Hydroxyapatite/titania hybrid coatings

on titanium by sol-gel process

Ki-Hyeong Im

Department of Medical Science

The Graduate School, Yonsei University

Hydroxyapatite/titania hybrid coatings

on titanium by sol-gel process

Directed by Professor Yong-Keun Lee

The Master’s Thesis

submitted to the Department of Medical Science,

the Graduate School of Yonsei University

in partial fulfillment of the requirements

for the degree of Master of Medical Science

Ki-Hyeong Im

December 2006

This certifies that the Master’s Thesis

of Ki-Hyeong Im is approved.

___________________________

Thesis Supervisor: Yong-Keun Lee

___________________________

[Kwang-Mahn Kim: Thesis Committee Member #1]

___________________________

[Keun-Taek Oh: Thesis Committee Member #2]

The Graduate School

Yonsei University

December 2006

ACKNOWLEDGEMENTS

오랫동안 정들었던 연세 교정과 신촌, 그리고 우리 생체재료공학 교실

에 남겨진 저의 소중한 추억들을 되새기며, 이러한 결실을 맺을 수 있도록

도와주신 많은 분들에게 감사의 말을 드립니다.

먼저, 모든 면에서 부족한 저에게 따뜻한 관심과 큰 가르침을 주신 이

용근 지도교수님께 진심으로 감사 드립니다. 이렇게까지 성장할 수 있도록

학문적인 가르침뿐만 아니라, 인간적인 가르침에 대해서도 깊은 감사를 드

립니다. 그리고 늘 배려와 격려의 말씀을 해 주신 김경남 교수님과 생소한

생체재료 분야에 대한 많은 가르침과 조언을 주시고 교실 구성원에 대해

많은 관심을 가져 주신 김광만 교수님께도 깊은 감사를 드립니다. 또한,

논문에 대한 깊은 열정으로 많은 질타와 깨우침을 주신 오근택 박사님께도

감사의 뜻을 전합니다.

교실의 에~엠을 자쳐하며 평가센터 뿐 아니라 대학원생들도 잘 챙겨주

신 연희동 연장자 남이 누나, 숙자라 부르며 오빠의 마음을 아프게 한 미

시완실 보영, 형들에게 최고의 귀여움을 받았던 이뿐 지혜, 다양한 역할을

소화하시며 바쁘신 와중에도 많은 도움을 주신 이상배 선생님, 서대문구를

떠나버린 구리얼짱 지효 엄마 해경, my lovely english tutor 혜옥 샘, 교실

최고의 striker & 스타 유져가 될 잠꾸러기 의국장 우현이, 교실 유일의

뱀띠 동기이자 최고령 violinist 지연이, 얼핏 김아중 같은 너무 착한 종숙

이, 최근에 교실 식구가 되어 친해질 수 있는 기회가 적어서 아쉬운 성제

씨 모두 고맙습니다.

나의 사랑스런 연구실 후배들 노문탁; 대학원생 막내로써 비타민 같은

존재였던 天常女子 지연, 長幼有序를 가장 잘 이해하고 몸소 실천하는 너

무 착한 문반장 승균, 항상 외부로 열심히 실험하러 다니는 경상도 사나이

탁사마 창우. 좋은 후배들이 되어 주어 너무 고맙습니다.

올림픽 멤버들; 술에서 벗어나 자기 관리에 조금씩 눈을 뜨기 시작한

이 시대 마지막 로맨티스트 미남Tommy 동국, 항상 내일처럼 잘 챙겨주었

던, 왠지 나중에 전화하면 ‘이따 전화할게’ 할 것 같은 쾌남 Farmer 재선,

후배들의 고민을 진심으로 이해하고 도와주던, 크게 성공할 것 같은 행복

한 남자 센스쟁이 Jerry영수 영일. 너무나도 좋은 형들이 있어 대학원 생

활이 즐거웠습니다.

The 3rd Sectors; 항상 가까이에서 우리의 학문적 도움 뿐 아니라 다방

면에서 대학원생들의 대변인이 되어 교실 복지 향상에 큰 역할을 해주신

건담쵸리 민철이 형, 살 찔 고민으로 살이 더욱 빠져버린, 하지만 인간적

으로는 가장 큰 근육을 가진 멋진 친구 Antoniho 세호, 같이 들어와서 먼

저 나가게 되어 미안한, 그러나 어느 누구보다 잘하리라 생각되는 만능 재

주꾼 동남왕자 병현. 당신들이 있어서 sector life가 행복하였습니다.

세라믹의 최고 형님이시고 많은 격려를 해 주신 연웅이 형, 지금 멀리

미국에서 열심히 공부하고 있을 자칭 세라믹ACE 멋쟁이 동박동현, 대학원

생활 초기에 많은 도움을 준 빠조지호, 코팅 장비사용에 많은 도움을 주신

준영이형, 그리고 95 세라믹 형들 석민, 준호, 동찬, 기훈 모두 고맙습니다.

그리고, 많은 일을 병행하면서도 열심히 실험하시고 동생처럼 챙겨주신 마

음이 따뜻한 누나 주혜, 상환 샘, 회사일과 학교일에 모두 최선을 다하는

모습이 아름다운 욱현이형. 정말 고맙습니다. 언제든지 찾아가면 반갑게

맞아 주시며 큰 스폰서가 되어 주신 술익는 마을 & 아파치의 co~ol한 형

님과 형수님께도 감사드립니다.

언젠가 멋진 모습으로 다시 만나기 위해 지금은 흩어져서 각자의 위치

에서 열심히 생활하고 있는 멋진 club 부티나 멤버들; 창시기, 용파리, 배

때기, 진수, 성칠이, 봉이, 스리. 너희들이 자랑스럽구나.

늦게 까지 공부하고 있는 친구를 위해 물심양면으로 지원해주며 큰 힘

이 되어 준 멋진 학성고 친구들; 소심 호성, 상담 환영, 우웅견마, 브랑카

용석, 메뚝 성준, 덩걸 기덕 그리고 대두회 아이들, 長幼有序를 위협하는

후배 호계 수준. 참 멋진 놈들이야.

더욱 멋진 사람으로 거듭나기 위해 많은 고민을 하고 있지만, 분명 멋

진 놈이 될 것 같은 나의 멋진 동생 Juan기석, 필요할 땐 언제나 달려와

주시고, 누구보다도 더 알차고 멋진 삶을 살고 있는 나의 정신적 지주 성

훈이형. 사랑하고 존경합니다.

내 인생 최고의 서포터이자 모든 면에서 본받을 점이 많은, 아버지와

같은 나의 형님과 나의 사랑스러운 조카 빈이의 엄마이자 형보다 나를 더

잘 이해해주시는 형수님께도 감사의 말을 전합니다.

끝으로, 지금의 제가 여기에 있게 해주신, 어떠한 말로도 그 큰 은혜를

표현할 수 없는, 세상 어느 누구보다 가장 사랑하는 나의 어머니와 하늘나

라에서 크게 기뻐하실 가장 존경하는 아버지께 이 논문을 받칩니다.

2006년 12월

임기형 드림

I

TABLE OF CONTENTS

LIST OF FIGURES ·····················································································ⅣⅣⅣⅣ

LIST OF TABLES························································································ⅥⅥⅥⅥ

ABSTRACT···································································································· 1

ⅠⅠⅠⅠ. INTRODUCTION

1. Titanium and titanium alloys for biomedical application ·····································3

2. Surface modification of titanium and titanium alloys ··········································5

3. Hydroxyapatite coatings on titanium····································································6

4. Hydroxyapatite/titania hybrid coating on titanium·············································10

5. The objective in this study ·················································································· 13

ⅡⅡⅡⅡ. MATERIALS AND METHODS

1. Materials ·············································································································14

A. Preparation of sols························································································ 14

(A) Hydroxyapatite sol ················································································· 14

(B) Titania sol ······························································································· 14

(C) Hydroxyapatite/titania hybrid sol··························································· 15

B. Sol-gel Coatings on titanium········································································ 17

C. MG63 cell ···································································································· 19

II

2. Methods ················································································································20

A. Phase identification of sol-gel powders ·························································20

B. Hybrid coatings on titanium substrate by sol-gel process······························20

(A) Characterization of coatings on titanium ·················································20

ⓐ Phase identification of coatings····························································20

