 37 OC for periods of 7 to

120 hours. After different times, the substrates were removed from the soIutions carefully

washed with deionized water to remove any non-adsorbed electrolytes and dried under

arnbient conditions.

2.1.2. Preparation of CaDDPICaP Composite Material on the Surface of

TiO,

Mono-n-dodecylphosphate was transformed to the water soluble fom,

potassium dodecylphosphate ( W D P ) with a solution of potassium hydroxide in a molar

ratio of KOWDDP = 2/1. The concentration of W D P in water was 0.1 %. The Ti02

subsû-ate with hydroxyapatite grown on it was placed vertically in a beaker containing

&DDP by clamping it with custom-made tweezen. The nibswte was kept in the

solution for haif an hour after which t h e one molar equivalent of 0.01 M solution of

CaCl, was slowly added to form the CaDDP lamellar phase. The rate of addition of CaCI,

solution was lW15 min. The solution was stirred at a constant rate at 50°C for 24 hours.

Substrates were rinsed with copious amounts of water and air dned.

2.2, Instrumentation

2.2.1. Sputter Coating

Sputter coating of the titanium substrate was performed using a Perkin

Elmer 2400 sputtering apparatus. Titanium dioxide was produced using a titanium target

and mixed argon/oxygen amiosphere. Sputtering was performed for different periods of

time at lOOOV DC, using 12 mtorr pressure of 6 sccm argon gas and 6 sccm oxygen gas.

By adjusting the sputtering conditions, the deposited film could be arranged to be

crystalline rutile and/or anatase or amorphous titania. It tums out that the nature of the

titania film did not have a dramatic effea on the structure, orientation or composinon of

the hydroxyapatite overlayers.

23.2. Powder X-ray diffraction (PXRD)

Powder X-ray diffraction (PXRD) data were obtained on a Siemens D5000 diffractometer

using Ni filtered Cu K a radiation (h = 154178 A) and a Kevex 2005-22 solid state

detector. Typicd accelerating conditions were 50 kV and 35 mA. For strongiy diffracting

samples, the intensity of the 100 % peak was fbst checked by quick manuai sans of the

angles in the region of the strongest peak under lower accelerating conditions, typically

40 kV and 20 mk The step size was 0.02' and the step time was 1.0 seconds, with a s a u i

range of 1 to 45" (28). A slit size of 1 mm was used on the X-ray tube, and 1 mm and 0.2

mm for the detector slits. The latter was changed to 0.1 mm for low angle 28 diffraction

( 1 to 7 O ). Samples were prepared by gluing the substrates on the back of the sample

holder using iittle amounts of the silicon paste.

223. Scanning electron microscopy (SEM)

Scanning electron microscopy (SEM) images were obtained on a E O L 840 scanning

electron microscope using an accelerating voltage of 15 kV or les. Substrates were cut to

a small piece and using conducting carbon paste, mounted to the sample holder. AU the

samples were gold coated in order to produce a conducting surface layer and reduce

charging effects which were cornmon for most of the samples.

2.2.4. X-ray photoelectron spectroscopy (XPS)

X-ray photoelectron spectrosmpy P S ) was perfomed using a Laybold MAX 200 W S

apparatus. A non-monochromatized Mg K a X-ray source was used at 15 kV and 20 mA.

Al1 the spectra were collected under a vacuum of l e s than 1 x 1 0 ~ Ton. Satellite line

subtraction was performed as well as correction for charging effects by calibrating the

position of the carbon 1s peak at 285 eV.

23.5. Surface Roughness Tester

Thickness measurements were performed on WYKO RST Surface Roughness

tester. The principle of this optical profilornetric technique is based upon constructive

interference between light that is refiected fiom the surface and a reference beam, and has

a quoted vertical resolution of ca. 1 nrn. The RST is a vertically scanning interference

microscope system that operates with one of seveml interchangeable magnincation

objectives. Each objective contains an interferorneter, consisting of a reference mirror and

beamsplitter. Interference fringes are produced when light reflected off the reference

rnirror combines with light reflected off the sample. When short-coherence-length white

light is used, these interference f i g e s are present only over a very shaUow depth on the

surface. The surface is pronled by scanning v e r t i d y so that each point on the surface

produces an interference signal and than locating the exact vertical position where each

signal reaches its maximum amplitude.

The RST starts the measurement sequerce by focusing above the top of the

surface being profiled and quickly scanning downward At evenly spaced intervals during

the scan, frames of interference data imaged by the video camera are captured and

processed by high-speed digital signal processing hardware. As the system s a n s

downward, an interference signal for each point on the surface is formed. The system

uses a senes of advanced amputer algorithms to precisely locate the peak of the

interference signal for each point on the surface. Each point is thus processed in parallel,

allowing the RST to determine surface height differences quickly and with an ememely

high degree of accuracy.

The software package provided with the instrument allows for full system control,

surface calculation, graphic display and statistical analysis. Analyses that can be

performed include average roughness (Ra), mis roughness (Rq), maximum surface height

(Rp), maximum surface depth (Rv), maximum peak-to-valley height (Rt), average peak-

to-valley height (Rz), while plots include culor contour, 3D isometrk, histograms, dopes

and 2D subandyses. Figure 2.1. shows the obtained profiling measunnent result of a

calcium phosphate sample obtained

Vertical scanning interference mode was used with a tungsten-halogen lamp as a

light source and solid-state CCD detector. Magnifications of 10 times and 20 times were

used which corresponds to 422 x468 mm and 211 x 234 mm , respectively. Vertical

measurement range is 10 nm to 100 p with a resolution of 3 m. Objectives wuld be of

either 2.5X 10X, 20X, and 40X magnification.