ⓑ Observation of surface morphology ·····················································20

ⓒ Measurement of surface roughness ······················································21

ⓓ Evaluation of the chemical composition ··············································21

ⓔ Measurement of thickness ····································································21

(B) Scratch test for bonding strength······························································22

(C) Bioactivity test using simulated body fluid··············································24

(D) In vitro cell test ························································································26

3. Statistical analysis·································································································28

ⅢⅢⅢⅢ. RESULTS

1. Phase identification of sol-gel powders································································29

2. Hybrid coatings on titanium substrate by sol-gel process ····································34

A. Characterization of coatings on titanium ·······················································34

(A) Phase identification of coatings ·······························································34

(B) Observation of surface morphology·························································34

(C) Measurement of surface roughness ························································37

(D) Evaluation of chemical composition························································39

(E) Measurement of thickness ········································································39

III

B. Scratch test for bonding strength····································································43

C. Bioactivity test using simulated body fluid····················································47

D. In vitro cell test ······························································································49

ⅣⅣⅣⅣ. DISCUSSION

1. Preparation of hydroxyapatite/titania hybrid sol···················································53

2. Properties of coatings on titanium substrate ·························································55

3. Scratch test for bonding strength between coatings and substrates ······················56

4. Bioactivity test 58

5. Biological performance·························································································59

ⅤⅤⅤⅤ. CONCLUSION ·································································································62

REFERENCES········································································································64

ABSTRACT (IN KOREAN) ················································································71

IV

LIST OF FIGURES

Figure 1. Stages of the spin coating process···································································9

Figure 2. Sol–gel process of hydorxyapatite/titania composite coatings ·····················12

Figure 3. Fabrication of hybrid sols by sol-gel process················································16

Figure 4. The image of spin coater···············································································18

Figure 5. The image of scratch tester ···········································································23

Figure 6. The photograph of (a) 96 well plate and (b) MRX microplate reader for the

MTT assay··································································································27

Figure 7. XRD patterns of hydroxyapatite sol-gel powders with (a) 0, (b) 2, (c) 3, and

(d) 4 wt % ammonium hydroxide·······························································30

Figure 8. XRD patterns of the powders; (a) PH, (b) HT10, (c) HT30, (d) HT50, (e)

HT70, and (f) PT powder ···········································································31

Figure 9. FT-IR spectra of the powder obtained by sol-gel process; (a) PH, (b) HT30,

and (c) PT sol-gel powder ··········································································33

Figure 10. XRD patterns of the coatings on Ti substrate; (a) PH, (b) HT10, (c) HT30,

(d) HT50, (e) HT70, and (f) PT coatings····················································35

Figure 11. SEM morphologies of the coating surface on Ti substrate; (a) PH, (b)

HT10, (c) HT30, (d) HT50, (e) HT70, and (f) PT coatings························36

Figure 12. EDS spectra of the coating on Ti substrate; (a) PH, (b) HT10, (c) HT30, (d)

HT50, and (e) HT70 coating ······································································40

V

Figure 13. SEM morphologies of the cross section of the coating layers on Ti

substrate; (a) PH, (b) HT10, (c) HT30, (d) HT50, (e) HT70, and (f) PT

coating layer ·······························································································42

Figure 14. Diagrams of friction force and acoustic Emission versus for the scratch

tests on the coatings; (a) PH, (b) HT10, (c) HT30, (d) HT50, (e) HT70,

and (f) PT coatings ·····················································································44

Figure 15. Scratch track and EDS mapping of (a), (c) HT10 and (b), (d) HT70

coatings·······································································································46

Figure 16. SEM morphologies of the (a) PH, (b) HT30, and (c) HT70 coatings after

immersion of specimens in SBF for 1 week, and (d) PH, (e) HT30, and (f)

HT70 coatings for 2 weeks·········································································48

Figure 17. MG63 cell attachment density on the coatings on Ti with culturing for 6

h··················································································································50

Figure 18. Proliferation of the MG63 cell on coatings Ti after culture for 2, 4, and

7days···········································································································51

VI

LIST OF TABLES

Table 1. Overview of surface modification methods for titanium and its alloys········· 5

Table 2. The composition of samples ········································································ 15

Table 3. Ion concentrations of simulated body fluid and human blood plasma ········ 25

Table 4. Surface roughness of the coating layers on Ti substrate······························ 38

Table 5. Thickness of coating layers on Ti substrate················································· 41

Table 6. The Critical load of coatings on Ti substrate ··············································· 45

１

ABSTRACT

Hydroxyapatite/titania hybrid coatings on titanium

by sol-gel process

Ki-Hyeong Im

Department of Medical Science

The Graduate School, Yonsei University

(Directed by Professor Yong-Keun Lee)

Sol-gel thin films of hydroxyapatite and titania have received a great deal of attention

in the area of bioactive surface modification of titanium implants. Sol-gel process

offers lots of advantages over other coating techniques, e.g. increased homogeneity

due to atomic level mixing; finer grain microstructure and lower temperature of the

crystallization. In this study, we fabricated hydroxyapatite/titania hybrid coatings on

titanium by sol-gel method to combine advantages of both materials: the adhesion

strength of the titanium dioxide on the substrate and the bioactivity of the

hydroxyapatite. Sol-gel coatings of pure hydroxyapatite and titania, and four

hydroxyapatite composites with 10~70 mol% titania were developed on titanium

substrates. The characteristics of coatings, such as crystallinity, thickness,

morphology, roughness and composition of surface, were observed by X-ray

diffraction, scanning electron microscopy, surface profiler and energy dispersive

spectroscopy, respectively. In order to obtain stable hydroxyapatite sols, different

concentrations (0, 2, 3, and 4 wt%) of NH4OH was added into the hydroxyapatite sol.

When 3 wt% NH4OH was added, typical apatite patterns with high intensity appeared.

Both powders and coatings on titanium by sol-gel methods showed the characteristic

２

peaks of pure hydroxyapatite and titania with anatase and rutile structures. The

surface of most coating layers was quite highly dense and homogeneous in

microstructure. When the titania amounts adding into the hydroxyapatite sol increased,

the morphology of the composite coatings became more porous and rougher. For

roughness value, composite samples with more than 30 mol% titania have

significantly higher (p0.05).

Key words: hydroxyapatite, titania, sol-gel method, hybrid coating, titanium, bonding

strength, bioactivity, cellular response.

３

Hydroxyapatite/titania hybrid coatings on titanium

by sol-gel process

Ki-Hyeong Im

Department of Medical Science

The Graduate School, Yonsei University

(Directed by Professor Yong-Keun Lee)

ⅠⅠⅠⅠ. INTRODUCTION

1. Titanium and titanium alloys for biomedical application

Titanium (Ti) is the newest metallic biomaterial among the three main metallic

biomaterials (i.e., titanium, stainless steel and Co-based alloy), and these remain the

most popular metallic biomaterials. Among the main metallic materials for

biomedical applications, Ti and its alloys are getting much attention in both medical

and dental fields because of good corrosion resistance, biocompatibility, light weight

and excellent balance of mechanical properties.1-5 There have been reported that the

surface oxide of Ti is inert when in contact with biological tissue and that the

formation of collagen on the surface of Ti components promotes the growth of new

bone tissues6 and also reported relatively high compatibility with bone and tissue as

compared with stainless steel.7

In this manner, Ti and its alloys have long been studied in the various field of

medicine such as dental and orthopedic implant materials due to their excellent

４

properties.1 However, the bone in-growth properties and implant fixation behavior

need to be improved in order to shorten the implant-tissue osseointegration time.8

2. Surface modification of titanium and titanium alloys

The bulk properties of biomaterials, such as non-toxicity, corrosion resistance or

controlled degradability, modulus of elasticity, and fatigue strength have long been

recognized to be highly relevant in terms of the selection of the right biomaterials for

a specific biomedical application.9-11 The events after implantation include

interactions between the biological environment and artificial material surfaces, onset

of biological reactions, as well as the particular response paths chosen by the body.12

The material surface plays an extremely important role in the response of the

biological environment to the artificial medical devices. In implants made of Ti, the

normal manufacturing steps usually lead to an oxidized, contaminated surface layer

that is often stressed and plastically deformed, non-uniform and rather poorly

defined.13

Such ‘‘native’’ surfaces are clearly not appropriate for biomedical applications and

some surface treatment must be performed. Another important reason for conducting

surface modification to Ti medical devices is that specific surface properties that are

different from those in the bulk are often required. For example, in order to

accomplish biological integration, it is necessary to have good bone formability. In

blood-contacting devices, such as artificial heart valves, blood compatibility is crucial.

In other applications, good wear and corrosion resistance is also required. The proper

surface modification techniques not only retain the excellent bulk attributes of Ti and

its alloys, such as relatively low modulus, good fatigue strength, formability and

machinability, but also improve specific surface properties required by different

clinical applications. According to the different clinical needs, various surface

modification schemes have been proposed and are shown in Table 1.14 These methods

are classified into mechanical, chemical and physical methods according to the

formation mechanism of the modified layer on the material surface.

５

Table 1. Overview of surface modification methods for titanium and its alloys14

Surface modification methods

Modified layer Objective

Mechanical methods Machining Grinding Polishing Blasting

Rough or smooth surface formed by subtraction process

Produce specific surface topographies; clean and roughen surface; improve adhesion in bonding

Chemical method Chemical treatment

Acidic treatment Alkaline treatment Hydrogen peroxide

treatment Sol-gel Anodic oxidation CVD

Biochemical methods

６

3. Hydroxyapatite coatings on titanium

Among various technologies, hydroxyapatite [HA, Ca10(PO4)6(OH)2] is

traditionally coated on Ti by plasma spraying method.15 Because of its similarity to

the mineral phase of natural hard tissues, artificial HA is considered to be a bioactive

material. Bone can be regarded as an organic matrix with inclusion of inorganic filler

with a crystal size in the submicron range. About 70% of the mineral fraction of bone

has a HA-like structure and the use of HA as an orthopedic biomaterial has been

suggested and clinically demonstrated. However, the mechanical properties of HA are

quite poor, making it unsuitable as bulk a material in applications where high load or

strain occurs.

The idea of using plasma spraying to produce HA coatings on endoprosthesis was

first proposed in Japan.16 It has been shown that implants with a HA surface develop a

strong connection with the bone tissue in a short time.17 The relatively poor bonding

between a plasma sprayed HA coating and Ti is one of main disadvantages. It is

obvious that the bonding strength of the HA coating on Ti decreases with longer

immersion time in simulated body fluids (SBF).18 The high residual stress resulting

from the mismatch of the thermal expansion coefficient between HA (13.3 × 10-6 K-1)

coating and Ti ((8.4-8.8) × 10-6 K-1) is thought to be responsible for the low bonding

strength between the two materials. Another important concern is the resorption and

degradability of HA coatings in a biological environment. This may lead to

disintegration of the coating resulting in the loss of both the coating–substrate

bonding strength and implant fixation. There also exists the threat of coating

delamination and formation of particulate debris.19 During plasma spraying, the

molten HA particles deposited on the metal substrate are quenched at a high cooling

rate leading to the formation of amorphous calcium phosphate compounds and the

reduction of the crystallinity of the coating. Some metastable compounds, such as

calcium-deficient hydroxyapatite (CDHA), oxyhydroxyapatite (OHA), calcium oxide

(CaO), a-tricalcium phosphate (a-TCP), b-tricalcium phosphate, and tetra-calcium

phosphate (TTCP) have been detected in the coating.20 It has been reported that the

amorphous and metastable compounds are more soluble than the crystalline HA

thereby accelerating fixation of the implant with the bone and promoting bone

７

remodeling and attachment.21

However, the coating with amorphous and metastable compounds causes excessive

dissolution. It affects the long-term reliability of the implants.22 From the perspective

of long-term performance of the implant, a high crystallinity in the plasma sprayed

HA coating is deemed desirable. However, there are some issues raised in the plasma

spraying process, such as coating strength, chemical homogeneity, and residual

porosity. The plasma-sprayed HA coating layer is known to be inhomogeneous in

structure and have low bonding strength (20-30 MPa) due to the high-temperature

processing and resultant thick coating layer.23,24 Recently, alternative methods have

been developed to produce thin HA films.

A colloid is a suspension in which the dispersed phase is so small (1-1000 nm) that

gravitational forces are negligible and interactions are dominated by short-range

forces, such as van der Waals attraction and surface charges. A sol is a colloid

suspension of solid particles in a liquid and a gel is a substance that contains a

continuous solid skeleton enclosing a continuous liquid phase.25 The sol–gel process

can be divided into five main steps: (1) hydrolysis and polycondensation; (2) gelation;

(3) aging; (4) drying; (5) densification and crystallization.26 Metal alkoxides, M(OR)n,

are the most common metalorganic precursors, where M is a metal or metalloid atom

and R is an alkyl group. Normally, the alkoxide is dissolved in alcohol and

hydrolyzed by the addition of water. Hydrolysis replaces the alkoxide ligands with

hydroxyl ligands as shown in the following reaction: M(OR)4+H2O à HO-M-(OR)3 + ROH (hydrolysis)

Two partially hydrolyzed molecules can link together via a condensation reaction:

(OR)3M-OH + HO-M(OR)3 à (OR)3-M-O-M-(OR)3+H2O (oxolation) Conversely, one hydrolyzed molecules can react directly with one alkoxide

molecule:

(OR)3M-OR + HO-M(OR)3 à (OR)3-M-O-M-(OR)3 + ROH (alcoxolation)

８

In both cases a small molecule (water or alcohol) is liberated. When a sufficiently

large number of interconnecting M–O–M bonds is formed in a region, they respond

cooperatively as colloidal particles or a sol. The colloidal particles gradually link

together to become a three-dimensional network. If one molecule reaches

macroscopic dimensions so that it extends throughout the solution, it is said to be a

gel and the process is called gelation. Gelation can also be produced by rapid

evaporation of the solvent. The structure and properties of the gel can be changed by

the aging process. During aging, polycondensation continues with dissolution and

reprecipitation of monomers or oligomers. At the same time, the wet gel is converted

into a dry gel, and considerable shrinkage occurs. In order to control the stress and

strains in the gel, the drying process must be carefully designed. Most gels are

amorphous even after drying, but many crystallize when heated. If the objective of the

process is to produce a pore-free ceramic, it is necessary to heat the gel to a high

enough temperature to induce sintering. Sol–gel thin films are usually produced using

the spin coating. The spin coating process consists of four stages: deposition, spin up,

spin off, and evaporation as shown in Figure 1.

The sol-gel process is widely used to deposit thin (

９

Figure 1. Stages of the spin coating process.27

SpinningSpinningSpinningSpinning

１０

The use of the sol–gel process in the biomedical area is recent. The first papers

came out about ten years ago, and the number of publications has noticeably

increased during the last five years. The sol-gel technique, being one of the thin film

methods, provides some benefits over plasma-spraying method, such as chemical

homogeneity, fine grain structure, and low processing temperature.28 Moreover, when

compared with other thin film methods, it is simple and cost efficient, as well

effective for the coating of complex-shaped implants. In the sol-gel HA coatings,

various precursors have been tried in pursuit of obtaining a well developed HA layer;

Ca(NO3)2, CaO, Ca(OH)2, and Ca alkoxide as a Ca source: P alkoxide such as triethyl

phosphate[P(OC2H5)3] (TEP) and triethyl phosphate [PO(OC2H5)3], P2O5, and H3PO4

as a P-source.28-34 Among these precusrsors, the Ca(NO3)2 and TEP combination was

recently introduced to obtain a pure HA phase in an ethanol-water mixed solution.32-34

The crystalline HA phase builds up between 350 ℃ and 400 ℃. It remains pure

even after the ceramic is sintered at a higher temperature, between 900 ℃ and

1100 ℃. The absence of additional phases such as beta-tricalcium phosphate, tetra

calcium phosphate or calcium oxide in this temperature region is the sign that the

stoichiometry of HA is respected.

In addition, the possibility of modifying the surface area, porosity, composition,

adsorption capacity, and dissolution rate using the sol-gel technique is very attractive

in the fields of medicine and dentistry. The sol-gel method is a relatively simple way

to prepare HA coatings on Ti and Ti alloys because of the easy formation of the oxide

coatings at a relatively low temperature. Kim et al. prepared HA and FHA coatings

using this method, and investigated their biological performance.35 The HA- and FHA

-coated Ti exhibited higher ALP expression levels compared to pure Ti, confirming

improved activity and functionality of cells.