Figare 2.1. Thiclaiess of the calcium phosphate gmwn for 4 days obtained by RST technique

2.2.6. Ion-Beam Milling

Cross sectional samples for TEM analysis were obtauied using an ion-beam

milling technique. In ion-beam thinning (milling), gas ions are accelexaîed to an energy

of severai keV and directed onto the specimen unda oblique incidence. Figure 22. is

schematic illustration of a . ion thinning apparatus. Two ion guns are arranged in a

vacuum chamber such that the specirnen is bombardecl h m both sides at the same t h e .

Argon is the gas most commonly used for the ion guns. A glow discharge is ignited when

a high voltage is applied between anode and cathode and a beam of particles is emitted

fiom the hole in the cathode. At a fixed gun voltage, the particle flux is controlled by

changing the flow rate of the gas.

Figure 2.2. Illustration of an ion-beam milling apparatus

To create a symmetrical dimple under ion bombardment, the specimen holder is

rotated about an axis normal to the specimen. Selection of the thinning parameters

represents a compromise between the highest possible thinning rate and an acceptable

damage level due to ion bombardment. Since ion d i n g is more than one order of

magnitude slower than chernical or electrochemicai polishing, the a h is generally to

achieve highest possible thinning rates. The thinning rate is diredy proportionai to the

flux of particles incident on the specimen. In order to maximize this flux, ion guns must

be aligned so that the center of the beam strikes the center of the specimen. Given a

constant flux of parcicles, the sputtering rate increases linearly with ion energy up to

about 10 keV. At high energies, the ions penetrate the specimen more deeply, causing ion

implantation and damage by way of specimen heaMg and irradiation damage. in practice,

smaller angles of incidence (about 10-15") of the beam are used to reduce the damage.

Momentum transfer and hence the sputtering rate are also dependent on the mas ratio

between the gas atoms in the incident beam and the target atoms. Both variables are

maximum when the masses are equai.

23.7. Transmission Electron Microscopy (TEM)

Transmission elecûon microscopy images were obtained on a JEOL 2010F field-

emmision electron microscope operating at 200 kV by Dr. Alehancira Perovic. Minimum

probe diameter was about 0.4 nm. Analytical techniques employed were high resolution

phase contrast irnaging and selected area electron diffraction. The samples were

embedded in a TAAB epoxy ma&, cured at 60 O C for 24 hours and sectioned using an

ion-beam milling technique. The thin sections were then mounted on either carbon

coated, plastic coated or 400 mesh size 3.05 mm copper support grids.

22.8. ET-IR Microscopy

FT-IR Microscopy is one of the primary methods for determining the rnolecular

composition of smdl amounts of material or small areas of large objects. It allows for

choosing an exact part of the sample to be analyzed. In a microscope measurement the IR

radiation is sent through a tiny opening so that the effects of diffraction must be

considered. Spatial resolution is the very important performance criterion and excellence

is achieved in the instrument used in this sîudy by the use of Spectra-Tech IR-Plan

laboratory microscope equip ped wi th Targeting and Redundant Aperturing . The sample

under consideration has two phases the centrai area of interest and a swounding matrix.

A conventional FT-IR microscope lacks a targeting aperture, that is an aperture between

the source and the sample. This results in a large percentage of the IR radiation passing

through the ma& resulting in spectral impdty due to the scattered radiation from the

matrix. In a Spectra-Tech microscope a targeting aperture is employed which restricts the

IR radiation to a much smaller area, greatly reducing scatter from the matrix. The second

redundant aperture further eliminates scatter fiom the matrix.

The Spectra-Tech microscope provides 15 mm of working distance , has a 100

Pm pinhole that determines the area to be analyzed and transmitted or reflected

illumination settings. For d l the sarnples, diffuse reflectance spectroscopy was used.

When infrared radiation is directed ont0 the surface of the solid sample, specular

reflectance or diffuse reflectance can occur. The specular component is the radiation

which reflects directly off the sample surface (it is the energy not absorbed by the

sample). Difhise reflectance is the radiation which penetrates into the sample and then

emerges. A diffuse reflectance accessory is designed so the difhisely reflected energy is

optimized and the specular component minimized. The optics collect the scattered

radiation and direct it to the hfiared detector. Several -ors exert a sigainant ifluence

over bandshape and relative/absolute intensity and these include rehctive index of the

sample ma&, particle size, sample homogeneity and concentration. Reffactive index

effects result in spedar refiectance contributions. The bandwidth decrease and relative

intensities are changed as particle size decreases. Peak intensities are changed if the

sample is not a homogenous and with high concentrations of the sampies, there is a

dramatic increase in spenilar contribution to the spectral data.

Chapter Three

Calcium Phosphate Based Bone Analogue Materials

Synthetic analogues of bone are king actively pursued as materials for biomedical

applications in the field of bone replacement, augmentation and repair, as discussed in

section 1 Al. Numerous stringent criteria have to be met for a biomaterial to be considered as

an acceptable bone implant, including the ability to integrate into bone and not cause any

deleterious side effects. Before discussing possible implant materials and their expected

response in the body environment, one has to take into account not ody the hierarchical

structure of bone but also the mode of bone formation and remodeiing thmughout life.