4. Hydroxyapatite/titania hybrid coatings on titanium by a sol-gel process

The HA coatings synthesized by the sol-gel method are typically bioactive but have

poor adhesion strength to the substrate. On the other hand, titanium dioxide coatings

strongly adhere to Ti substrate, but their bioactivity is limited. Hence, a composite

HA/TiO2 coating can take advantage of the high adhesion strength of TiO2 and

bioactivity of HA. Kim et al. deposited HA on Ti by means of a titania (TiO2) buffer

１１

layer using the sol-gel method.36 The HA layer showed a typical apatite phase at

400 ℃. The phase intensity increased above 450 ℃. The HA and TiO2 films of

thicknesses of approximately 800 and 200 nm, respectively, adhered tightly to each

other and to the Ti substrate. The bonding strength of the HA/TiO2 double layer was

markedly improved when compared to that of a single HA coating on Ti. The

maximum strength of the double layer coating of 55 MPa was attained after heat

treatment at 500 ℃. The improvement in the bonding strength with the insertion of a

TiO2 buffer layer was attributed to the enhanced chemical affinity of TiO2 towards the

HA layer as well as Ti substrate. Human osteoblast (HOS)-like cells cultured on the

HA/TiO2 coating surface proliferated in a similar manner to those on the single TiO2

coating and Ti. However, the alkaline phosphatase (ALP) activity of the cells on the

HA/TiO2 double layer was expressed to a higher degree than the single TiO2 single

coating and Ti. The corrosion resistance of Ti was also improved by the presence of

the TiO2 coating as confirmed by potentiodynamic polarization tests.

Recently, HA/TiO2 composite coatings on Ti have been studied to control he

surface properties of coatings, such as roughness and chemical composition.

Practically, the HA/TiO2 composite coating system has been tried by a few other

groups.37-40 The layers were reported to have improved mechanical properties or in

vitro cellular responses. Milella et al. prepared HA/TiO2 composite coatings as

illustrated schematically in Figure 2.39 TiO2 sol was prepared by mixing titanium

isopropoxide, acetyl acetone, nitric acid, n-propane alcohol and distilled water. HA

powders were then added to anhydrous ethanol in the same ratio to prepare the HA

sol. Two sol solutions could be mixed in different ratios. A homogeneous, rough and

porous HA/TiO2 composite coating was produced on the Ti substrate and hydroxylic

groups as Ti-OH were detected on the surface of the coatings. The film adhesion

strength determined using the pull-test was 39.8±3.75 MPa. The biocompatibility of

HA/TiO2 coatings prepared by the sol–gel process was investigated by Ramires et

al..40 They suggested that the composite coatings were not cytotoxic and allowed the

differentiation of cells stimulating the expression of some peculiar osteoblast

biochemical markers: alkaline phosphatase activity, collagen and osteocalcin

production.

１２

Figure 2. Sol–gel process of hydorxyapatite/titania composite coatings.40

１３

In particular, the HA/TiO2 coatings increased some differentiation markers of the

osteoblastic phenotype. The good performances of these coatings can be explained by

the characteristics of the chemical composition and the deposition technique. The

HA/TiO2 coatings are bioactive due to the presence of hydroxyl groups on the surface

promoting calcium and phosphate precipitation thereby improving the interactions

with osteoblastic cells.

However, in most cases, the HA/TiO2 composite coatings were made by a plasma

spraying method, or else the HA powder was mixed with TiO2 sol. In such cases, the

resultant layer was thick, rough, and inhomogeneous. Practically, there have been

little reports on the HA/TiO2 composite film produced by using sol-gel precursors of

both HA and TiO2.

5. The objective in this study

The objective of this study is to fabricate HA/TiO2 hybrid coatings on Ti substrate

by sol-gel method to combine advantages of both materials: the adhesion strength of

the titanium dioxide on the substrate and the bioactivity of the HA. Various mol %

amounts of TiO2 sols were added to HA sols to make hybrid sols. Ti substrates were

coated with the composite sols by spin-coater. The phase identification, morphology,

roughness, thickness, and chemical composition of sol-gel powders and coatings on

Ti were determined, and the bonding strength between coatings and substrates were

tested by scratch test. Biological performances such as cell attachment and

proliferation were evaluated using MG63 osteoblast cell. Therefore, the purpose is to

optimize the composition of coatings with not only excellent bonding strength but

also excellent bioactive and biological properties.

１４

ⅡⅡⅡⅡ. MATERIALS AND METHODS

1. Materials

A. Preparation of sols.

(A) Hydroxyapatite sol

Calcium nitrate tetrahydrate (Ca(NO3)2·4H2O)) and triethyl phosphite [P(OC2H5)]3

as the Ca and P source were used to obtain a pure HA phase in an ethanol-water

mixed solution. In order to make Ca solution, calcium nitrate tetrahydrate (Sigma-

Aldrich, USA) of 2 M was dissolved in ethanol by stirring for 24 h. A stoichiometric

amount (Ca/P ~ 1.67) of triethyl phosphite (Sigma-Aldrich, USA) was hydrolyzed in

ethanol-water mixed solution (at a ratio of 6) and the mixture was stirred for 24 h to

prepare P solution. After mixing the Ca and P solutions and stirring for 30 min,

different concentrations of NH4OH (0, 2, 3, 4 wt%) were added to the mixed solutions,

and the solutions were stirred for an additional 30 min. Finally, the HA sol was

obtained by aging at 60 ℃ for 72 h.

(B) Titania sol

Titanium propoxide [Ti(OCH2CH2)2NH] was used as Ti alkoxide to make a TiO2 sol.

To produce a TiO2 sol, titanium propoxide (Sigma-Aldrich, USA) of 1 M was

hydrolyzed within an ethanol-based solution, containing diethanolamine

[(HOCH2CH2)2NH, Aldrich, USA] and distilled water. The molar ratios of the

diethanolamine/Ti and water/Ti were 1 and 2, respectively. The sol was stirred at

room temperature for 24 h and, the TiO2 sol was obtained by additional aging at

50 ℃ for 72 h.

１５

(C) Hydroxyapatite/titania hybrid sol

Figure 3 shows the fabrication of hybrid sols by sol-gel process. The prepared TiO2

sol was added to the prepared HA sol with various ratios (10, 30, 50, and 70 mol %).

and the mixtures were stirred vigorously for 6 h and slowly for a further 3 h at 40 ℃,

to finally obtain HA/TiO2 composite sols. HT10, HT30, HT50, and HT70 were

regarded as the TiO2 amount of 10, 30, 50, and 70 mol % added to HA sol as shown

in Table 2. For the purpose of comparison, pure HA (PH) and TiO2 (PT) sols were

prepared without mixing.

Table 2. The composition of samples

Samples The amount of TiO2 added to HA sol.

(mol %)

PH Pure hydroxyapatite Control

Group PT Pure titania

HT10 10

HT30 30

HT50 50

Experimental

Group

HT70 70

１６

Figure 3. Fabrication of hybrid sols by sol-gel process.

１７

B. Sol-gel coatings on titanium substrate

Commercially pure Ti (c.p. Ti, grade III) disc was purchased from Dynamet, USA.

Pure Ti disks were cut into 10 × 10 mm size as coating substrates, and were prepared

after polishing with silicon carbide paper (#1500 grit), and then cleaning in acetone

and ethanol. The prepared composite sols were dropped onto the Ti substrate and then

spin-coated at 2000 rpm for 10 s. After drying at 80 ℃ for 24 h, samples were heat

treated at 500 ℃ for 2 h in air at a heating and cooling rate of 1 ℃/min. For the

purpose of comparison, pure HA and TiO2 coatings were also prepared at the same

coating and heat treatment conditions.

１８

Figure 4. The image of spin coater.

１９

C. MG63 cell.

MG63 osteoblast-like cells, originally isolated from a human osteosarcoma, were

purchased from Korea Cell Line Bank. The cells were cultured in Dulbecco Modified

Eagle’s Medium (DMEM; Gibco, Grand Island, NY, USA), supplemented with 10%

Fetal Bovine Serum (FBS; Gibco, Grand Island, NY, USA) in 75 cm2 tissue culture

polystylene flasks (BD Falcon, Bedford, MA, USA) in a 37 ℃, 5% CO2 humidified

atmosphere incubator (VS-9180MS, Vision Scientific Co., Korea). Cells used for

evaluations were between the 5th and 15th passages.

２０

2. Methods

A. Phase identification of sol-gel powders

In order to make powders, the prepared HA, TiO2, and HA/TiO2 sol were dried the

at 80 ℃ for 24 h in air, and were heat treated at 500 ℃ for 2 h in air at a heating

and cooling rate of 1 ℃/min. The phase of the synthesized powders was investigated

using X-ray diffractometer (XRD; X’pert PW1830, Philips, Japan) with Ni-filtered

Cu-kα ray (λ=1.5406 nm). Diffraction data were collected between 20° and 80° at a scan rate of 4°/min. The crystalline structures were identified according to JCPDS

software (PCPDFWIN1.30, JCPDS-ICDD). The chemical bonding structure of the

powders was analyzed with Fourier transformed infrared spectroscopy (FT-IR;

AVATAR360, Nicolet, USA).

B. Hybrid sol coated layer on Ti substrate

(A) Characterization of coatings on Ti

ⓐ Phase identification of coatings

The phase of the HA, TiO2 and composite sol coated layer on Ti was investigated

using X-ray diffractometer (XRD; X’pert PW1830, Philips, Japan) with Ni-filtered

Cu-kα ray (λ=1.5406 nm). Diffraction data were collected between 20° and 80° at a scan rate of 4°/min. The crystalline structures were identified according to JCPDS

software (PCPDFWIN1.30, JCPDS-ICDD).

ⓑ Observation of surface morphology

The HA, TiO2 and composite coated film were prepared to determine the surfaces

using scanning electron microscopy (SEM; Hitachi, Tokyo, Japan). The surfaces were

gold-coated with the thickness of 300 Å using ion sputter (Eiko IB-3, Tokyo, Japan),

２１

and examined using SEM at an accelerating voltage of 3 KV, and digital images were

captured at varying magnification.

ⓒ Measurement of surface roughness

Roughness of the coating layer was measured at a speed of 0.1 mm/s using a

surface profiler (Surfcorder SEF-30D, Kosaka Laboratory Ltd., Japan). Roughness

parameter, Ra (average height above center line), were obtained. Three samples were

tested, and average surface roughness measurements were taken at three different

areas for each sample to determine the mean±standard deviations (n=9).

ⓓ Evaluation of the chemical composition

Energy dispersive spectroscopy (EDS; Inka, Oxford, UK) was used to determine

the Ca/P ratio of coating surface on Ti.

ⓔ Measurement of thickness

The coating thickness was measured by scanning electron microscopy (SEM;

Hitachi, Tokyo, Japan). The cross section of layers was gold-coated with the thickness

of 300 Å using ion sputter (Eiko IB-3, Tokyo, Japan), and examined using SEM at an

accelerating voltage of 3 KV, and digital images were captured at varying

magnification. The average value was taken from arbitrary five areas and represented

as the mean±standard deviation.

２２

(B) Scratch test for bonding strength between coatings and substrates.

The scratch test is a common technique used to evaluate the adhesion of a thin film

on the substrate. The adhesion of the films obtained was assessed using a scratch

tester (CSEM-REVETEST, Swiss) with a spherical Rockwell C diamond stylus of

200 µm-radius. The scratches were generated on the coatings by constantly increasing

the load at the rate of 50 N/min from initial load 2 N while the specimen was

displaced at the constant speed of 5 mm/min. The point of adhesion failure of the

coating from the substrate was detected by a burst increase in the acoustic signal and

friction force from the sample. The load at which the first damage is made in the

coating is called the “lower critical load”; and the load at which total peeling-off of

the coating from the substrate occurs is referred to as the “upper critical load”. In hard

coatings, usually the “lower critical load” is used a measure of the adhesion strength.

To evaluate the coating/substrate adhesion, we choose to use the “upper critical load”

as an indication of the adhesion strength. Three readings of such critical load values

were averaged for each sample, and the standard deviation of these data was taken as

the error. At each composition, two samples were tested. The scratch track was

observed using the EDS mapping of Ca.

２３

Figure 5. The image of scratch tester.

２４

(C) Bioactivity test using a simulated body fluid

The bioactivity tests were performed using a simulated body fluid (SBF) that was

prepared by dissolving reagent chemicals of NaCl, NaHCO3, KCl, K2HPO4· 3H2O,

MgCl2·6H2O, CaCl2, and Na2SO4 into deionized water.41,42 The fluid was buffered at

physiologic pH 7.40 at 37 ℃ with tris(hydroxymethyl)-aminomethane and

hydrochloric acid.41 SBF with ionic concentration approximately the same as that of

human blood plasma, as reported in Table 3, were used.43

Each specimen was immersed in SBF solution in a closed container. The ratio of the

coating surface area (microscopic; SA) to SBF solution volume (V) was fixed to 0.1

cm−1.44 The coatings were immersed in the SBF fluid for 1 week, the bottles being

placed having a constant temperature of 37 ℃. After immersion in SBF for 1 and 2

weeks, the specimens were rinsed in distilled water, dried in a vacuum desiccator.

Scanning electron microscopy (SEM) was used to observe the morphology of the

precipitates after gold sputtering.

２５

Table 3. Ion concentrations of simulated body fluid and human blood plasma43

Ion concentrations (mM) Ion

Blood plasma SBF

Na+

K+

Mg2+

Ca2+

Cl-

HCO3-

HPO42-

SO42-

pH

142.0

5.0

1.5

2.5

103.0

27.0

1.0

0.5

7.2-7.4

142.0

5.0

1.5

2.5

147.8

4.2

1.0

0.5

7.40

２６

(D) In vitro cell test

All specimens for the cell tests were prepared after sterilization at 121 ℃ for 20

min. For initial cell attachment tests, the cells were plated at a density of 5 × 104

cells/mL in 0.1 mL medium on all the specimens (10 × 10 mm) in individual wells of

a 24-well plate and cultured for 6 h to allow the cells to attach. At the period, the cells

were washed with PBS solution to eliminate the non-adherent cells.45

Then MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide) reagent

(5 mg/ml MTT in PBS) was added onto the specimens. The mitochondria

dehydrogenated from the living cells would reduce the MTT reagent into water-

unsoluble blue crystals.46 After removal of the media, dimethylsulfoxide (DMSO) was

added onto the specimens to dissolve the blue crystals. The optical density (OD) of

the dissolved solute was then measured by an ELISA reader under a light source of

570 nm wavelength. The percentage of OD value (OD for specimen/OD for control

group 100%) was calculated. A larger OD value represented higher cell adhesion.

There were three specimens used for each test condition.

For cell proliferation tests, the cells seeded at a density of 5 × 104 cells/mL were

allowed to attach for 6 h, and then the samples were placed into new plates and

cultured for up to 7 days in 1.5 mL medium in an incubator humidified with 5% CO2

and 95% air at 37 ℃. At each culture period (2, 4, and 7 days), MTT was added to

each well and incubated at 37 ℃ for 4 h. The blue formazan product was dissolved

by DMSO, and the absorbance was measured at 570 nm using an ELISA reader. The

relate value to initial density (OD for proliferation/OD for initial attachment) was

calculated. The cell morphology was observed with SEM at an accelerating voltage of

25 kV, after fixing the cells with 2.5% glutaraldehyde, dehydrating them with a series

of graded ethanol (70, 90, and 100%), critical point drying with CO2, and gold

coating the cells.

２７

(a)

(b)

Figure 6. The photograph of (a) 96 well plate and (b) MRX microplate

reader for the MTT assay.

96 well plate

２８

3. Statistical analysis

The statistical significant difference of the results between the experimental and

control groups were analyzed using one-way Anova and Tukey statistical test at a

level of 0.05.

２９

ⅢⅢⅢⅢ. RESULTS

1. Phase identification of powders

In order to obtain stable HA sols, different concentrations (0, 2, 3, and 4 wt%) of

NH4OH was added into the HA sol. The XRD patterns of the sol-gel HA powders

after heat treatment at 500 ℃ for 2h in air, which were obtained using the sols aged

for 72 h, are shown in Figure 7. When the higher concentrations (3 and 4 wt%) of

NH4OH were added, typical apatite patterns with higher intensity appeared, but weak

apatite peaks were observed without NH4OH and with 2 wt% NH4OH. However,

when 4 wt% ammonium hydroxide was added, the HA sol was suspended.