Kimwledge of the underlying principles of such biornineral reamstructruction chemistry

could lead to the production of novel, self-organizing structures, able to remodel and

optimize in response to the way that they are being used. in particular, a bone analogue

material could be synthesized with new and/or enhanced bioactivity and biocompatibility

characteristics that is capable of succesfil integration and remodeling with bone itself. One

of the most promising approaches to bone implant materials is the use of composite organic-

inorganic materials since they can be designed to be compositional, structural and functional

analogues of bone. Moreover, because organic and polymenc structures are used as a targeted

dmg delivery systems, it is easy to imagine how composite inorganic-organic bone analogue

materials could fatilitate bone repair and overcome infection.

In this chapter, a materials chemistry approach to synthesinng a new type of bone

analogue material is described. The strategy involves the spontaneous growth, under aqueous

physiological pH conditions, of an oriented hydroxyapatite film with micron dimension

porosity, on the surface of a Layer of TiO, that has been sputter deposited on Ti metal. The

results indicate strong adhesion and rapid growth of calcium phosphate on sputtered TiOt

surface within hours, rendering them as a viable option for bone implant matenals. Bath

oriented octacalcium phosphate and hydroxyapatite phases are found to CO-crystallize as a

macroporous nIm indicating an epitaxial growtb process. It is suggested that OCP acts as a

prewsor nucleation phase at the TiO, surface, that hydrolyzes to hydroxyapatite and

continues to grow epitaxially. In an attempt to rnimic the hierarchical organic-inorganic

composite construction of bone, a dc ium dodecylphosphate ester rnesolamellar phase has

been integrated into these oriented porous nIms to create a CaDDP-OHApOCP-TiQ-Ti

multi-layer architechne. A multi-analytical approach is employed to follow the growth

process and characterize the composition and structure of the calcium phosphate based films

on the titania-titanium substrate. The relationship of the spthetic calcium phosphate

composite materials to bone as weli as other bone analogue materials is discussed.

It should be noted that understanding underlying principles of self-assembling

organic structures is a necessary pre-requisite for an appreciation of the hierarchical organic-

inorganic stmctures that are produced either in the laboratory or by Nature. In addition,

possible precursors to the formation of inorganic bone phase have to be taken into account

w hen discussing bone analogue/implant matenals.

33. Background

3.2.1. Surfactant Science

Surfactants or surfaceactive agents are amphiphilic molecules that function

predominantly at the interface between two separate phases. Amphiphilic nature denotes that

they are comprised of two dissimilar groups namely, a lyophobic (solvent repelling) and a

lyophilic (solvent attracting) rnoiety. When water is used as a solvent, these are referred to as

the hydrophobie and hydrophilic groups. Thus, there are two parts with extremely different

solubility properties withui one molede. Depending on the nature of the polar functionality,

surfactants are divided into three groups : cationic, anionic and nonionic. Surfactant behavior

in a solvent is characterized by enhanced adsorption at surfaces/interfaces due to the

distortion of the solvent phase." Lyophobic group causes a distortion of the solvent liquid,

increasing the overail energy of the system. For the present purposes, behavior of surfactant

moledes in water will be discussed. Water molecules form inter-molecular hydrogen bonds

which leads to a teaahedral network with a relatively low density. A large number of ailowed

positions of the hydrogen atoms due to the tetrahedral surroundings maintaincd by hydrogen

bonds is responsible for the large entropy of the phase. The presence of hydrophobic groups

cause water molecules to rearrange and form cage structures around the solute.' This process

is costly in entropy since the water is more ordered. Therefore, surfactants either adsorb at the

surfaces/interfaces or they aggregate into structures such as micelles, liquid crystals or they

crystaliize out from solution. In the case of micelle and liquid crystal formation, water

molecules are repelled h m the hydrophobic part of the amphiphile. This entropy driven

process lowen the energy of the systern and favoa an increase in solubility. It should be

noted that micelles only form above a so called critical micelle concentration, "cmc".

Ionic surfactants in an aqueous solvent dissociate completely at low concentrations.

Without the shielding action of nearby counterions an aggregation of amphiphilic ions wodd

be impossible because of the stmng repulsive electrostatic forces between neighboring

headgroups. In general, the behavior of ionic surfactants is dependent on the length of the

hydrophobic tail, the nature of the head group, the valency of the counterion and the solution

conditions. It is often observed that the solubility of ionic surfactants in water solutions

undergo a sharp, discontinuous increase at some characteristic temperature, referred to as the

K r a temperature, T,. Below that temperature, the solubiiity of the surfactant is determined

by the crystd lanice energy and heat of hydration of the system. This temperature is defined

by the intersection of the crnc and the solubility curves in temperature versw concentration

diagrams, Figure 3.1 .'

The rnicellar structure is the initial aggregate observed upon increasing the

concentration (above crnc). For ionic surfactants, the system has to be above T, as weli. Upon

further increase in concentration of solute, larger aggregate structure are formed such as

cylindncal micelles and liquid crystal phases. The latter can form hexagonal, lamellar and

cubic structures. Starting h m an isotmpic solution of anisometric micelles the first step to

achieve a long-range orientational order is to let certain preferred axes of rod-üke micelles

concentration

3.1. Critical micelle concentration, solubility and definition of Kmfft-point

anange parallel to each other in time, ~igure3.k' The resulting phase is called "nematic".