The XRD patterns of the PH, HT10, HT30, HT50, HT70, and PT powders obtained

by sol-gel process are shown in Figure 8(a)-(f), respectively. All the diffraction peaks

matched the ICDD standard (PCPDFWIN1.30, JCPDS-ICDD). In the Figure 8(a),

XRD pattern of PH powder appeared to identical to that with main characteristic

peaks (2θ=25.9, 31.82, 32.33, 32.91 and 39.78˚) of HA. There were no secondary phases, such as TCP or CaO, detected. As can be seen in Figure 8(f), the XRD pattern

of PT powder represented the main peaks of pure TiO2 with anatase (2θ=25.22, 36.94, and 47.98˚) and rutile (2θ=27.38 and 35.98˚) structures. In the Figure 8(b)-(e), it can be seen that all patterns of composite powders (HT10, HT30, HT50, and HT70)

showed the characteristic peaks consisting of the phase of pure HA and TiO2. The

more TiO2 added into the HA and TIO2 composite sol, the higher intensity of TiO2

peaks appeared in the HA and TiO2 compound. It means that the all composite

powders were well-mixed with pure hydroxyapatite and titania.

３０

20 25 30 35 40

(d)

(c)

(b)

(a)

HA

Inte

nsit

y (A

rb. U

nit)

Diffraction angle (2θθθθ)

Figure 7. XRD patterns of hydroxyapatite sol-gel powders with (a) 0, (b) 2, (c) 3, and

4 wt % ammonium hydroxide.

３１

20 25 30 35 40

TiO2 rutile

TiO2 anatase

HA

(f)

(e)

(d)

(c)

(b)

(a)

Inte

nsit

y (A

rb. U

nit)

Diffraction angle (2θθθθ)

Figure 8. XRD patterns of the powders; (a) PH, (b) HT10, (c) HT30, (d) HT50, (e)

HT70, and (f) PT powder.

３２

The FT-IR spectrum of the powders obtained by sol-gel process is shown in Figure

9(a) PH, (b) HT30, and (c) PT powder. A weak OH band is obscured by a broad band

between 3200 and 3800 cm-1, which is characteristic of moisture in the all samples.

Absorption bands at 3567 and 633 cm-1 correspond to the symmetric stretching and

bending vibration of the OH group.47 In Figure 9(a), the absorption band at 1382 cm-1

is attributed to the presence of carbonate and/or hydrogen phosphate in the sample.48

The carbonate v3 band also is observed as three peaks (at 1648, 1454, and 1413 cm-1)

and the carbonate v2 band as a single peak (at 875 cm-1) for PH and HT30.49

The intense broad peak between 900 and 1100 cm-1 is assigned to PO43 .The

phosphate v3 and v1 bands are identified by two peaks(at 1100 and 1030 cm-1) and

peak of v4 band, the stretching and the bending modes of PO43-, appeared at 602 and

571 cm-1 as intense sharp peaks. Figure 9(c) shows that PT have broad bands in the

range of 800–400 cm-1 which can be observable in crystalline TiO2.50 With the

composition of TiO2 adding to HA, the intensities of PO43- absorption bands at 900-

1100 and 560-630 cm-1 decrease and TiO peak at 471 cm-1 appears in Figure 9(b).

３３

4000 3500 3000 2500 2000 1500 1000 500

PO4

TiOOH

PO4

OH

OH

(c)

(b)

(a)

Tra

nsm

itta

nce

(Arb

. Uni

t)

Wavenumber (cm-1)

Figure 9. FT-IR spectra of the powder obtained by sol-gel process; (a) PH, (b) HT30,

and (c) PT sol-gel powder.

３４

2. Hybrid sol-gel coatings on titanium substrates

A. Characterization of coatings on titanium

(A) Phase identification of coatings

Figure 10 shows the XRD patterns of the (a) PH, (f) PT, and (b)-(e) composite

coatings on Ti after heat treatment at 500 ℃ for 2 h in air. The XRD patterns (Figure

10(a), (f)) of the HA and TiO2 coating layers had the characteristic peaks of the pure

HA and TiO2 phase with anatase, respectively. The characteristic HA and TiO2 peaks

were well developed after heat treatment. There were no secondary phases detected.

In Figure 10(b)-(e), the composite coatings had the representative peaks of the HA

and TiO2 anatase phase.

(B) Observation of surface morphology

Figure 11 represents the SEM morphologies of the (a) PH, (f) PT, and (b)-(e)

composite coating surfaces on Ti substrate after heat treatment at 500˚ for 2h in air.

As can be seen in Figure 11, the surface of all coating layers except for TiO2 was

quite highly dense and homogeneous in microstructure. When the TiO2 amounts

adding into the HA sol increased, the morphology of the composite coatings became

more porous and rougher.

３５

20 25 30 35 40

TiTiO

2 rutile

TiO2 anatase

HA(f)

(e)

(d)

(c)

(b)

(a)

Inte

nsit

y (A

rb. U

nit)

Diffraction angle (2θθθθ)

Figure 10. XRD patterns of the coatings; (a) PH, (b) HT10, (c) HT30, (d) HT50, (e)

HT70, and (f) PT coating.

３６

(a)

(b)

(c)

Figure 11. SEM morphologies of the coating surface on Ti substrate; (a) PH, (b)

HT10, (c) HT30, (d) HT50, (e) HT70, and (f) PT coatings. Magnification 500 ×

(left), 5,000 × (right).

３７

(d)

(e)

(f)

Figure 11. (continued)

３８

(C) Measurement of surface roughness

The roughness of the coatings was analyzed with a surface profiler. Table 4 showed

the roughness of samples. The Ra values of all the coatings were in the range of 0.832

~ 0.923 µm. There was no significant difference (p>0.05) between PH and HT10

coatings. HT30, HT50, and HT 70 samples have significantly higher (p

３９

(D) Evaluation of chemical composition

Chemical composition of the coatings was observed with EDS. Figure 13 shows the

EDS spectra of the coating layer where HA sol was added. Calcium and phosphorous

were observed on the layers. In Figure 13(a)-(e), The Ca/P ratio of the PH, HT10,

HT30, and HT70 coatings was 1.73, 1.67, 1.58, 1.54, and 1.58, respectively.

(E) Measurement of thickness

The results of the coating thickness on Ti substrate can be shown in Table 5. The

average thickness of the composite layers was measured, as shown in Table 5. All the

coatings were in the range of 1.291-1.482 µm, except for pure TiO2, which was much

thinner (< 1 µm). The thickness of the coating layers decreased as the amount of TiO2

added to HA increased. In Figure 13, cross-section images of the coatings on Ti

substrate clearly depict the formation of a highly dense and homogeneous structure.

The coating layer appeared to consist of very small grains throughout the coating

layers. Moreover, there was no delamination at the interface, suggesting a tight bond

between the film and the substrate. However, the HA and TiO2 phases could not be

discerned.

４０

(a) (b)

(c) (d)

(e)

Figure 12. EDS spectra of the coating on Ti substrate; (a) PH, (b) HT10, (c) HT30, (d)

HT50, and (e) HT70 coating.

４１

Table 5. Thickness of coating layers on Ti substrate

Samples Thickness of coating layers (µm)

PH 1.482±0.104 a

HT10 1.467±0.123 ab

HT30 1.414±0.138 ab

HT50 1.368±0.092 ab

HT70 1.291±0.095 b

PT 0.908±0.072 c

Significant differences (p < 0.05) between a, b, and c.

４２

(a) (b)

(c) (d)

(e) (f)

Figure 13. SEM morphologies of the cross section of the coating layers on Ti

substrate; (a) PH, (b) HT10, (c) HT30, (d) HT50, (e) HT70, and (f) PT coating layer.

４３

B. Scratch test for bonding strength.

The adhesion strength between the coating and the substrate was evaluated using a

scratch tester. The diagram of friction versus load for the scratch test performed on

the specimens is shown in Figure 14. In Figure 14(a) and (b), the diagram for PH and

HT10 coatings represented higher slope and rougher striation than that of others from

the initial scratch load of 2N. Obviously, the HT10 coated film was scratched off the

substrate and the wear debris were found on both sides of the scratch line at the initial

stage in Figure 15(a). The EDS mappings of Ca also indicate that most of HT10

coating was removed from the specimen at the beginning of scratch, as shown in

Figure 15(b). In other words, the adhesion between HA and Ti is too weak to bear the

initial scratch load of 2N.