With increasing order rod-like aggregates cari anange in a two-dimensional hexagonal lattice,

Figure 3.2b.' Then, a long-range positional order occurs in addition to the orientational order

of the rods. If the amphiphile builds up paralle1 anay of aggregates then a lamellar phase may

be fonned, Figure 3.2c.' Aliphatic chahs can be in a disordered liquid-like or solid-like a- trans conformation. The lamellae can be built Born bilayen of monomen and interdigitation

of molecules in the bilayer may occur. Cubic lattices are built of globular micelles possessing

long-range orientational order ( in addition to positional order), Figure 3.2d.'

Geornetric shape constraints imposed by the individual amphiphiles restricts their

packing into aggregates of different structures. Two opposing major forces, acting mainly in

the water-amphiphile interfacial region are considered to govem the self-assernbiy of

amphiphiles into liquid crystal structures. The hydrophobic attraction causes the molecdes to

associate whereas the hydrophilic ionic or stenc repulsion of the headgroups has the opposite

effect of keeping them in contact with water. The former tends to decrease whereas the latter

tends to increase the interfacial area per molecule. These two opposing forces lead to the

Figure 3.2. Models of lyotropic liquid crystalline phases: (a) nematic order of rod-like micelles; @) hexagonal array of aggregates; (c) lamellar amy; (d) cubic armogement of spherical aggregates

concept of an optimal area per headgroup at which the total interaction energy per amphiphile

molecule is minimum. Geometric packing properties of different amphiphiles are expressed

in terms of the 'packhg parameter" VI& where v= chain volume, a,,= head group ma, le=

chain length, which determines the type of aggregate that forms.

33. Chemistry of Octacaicium Phosphate

The presence of HPO:- groups in bone apatites has suggested that octacalcium

phosphate (OCP) is a likely precursor to hydroxyapatite (OHAp) formation in bone. In

fact, spontaneous hydrolysis of OCP has been observed to result in the formation of

nonstoichiometric OHAp (ns-OHAp)." Nonstoichiometric apatite cm exist in three forms:

OCP-OHAp intergrowths, imperfectly hydrolyzed OCP, and OHAp formed directly

without a precursor.' Since the third possibility is not probable under physioiogical

conditions, the presence of nonstoichiometric apatites in the early stage of bone formation

as well as in some mature ones has suggested an OCP as a precursor. OCP serves to

establish the final morphology, composition, solubility, and interfacial energy of apatitic

materials, as well as controlling the nucleation and growth of OHA~: These properties result

from the f a a that OCP and OHAp have similar crystal structures and are able to epitaxially

grow together. As a consequence of this relationship, a prime consideration in the preparation

and performance of a bone implant materid is the selection of a surface that is able to induce

nucleation of OCP.

The structure of OCP is shown in Figure 3.3: Every second layer is the same as in

the hydroxyapatite (OHAp) lattice and the layers sandwiched in between are the "ydrated

layers". OCP and OHAp phases are able to grow together to form interlayered mixtures

without a concomitant large increase in interfacial energy. Interlayered mixtures, depending

on the thickness of the layes c m give two kinds of X-ray diffraction patterns. When the

layers are relatively thick, the two kinds of layen ciifha independently, givhg the

appearence of a physical mixture of the two salts. When the layers are thin and randomly

variable in thichess, the dm peaks of OCP interact with those of OHAp ousing the

3-7

variable in thickness, the d, peaks of OCP interact with those of OHAp causing the

positions of the mmbined peaks to shift with tbe Ca/P ratio of the interlayered ~rystals.~

Figure 33. Projection onto the (001) plane of the structure of OCP; the unit cells of

OCP (lower left) and the apatite (upper right) are shown

In addition, due to the similarity in the crystal structures, the diffraction patterns of the two

phases are very simüar (Appeodix A). Therefore, if the characteristic d,, peak of OCP is

absent (due to preferential orientation, for example) misassignment is often made in that

presence of nonstoichiometric apatite is assumed instead of mixtures of OCP and OHAp.

An important feature of the morphology of OCP is that it alrnost invariably occurs

as platy (100) crystals. In the cases where OCP acts as a precursor for the formation of

apatite, the latter crysmllizes in a plate-like fashion even though hexagonal needles are to be

expected. Therefore, OCP serves as a template for the growth of OHAp. The platy nature of

OCP expresses itself variously in the fom of platelets, elongated blades, nbbons, f h s y

sheers or broad plates, frequently clustered in the f o m of rosettes.

Except under conditions of very high pressures, OCP is thermodynamically less

stable than OHAp and would not f o m were it not for its ability to grow rapidly? This

kinetic advantage is probably the main reason why OCP frequently foms as the initial phase

instead of OHAp and undergoes spontaneous hydrolysis to the more stable phase OH&. As

a matter of experience, it appears that the hydrolysis results in the formation of non-

stoichiometric apatite that contains impurhies and defects." The imperfectly hydrolyzed

OCP *in be divided into two types: crystals formed under conditions where growth of OCP

is not as fast as its hydrolysis so that the product contains mostly defective OHAp and very

little OCP or the crystals formed when growth of OCP is fast in cornparison to the

hydrolysis. In the latter case, a substantial amount of OCP could be present and its

subsequent hydrolysis would produce highly non-stoichiornetric hydroxyapatite. The

following equations, (1) and (2) respectively, represent proposed models for hydrolysis

reactions in the presence and absence of a caZ+ source:

It can be seen nom equation (2) that the hydrolysis process in the absence of calcium ions

involves l o s of phosphate ions fiom the crystal and the =action may cease after

approximately 50% of the phosphate has been released by the crystal. This is an additional to

the effects that result in a formation of a non-stoichiometric apatites.

Refer to Chapter 2, Section 2.1.

3.5. Results and Discussion

3.5.1. Calcium Phosphate Growth on TiO, Surface

Figure 3.4. displays SEM images of samples obtained after 7,22,30 and 48 hours of

growth. During the early stages (7 hours) srna11 1 pm sized aggregates are observed, which

upon increased growth time sente as nucleation sites for subsequent growth of larger plate-

like crystds. The crystal plates appear oriented perpendicular to the substrate and upon

further growth increase in size and coalesce to fom "rose-like" structures. The entire

substrate surface is unifody covered within approximately 30 hours. Lri addition, the

coating does not have the texture of a dense calcium phosphate film but instead displays a

porous structure with an average pore size of approximately 1-4 p. As remarked in Chapter

1, section 1.4.1, a porous apatitic structure is advantageous in certain bone implant situations.

It should be noted that OCP crystds and not OHAp are commonly associated with a plate-

based rosette morphology?

Figure 3.4. SEM images of samples grom after (a) 7 hours; @) 22 hours growth time

Figure 3.4. SEM images of the samples grown aher (c) 30 hours; (d) 48 hours growth tirne

.S. Cross :ts the int

Figure 3.4. (e) SEM image of sample after 72 hours growth t h e

#owth time; das te film

Cross-sectional TEM imaging studies of these films reveded direct contact of the

calcium phosphate crystals with the T i 0 substrate, Figure 35. The crystals appear oriented

with the long c-axes perpendicular to the substrate, consistent with the PXRD and SAED

results (see below). The integrity of the resulting nIms, irrespective of the growth time, was

maintained even &ter exîensive washing and the films have excellent shelf-Me. It required

the application of considerable force with a razor blade to even remove part of the film from

the substrate, indicating an impressive strength of adhesion.

In order to identify the phase(s) formed, PXRD studies were conducted on the

obtained film samples. The resuits are displayed in Figure 3.6. The PXRD pattern in Figure

3.6a is of the film on the Ti02 surface after 72 hours growth t h e . This pattern was observed

after 30 hours and before this period nothing was discemible as the N m was insufficiently

thick to observe a diffraction pattern. The peaks are well indexed to a hydroxyapatite phase

and the high relative intensity of the 002 reflection indicates preferential orientation of the

crystds dong the (001) plane (Appendix A). If the sample contained OCP oriented with the

(001) plane parallel to the surface then the most intense reflection from the (100) set of planes

would not be seen. When the sample was scraped from the surface and ground, a peak at 18.8

A was observed that is assigned to the 100 reflection of OCP, Figure 3.6b. Aside from this

low angle PXRD peak, the remaining peaks are indexable to a hydroxyapatite and oot an

octacalcium phosphate phase. From these results it can be concluded that the material which

forms on the TiO, surface most likely consists of an integrown mixture of OCP and OHAp,

the latter being the predominant phase and the former a minor one.

The thickness of the calcium phosphate films grown on the TiOl surface as a

funaion of growth M i e is s h o w in Figure 3.7. The values were obtained by surface opticd

profïlometry, (refer to Chapter 2, Section 225). During the first 22 hours, the surface is not

completely covered and the growth rate is determined primarily by the slow nucleation step.

Upon increasing the growth t h e , the rate appears to linearly increase and suggests that the

growth rate is no longer controlled by the nucleation rate on the substrate but is instead

18-98 A = (100) OCP 825 A = (100) OHAp 3.43 A = (002) O H A ~ 3.17A=(102)OHAp 2.81 A = (21 1) OHAp 2.77 A = (1 12) 0HAp 2.71 A = (300) 0HAp

Figure 3.6. PXRD patterns of (a) film on the TiOz surface after 72 hours growth tirne; (b) film scraped fiom the substrate and ground into randody oriented powder.

O 20 40 60 80 100 1 20 140

time (hours)

Figure 3.7. Thickness of calcium phosphate films as a function of growth time obtained by surface pronlometry; emor bars too srnall to be seen

dependent upon the production of the earlier forrned hydroxyapatite crystals.