The diagrams of friction versus load for scratch on the HT30, HT50, HT70, and PT

coatings indicate two regions, one of a low slope with smooth striation, and the other

of a high slope with rough striation, as shown in Figure 14(c), (d), (e), and (f),

respectively. As can be seen from the diagram, TiO2 coatings had the highest adhesion

strength, and the adhesion strength between coating and substrate slightly increased to

3.151, 4.168, and 5.389N when the amount of TiO2 added into HA sol increased to 30,

50, and 70 mol%, respectively (Table 6). Consistently, SEM observations of the

scratch line also exhibit two stages in the Figure 15(b). The first, HT70-coated film

bore the wear stress and little wear debris was found (2N-5N). Secondly, HT70

coatings was crushed in advance and greatly worn away. The EDS mappings of Ca

also indicate that little HA was worn away at the first stage, but much more at the

second stage, as shown in Figure 14(d).

４４

0 5 10 150

5

10

Critical load

Load(N)

Fri

ctio

n F

orce

(N)

0 5 10 150

5

10

Critical load

Load(N)

Fri

ctio

n F

orce

(N)

(a) (b)

0 5 10 150

5

10

Critical load

Load(N)

Fri

ctio

n F

orce

(N)

0 5 10 150

5

10

Load(N)

Fri

ctio

n F

orce

(N)

Critical load

(c) (d)

0 5 10 150

5

10

Critical load

Laod(N)

Fri

ctio

n F

orce

(N)

0 5 10 150

5

10

Critical load

Load(N)

Fri

ctio

n F

orce

(N)

(e) (f)

Figure 14. Diagrams of friction force versus load for the scratch tests on the coatings;

(a) PH, (b) HT10, (c) HT30, (d) HT50, (e) HT70, and (f) PT coatings.

４５

Table 6. The Critical load of coatings on Ti substrate

Samples Critical load (N)

PH < 2 a

HT10 < 2 a

HT30 3.151±0.931 b

HT50 4.168±1.244 bc

HT70 5.389±1.465 c

PT 9.636±2.347 d

Significant differences (p < 0.05) between a, b, c, and d.

４６

(a) (b)

(c) (d)

Figure 15. Scratch track of (a) HT10 and (c) HT70, and EDS Ca-mapping of (b)

HT10 and (d) HT70.

2 N2 N2 N2 N

2 N2 N2 N2 N 5555 N N N N

４７

C. Bioactivity test using simulated body fluid.

Bioactivity of the coatings with various compositions was evaluated by immersion

of specimens in SBF for 1 and 2 weeks, and the change of the coating surface was

observed. Figure 16 shows SEM images of the coating surfaces on the Ti substrate

after immersion for 1 and 2 weeks. For 7 days, much more bone-like apatite was

deposited on PH coatings than others. However, after immersion for 14 days, we

observed apatite with needle shape on all PH, HT30, and HT70 coatings.

４８

(a) (b) (c)

(d) (e) (f)

Figure 16. SEM morphologies of the (a) PH, (b) HT30, and (c) HT70 coatings after

immersion of specimens in SBF for 1 week, and (d) PH, (e) HT30, and (f) HT70

coatings for 2 weeks.

４９

D. In vitro cell test

The in vitro cellular responses to the sol-gel-derived coatings on Ti substrate were

assessed in terms of cell attachment and proliferation.

The initial cell attachment on the films is quantified in Figure 17 after culturing for

6 h. With respect to the cell attachment on the coatings for 6 h, cell numbers attached

on all the composite coatings except HT70 have no statistical difference, compared to

that on PH coatings(p>0.05).

The proliferation levels on the films with culturing for up to 7 days is shown in

Figure 18, as assessed by an MTT method. The cells on all samples proliferated

actively with culture period. Statistically analysis revealed that there were no

significant differences between the proliferation of osteoblastic cells on the various

materials(p>0..05).

５０

HA HT10 HT30 HT50 HT70 TiO20.0

0.1

0.2

0.3

0.4

0.5

0.6

#

##

**

**

Cel

l att

achm

ent d

ensi

ty(X

103 /c

m2 )

Samples

Figure 17. MG63 cell attachment density on the coatings on Ti with culturing for 6 h.

Significant differences (p < 0.05) between * and #.

５１

HA HT10 HT30 HT50 HT70 TiO20

1

2

3

4

5

6

Cel

l Pro

lifer

atio

n(r

elat

ive

to in

itia

l den

sity

)

Samples

Day2 Day4 Day7

Figure 18. Proliferation of the MG63 cell on coatings Ti after culture for 2, 4, and 7

days.

５２

IV. DISCUSSION

Ti and Ti alloys are proven to be potentially very suitable materials for load bearing in

bioimplant applications due to their good and reliable mechanical properties.

Unfortunately, like most metals, Ti exhibits poor osteoinductive properties. From a

biochemical point of view, Ti and its alloys are considered nearly inert materials,

because they do not chemically or biologically bond with the tissues when implanted,

but they are simply surrounded by them. In this case, the bone bonding ability of the

bioactive surface is associated with the mechanical properties of the metallic

substrate.51 Therefore, the metallic implants coated with bioactive materials are of

interest in the biomedical application. Among the bioactive coating materials, HA

coatings on Ti have shown good fixation to the host bone.52,53 The improved

biocompatibility driven by the HA coatings was attributed to the chemical and

biological similarity of HA to host tissues, as well as its osteoconductivity.54 Many

studies have been carried out to develop stable HA coatings for their capacity to

establish a bond with tissues in vivo, and to promote rapid attachment and cell growth

in vitro.55-57 Various techniques, such as plasma spray58, ion beam assisted

deposition59, radiofrequency magnetron sputtering60, have been used to produce

coatings on implants. Some drawbacks have been noticed regarding the long-term

performance of the obtained coatings: coating resorption, poor mechanical properties,

high thickness, non-homogeneity, lack of adherence.61 Sol-gel processing represents

an alternative approach for the coating preparation with potential advantages, such as

higher purity and homogeneity, lower processing temperatures, reduced thickness,

simple and cheap method of preparation. Moreover, materials prepared by sol-gel

process have shown to be more bioactive than those with the same composition but

prepared with different methods. Recently, HA and TiO2, which have high chemical

affinity between HA and Ti, composite coatings have been studied to improve the low

strength of pure HA coatings. There are hydroxylic groups as Ti-OH in TiO2 sols.39

The OH groups could bond the coatings to Ti substrate more strongly. It can be

expected to strengthen the bonding ability. In this study, we fabricated HA/TiO2

hybrid coatings on Ti by sol-gel method to improve the mechanical and biological

properties.

５３

1. Preparation of HA/TiO2 hybrids sol

The sol–gel method is considered as one of the most attractive procedures because

of its low processing temperature and thin coating layer. However, the processing

condition should be carefully controlled to obtain satisfactory sol status and further

coating properties. The selection of initial precursors, appropriate chemical reactions

of sol, and final heat treatment must be considered and tailored. As a HA-coating

precursor, calcium nitrate [Ca(NO3)2] and TEP coupling was previously found to be

promising.62 Liu et al. also reported the effectiveness of TEP as a HA precursor.33

When using the precursors, the aging step is always needed to obtain relatively stable

coating layer. When aging proceeds, the mixed Ca and P precursors react and form [–

Ca–O–P–] chains by dehydration and polymerization process. However, in most cases,

the aging for prolonged periods (over 3-7 days) and high temperatures (>40 ℃) was

necessary to crystallize HA phase. More seriously, considerable amounts of secondary

phases, such as TCP, CaCO3, and CaO, were formed in the coating when the sol was

incompletely aged. The secondary phases need to be avoided since they can cause

adverse effects on the coating properties biologically and mechanically. To overcome

this, the initial sol stability should be improved by means of accelerating the [–Ca–O–

P–] chain formation, i.e., enhancing polymerization and gelation process.