In order to further elucidate the role of OCP and OHAp in the formation of these

films, time dependent Fï-IR microscopy studies were performed. It should be noted that the

FT-IR spectra of hydroxyapatite have two ranges of interest: v, (680-710 cm-')' v, (1000-

1200 cm").' Slight changes in the phosphate ion environment are readily observed by the

splitting of the degeneracy of the v, vibration." This region can be further subdivided into

two components: 1000-1085 and 1085-1200 an-' where the lower mmponent is generally

more intense. In stoichiometric apatites (s-OHAp) vibrations anse from symmetric v, and

antisymmetric v, P-O stretching modes in the region 950-1200 cm-' and v, antisymmeaic P-

O bending modes at 680-710 cm". In nonstoichiometric apatites (ns-OHM), the presence of

HPo,'' groups and vacaacies distort the pattern of P-O stretching modes. These give rise to

additional bands which have been extensively assigned and tabulated in the literat~re."~'~

- 7 hours - 22 hours - 30 hours - 48 hours - 72 hours 96 hours

8 0 0 900 1000 1100 1200 1 3 0 0

W ave numbers (cm-')

Figure 3.8. FT-IR mimscopy spectra of calcium phosphate films after different growth t h e

In addition, the OCP phase gives rise to IR spectra that are quite distinct h m s-OHAp and

ns-OHAp." The presence of bands at 830, 910 and 1200 cm-' are diagnostic of the OCP

phase.

Figure 3.8 displays the FT-IR microscopy spectra of the nIms grown on titania. It

can be seen that after 7 hours, characteristic bands of ~0,'- ions are present in the sample.

This signals the presence of a calcium deficient ns-OHAp phase. in addition, characteristic

peaks from OCP are also present, specifîcally at 867 an", 917 cm-' and 1200 cm-'. These

bands were however, only observed after 48 hours of growth time. It shouid be noted that

under the synthesis conditions used to grow these nIms, OCP is commody believed to be a

precursor to OHAp. The observance of OCP only after 48 hours implies that at earlier h e s

in the growth process, the amount of OCP present was insufncient to be detected by this FT-

IR microsmpy technique. It may be suggested that the absence of OCP is likely due to i a

hydrolysis to OHAp. As was mentioned for Figure 3.7, initially, the growth rate of the film is

slow. During this induction t h e OCP is able to convert to OHAp and hence rnay account for

it not being detected by FT-IR. However, once the growth rate increases (after 30 hours), the

rate of hydrolysis of OCP to OHAp is comparably slower and there exists the likelihood of

formuig an intergrown OCP/OHAp mixture a s recognized by the presence and subsequent

increase of OCP P-O vibrational modes in the FT'-IR specaum. This proposa1 receives

support from selected area electron diffraction studies of the films as described below.

Electron diffraction (ED) patterns obtained are displayed in Figure 3.9. These were

obtained using a small aperture size in order to ensure that the cross sectional area examined

belonged to a selected small region of one crystal and was approximately 0.1 pm in size. This

was also done to prevent contributions from streak and ring distortions of the ED patterns

arising from the expected mosaicity of the samples. The patterns obtained were indexed to

the hexagonal lanice of hydroxyapatite viewed dong the 4210, and c1213> zones, Figure

3.9a. Electron diffraction of the triclinic lattice down the OCP <010> zone was also

observed, Figure 3.9b. It should be noted that on scanning numerous areas of the films, the

F i r e 39. Selected Area Electron Diffraction (SAED) of (a) <1210> zone of hydroxyapatite; (b) <01b zone of octacalcium phosphate

rnajority of the patterns obtained belonged to the OHA. phase and the minonty were for

the OCP phase. The zones observed mnfïmed the presence of preferred orientation of both

phases. This is because the electron beam was parallel to the substrate which suggested that

the zones observed contained ody the information regarding planes that were parallel to the

sub~trate. Therefore, the (001) and (101) planes for both OHAp and OCP phases are parael

to the substrate as they are contained within the zones observed (Appendix B). In addition,

the OCP electron diffraction patterns were obtained from crystals mostly on the edges and

top parts of the hyàroxyapatite crystals that were the furthest away fiom the substrate. This

further confirmed the m'-IR resdts that the presence of OCP could only be detected after

prolonged growth times.

In order to further determine the interaction between the substrate and the calcium

phosphate film, W S studies were performed. Table 3.1 lists the atornic percentages and

binding energies of the representative elements. The presence of both calcium and

phosphorus was confinned after only one hour of deposition t h e with a ratio of Ca/P-111.

This can be explained by the fact that at pH = 7.5, the surface of Ti02 is known to be

hydroxylated and Br@nsted acid and base hydroxyl sites exist at the surface that cm aa as

nucleation/anchoring centen for calcium and phosphate ions.

An illustration of a clean and hydroxylated (110) Ti02 surface" is shown in Figure

3.10. Such a TiO, surface has 2- and 3-coordinate bridging oxygens and 5- and 6-coordinate

titanium. M e r exposure to water the surface becumes hydroxylated and two different types

of OH groups exist on the surface. One is denoted as acidic, the other as basic and these are

distinguishable by means of XPS as seen in Figure 3.11a and wnfirmed by reference to the

published ~iterature.'~"

\

-

4

6 Cri in

6 - m iA

- * Cr)

2 - 2 TP Cc,

Clean Surface

2-fold 3-fold 6-fold Ti bndging O bridging O 1 5-fold Ti

I I

hydroxyl groups hydroxyl groups i l

Hydro~lated Surface

acidic basic

Figure 3.10. Representation of clem and hydroxylated (110) surfaces of rutile

Binding Energy (eV) 7 Hours

Binding Energy (eV)

L -535 -530 -525

Bmding Energy (eV) -536 -532 -538 -524 -520

Binding Energy (eV)

Figure 3.11. XPS spectra of O 1s (a) TiO, sinface before immersion in calcium phosphate solution; (b) T i 4 sinface at different growth tunes

Figure 3.12. XPS spectra of Ca 2p core electrons after different growth periods

Quantitative WS studies of the nucleation and growth of these films shows that the atomic

concentrations of Ca and P do not change in the first three hours after which they increase

slightly with the . The h t 3 hours can therefore be considered as an induction-nucleation

period, followed for the next 30 hours by a nucleation-crystal growth period. Note the

disappearance of the titanium signal (Table 3.1.) indicahg cornplete average.

The oxygen 1s binding energy spectra are shown in Figure 3.11b. During the first

three hours, the ratio of the bulk to surface oxygen signal decreases due to the higher degree

of hydration of TiOl in the aqueous solution and du, because of the presence of phosphate

ions. Also after this tirne differential charging of the substrate and film is apparent. There are

two sets of signals ffom equivalent atorns, one fiom atoms close to the surface that are not

charging and the other from atoms hirther from the surface that are charging. Signals from

oxygen in hydroxyapatite overlap with oxygens from the substrate surface. This charging is

therefore indicative of the formation of the calcium phosphate crystals. M e r 48 hours of

growth time the surface is completely covered and the W S signal originates solely ffom the

calcium phosphate. Two peaks are observed with binding energies at 531 eV which is the

reported value for oxygen associated with the phosphate group in hydroxyapatite and the

second at 533 eV which is characteristic of oxygen in the HPO,~' group and intercrystalline

water.lC16 The same charging effect is seen with the calcium and phosphorus signals. Figure

3.12. is showing representative XPS spectra of Ca 2p core electmns. The binding energies of

Ca 2p and P 1s at 347.3 eV and 133.3 eV respecîively, are in agreement with the values

reported for hydroxyapatite. "16 The ratio of Ca/P oscillates around the value of 1.3 in

samples obtained after the nucleation penod was over (longer than 3 hours) which points to

the presence of a ns-OHM. This is in agreement with the FT-IR microscopy resdts

presented above.

In order to m e r prove that OCP is a prerequisite for the growth of the obtained

OHAp phase, controi experiments were perfonned. When the conditions of the synthesis

were changed in a way so as to favor OHAp nucleation rather then OCP such as, higher pH,

higher temperature and the presence of fluoride, no crystal growth was obsenred even a€ter

five days. W S spectra showed an interesting result whereby the presence of calcium and

phosphorous was in fact observed, albeit trace arnountso and in a 111 ratio. These values

however, did not change upon increased growth tirne. Conversely, when a lower pH (< 7)

was used which would ailow for nucleation of OCP, again film formation was not observed.

XPS spectra showed the absence of calcium or phosphorus on the substrate surface.

Taken together, these resdts cm be explained by the faa that in the case of

favorable OHAp nucleation, the surface of the substrate has both acidic and basic

aaivation/ancho~g sites to accommodate both calcium and phosphate species but the

inability of OCP to nucleate prevented any crystd growth. On the other hand, with a lower

pH, OCP could nucleate but the surface was mainly acidic which prevented calcium and

phosphorous from simul taneously binding to the surface. It is concluded therefore that the

presence of OCP as a precursor phase and the pH of the initial solution are both pivotai

factors that have to be controlled for the successful growth of porous, oriented and thick

crystalline hydroxyapatite films on the surface of TiO, on Ti.

35.2 CaDDP-Cap Composite on TiO, Surface

In Figure 3.13 PXRD diffraction patterns of Ti-TiO-OHAp and Ti-TiO-OHAp-

CaDDP films are displayed. It is evident that the phosphate ester surfactant forms a CaDDP

lamellar phase, with interlayer d-spacings of 40 A and 37 A under the experimental

conditions, and that its presence does not advenely affect the growth of the OHAp mineral

phase. The two Werent interlayer d-spacings probably correspond to distinct degrees of

interdigitation or tilting of the hydrophobic tails of the surfactant in the bilayer. The intensity

of the characteristic peaks of hydroxyapatite did not decrease upon formation of the CaDDP

which implies that no dissolution occurred during the hlm growih process.

A SEM micrograph of the calcium dodecylphosphate-hydroxyapatite (CaDDP-

OHAp) sample after one day growth tirne is s h o w in Figure 3.14. It can be seen that the

CaDDP phase is coating the füm of porous oriented hydroxyapatite crystals and that it is

without CaDDP

w

Figure 3.13. PXRD diffraction patterns of Ti-TiO-OHAp (top) and Ti-TiO-OH@-CaDDP @onom)

effectively following the contour surfaces of the crystals. The thickness of the CaDDP

lameliar phase varies £rom approximately 1 to 5 W. Figure 3.15 displays a TEM micrograph

of the same sample. It can be seen that the surface of the growing CaDDP lamellar phase is

onented parallel to the OHAp mineral long c-axes and moreover is organized in close contact

with the crystals. This micrograph is representative of the sample as a whole. It should be

noted that the same type of CO-assembly phenornenon was observed previously with

powdered foms of the CaDDP-OHAp composite material." The authors of this earlier work

proposed a model which could rationalize the formation and preferred orientation of the

composite material. The model assumed stereochernical, geometrical and

Figure 3.14. SEM micrographs of the calcium dodecy lphosphate-h ydroxy apatite composite film

Figure 3.15. Crosssectional TEM micrographs of the calcium dodecylphosphate- hydroxyapatite composite film

charge mmplementarity between the calcium and phosphate ester sites in the surface region

of the CaDDP 1ameUa.r phase and calcium and phosphate sites in the [O011 face of the

hydroxyapatite crystal lattice, with some degree of protonation of the interfacial phosphate

groups. Since hydroxyapatite fomed on TiO, is a nonstoichiometric HPO,~ containing

phase, it is expected that some of these HPO:- groups wiII also be located in the s d a c e

regions, particulariy because the source of the phosphate used is K-O, and the pH of the

solution was 7.4. This provides credence for the previously proposed mode1 of the CDDP-

OHAp chernical composite material." An illustration of the architecture of the CDDP-

OHAp-TiO-Ti composite Nm established in this study is shown in Figure 3.16.

Figure 3.16. Illustration of the architecture of CaDDP-OHAp-Ti0,-Ti composite material

The andogy between the construction of this purely synthetic multi-layer calcium

phosphate-based organic-inorganic composite film and the hierarchical structure of bone

grown in vitru on a planar substrate h m a bone cultue medium, can be appreciated by

cornparison of Figures 1.7. and Figure 3.16. Although the a d constituents and building

niles are not the same, the design and assembly phciples of the films have aspects in

common.

Fast growth of oriented hydroxyapatite crystalline films on a TiO, surface occurs as

a consequence of precursor OCP nucleation on distinct Brmsted acidic and basic anchoring

sites of the substrate and its in-situ hydrolysis to the more thermodynamically stable OHAp

phase. Porosity, nonstoichiometry and the rate of the OHAp film formation under

physiologicai conditions on the sputter deposited TiO, surface render this materiai as a viable

mating for bone implant materials. The presence of HPO,~ groups are found to be crucial not

only for ensuring chemical compatibility with the bone mineml phase but aiso as a key

cornponent for facilitating the formation of an organic-inorganic composite material. The

synthesis protocol of growing the OHAp and CaDDP phases separately, provides a hi@ level

of control over the thiclmess and composition of each component. It also allows for

alternation of these phases in a rnulti-layer composite construction, as well as a means for

incorporatiodrelease of bioactive substances fiom the hydrophobie region of the CaDDP

lamellar phase, that could improve the behaviour of this type of material in a body

environment.

1. D. Mayen, Suqactant Science and Technology; VVCH: New York, 1988

2. H. Baumgartel, E. U. Franck, W. Gmbein (eds.), Liquid Crystak; Steinkopff, Darmstadt,

1994

3. R. 2. LeGeros, G. Daculsi, 1. Orly, T. Abergas, W. Torres, Scanning Mimoscopy, 1989,

3(1), 129-138

4. W. E. Brown, M. Mathew, L. C. Chow: Roies of Octuculciwn Phosphate in Swface

ChemLFRy of Apotites; in Misra D. N. (ed): Adrorption on and Surfnce ChemL«ry of

Hydroxyupatite; Plenum press: New York, 1984,13-28

5. J. C. ELliot, Structure and C h e m i s ~ of the Aptites and Other Calcium Orthophosphntes;

Elsevier: New York, 1994

6. W. E. Brown, L. W. Schroeder, J. S. Fems, J. Am. Chem Soc., 1979,83(11), 1385-1388

7. C. Rey, M. Shimizu, B. Collins, M. J. Glimcher, Calcif. Tissue Znt., 1990,46,384-394

8. C. Rey, M. Shimizu, B. Collins, M. J. Glimcher, Calcif: Tissue Int., 1991,49,383-388

9. S. R. Radin, P. Ducheyne, J. Biomed mater. Res., 1993,27,35-45

10. N. Pleshko, k Boskey, R. Mendelsohn, Biophys. J., L991,60,786-793

1 1. B.O.Fowler, M.Markovic, W.EBrown, Chon Mater., 1993,5,1417-1423

12. E. L. Bullock, L. Patthey, S. G. Steinemann, Surface Science, 1996,352-354,504-510

13. T . K. Sham, M. S. Lazarus, Chem. Phys. Lett., 1979,68,426-432

14. P. W. Brown, B. Constantz, Hydroxypatite and Related Materials, 1994

15. S. Sugiyama, T. Mïnami, T. Moriga, H. Hayashi, K. Koto, M. Tanaka, J. B. Moffat, J.

Mater. Chem., 1996,6(3), 459-464

16. J. L. Ong, L C. Lucas, G. N. Railcar, J. J. Weimar, J. C. Gregory, Culloids Suflaces A:

Physicochem Eng. Aspecrs,l994,87,15 1-162

17. G. A. Ozin, N. Varaksa, N. Coombs, J. E. Davies, D. Perovic, M. Züiox, J. Mater.

Chem., 1997,7(8), 1601-1607

Appen& A: PXRD patterns of OCP (top) and OHAp @ottom)

Appendb B: Simulated e l emn diffraction patterns for [O101 zone of OCP (top) and

[1210] zone of OHAp

IMAGE EVALUATION TEST TARGET (QA-3)

APPLIED 1 IMAGE. lnc - 1653 East Main Strwt - -. - - Rochester. NY 14609 USA -- -- - - p h ~ : 71 6/42-0300 -- -- - - Fax: 71 6/28&5989