When a Ca solution is added to the P solution and the mixture is aged, [–Ca–O–P–]

chains are formed, and H+ ions are provided in the solution. When the aging proceeds,

multiple chains are formed and a large amount of H+ is produced. In this study, we

used a basic additive, NH4OH to capture H+ ions. The basic additive provides a large

number of OH- ions to the solution. The OH- surely acted as capturing the released H+,

rendering the solution less acidic and more importantly accelerating the reaction in

forward direction. Moreover, since the H+ changes the chain surface to be positive

and prevent further polymerization driven by electrostatic repulsive forces.21 the

capturing of H+ ions improves the polymerization reaction effectively. In Figure 7,

when the higher concentrations (3 and 4 wt%) of NH4OH were added, typical apatite

patterns with higher intensity appeared whereas weak apatite peaks were observed

without NH4OH and with 2 wt% NH4OH. Based on the findings, i.e., the

improvement of the polymerization and gelation of the NH4OH added sol due to

５４

neutralization of H+, the higher NH4OH addition might be considered. However, the

higher addition resulted in precipitates from the solution, thus the upper limit of the

NH4OH addition was considered to be ~3% in order to obtain a clear sol.

HA/TiO2 sols were synthesized as adding various amount of TiO2 into the prepared

HA sols. Figure 8 shows the XRD patterns of all sol-gel powders after heat treatment

at 500 ℃ for 2h. In the Figure 8(a), XRD pattern of PH powder appeared to identical

to that with main characteristic peaks of HA. There were no secondary phases, such

as TCP or CaO, detected. Additionally, the XRD pattern of TiO2 powder in Figure 8(f)

represented the main peaks of pure TiO2 with anatase and rutile structures. It means

that the pure and stable HA and TiO2 sols were synthesized. In the Figure 1(b)-(e), it

can be seen that all patterns of composite powders (HT10, HT30, HT50, and HT70)

shows the characteristic peaks consisting of the phase of pure HA and TiO2 without

any by-products. The more TiO2 added into the HA and TiO2 composite sol, the

higher intensity of TiO2 peaks appeared in the HA and TiO2 compound. It means that

the all composite powders were well-mixed with pure HA and TiO2.

Van der Houwen et al. reported identification of HPO42- based on its water

absorption bands at 1210 cm-1 together with one at 1130 cm-1 (shoulder to the

absorption band at 1099 cm-1) and a clear absorption band at 879 cm-1. However,

these features are not clearly evident in the spectrum, as the absorption band due to

PO43- of HA is intense and broad. But the distinct shoulder at 868 cm-1 indicates the

presence of HPO42- in the structures. Van der Houwen et al. reported dissolving of

atmospheric CO2 yielding CO32- based on the peaks at 1456 and 873 cm-1 as the

crystallization was carried out in alkaline range (pH=10.5).63 Since the present

experiment was also done in the alkaline range similar process should also occur, but

in the IR spectrum, there is no sharp peak at 1456 and 873 cm-1. However, a close

observation of the spectrum shows a shoulder at about 1456 cm-1 in the spectrum of

the as-prepared material. However, this peak is somewhat more evident in the spectra

of sintered materials. Since the decomposition of CaCO3 is reported to occur at

580 ℃,64 carbonate stretching should be absent in the spectra of the materials taken

at 500 ℃. However, even in these spectra the peak is evident. It might be due to

chemisorption of atmospheric CO2 on the surface of HA, but there are no separate

patterns for CaCO3 in XRD spectra of the sintered materials at 500 ℃. Hence it

５５

could be expected to be present in traces in the sample. Elliot et al. reported

absorption of non-structural carbonate at the surface of HA at 1465, 1412 and 873 cm-

1.65 The Figure 9(c) shows that TiO2 have broad bands in the range of 800-400 cm-1

which can be observable in crystalline TiO2.50 In the range of 400-700 cm−1, the

adsorption band broadens into a absorption terrace. There are two reasons for the

terrace. Firstly, the dimension of the grains distributed in a range can result in the

different surface tension and the different lattice deformation, so the bond length has a

distribution range. In addition, the interface volume of nano-material holds a big ratio

and the bond of inner is different from that of interface for the surface atoms

coordination number is inadequate.47 With the composition of TiO2 adding to HA, the

intensities of PO43- absorption bands at 900-1100 and 560-630 cm-1 decrease and TiO

peak at 471 cm-1 appears as shown in Figure 9(b).

2. Properties of coatings on titanium substrate

Commercially pure Ti (grade III) disks were cut into 10 × 10 mm size as coating

substrates, and were prepared after polishing with silicon carbide paper (#1500 grit),

and then cleaning in acetone and ethanol. The prepared composite sols were dropped

onto the Ti substrate and then spin-coated at 2000 rpm for 10 s. After drying at 80 ℃

for 24 h, samples were heat treated at 500 ℃ for 2 h in air at a heating and cooling

rate of 1 ℃/min. Figure 10 shows the XRD patterns of the (a) PT, (f) PT, and (b)-(e)

composite coatings on Ti after heat treatment at 500 ℃ for 2 h in air. The XRD

patterns (Figure 10(a), (f)) of the PH and PT coating layers had the characteristic

peaks of the pure HA and TiO2 phase with anatase, respectively. The characteristic

HA and TiO2 peaks were well developed after heat treatment. There were no

secondary phases detected. In the Figure 9(b)-(e), the composite coatings had the

representative peaks of the HA and TiO2 anatase phase. No other peaks were

produced in the composite coatings suggesting a high chemical and thermal stability

of both HA and TiO2 phases.

Figure 11 represents the SEM morphologies of the all coatings on Ti after heat

treatment at 500 ℃ for 2 h in air. As can be seen, the surface of most coating layers

was quite highly dense and homogeneous in microstructure, but the SEM image of

TiO2 surface showed a cracked structure. It can be seen that a number of islands

５６

separated by large cracks on the surface. This is caused by the mechanical constraints

occurred during the heat treatment process.66 When the TiO2 amounts adding into the

HA sol increased, the morphology of the composite coatings became more porous and

rougher. This was believed to be attributable to the high content of an organic additive,

mainly the diethanolamine, which was added to preserve the stability of TiO2 sol

during hydrolysis. Thus, evaporating of organic compounds contained in TiO2 sols

causes the surface to be rougher. In the same reason, roughness of coating surface

increased with the increase in TiO2 contents. The roughness of HA/TiO2 coatings with

more than 30 mol % TiO2 sols have significant difference (p

５７

3. Scratch test for the bonding strength

The scratch test is a common technique used to evaluate the adhesion of a thin film

on the substrate. The point of adhesion failure of the coating from the substrate was

detected by a burst increase in friction force from the sample. The load at which the

first damage is made in the coating is called the “lower critical load”; and the load at

which total peeling-off of the coating from the substrate occurs is referred to as the

“upper critical load”. In hard coatings, usually the “lower critical load” is used a

measure of the adhesion strength. To evaluate the coating/substrate adhesion, we

choose to use the “upper critical load” as an indication of the adhesion strength.

The diagram of friction versus load for the scratch test performed on the specimens

is shown in Figure 14. The diagram for PH and HT10 coatings represented higher

slope and rougher striation that that of others from the initial scratch load of 2N.

Obviously, the HT10 coated film was scratched off the substrate and the wear debris

were found on both sides of the scratch line at the initial stage[Figure 14(a)]. The

EDS mappings of Ca also indicate that most of HT10 coating was removed from the

specimen at the beginning of scratch, as shown in Figure 14(a). In other words, the

adhesion between HA and Ti is too weak to bear the initial scratch load of 2N. Yen et

al. reported HA coatings on Ti was removed from the substrate at the initial 2N

loading, and the wear debris were found near the scratch line.70

The diagrams of friction versus load for scratch on the HT30, HT50, HT70, and

TiO2 coatings indicate two regions, one of a low slope with smooth striation, and the

other of a high slope with rough striation. The adhesion strength between coating and

substrate slightly increased. Among the composite coatings, the critical load of HT70

coatings was higher that that of HT10 and HT30.(p

５８

MPa) and TiO2 (70 MPa), and with increasing TiO2 content, the strength increased.

The higherst strength was approximately 56 MPa with 30 mol % TiO2 addition, and

this value was an improvement of approximately 50% with respect to pure HA

coatings.71 Liu et al. reported that the bonding strength of wollastonite/TiO2

composite coatings increased with the increase in TiO2 contents.72 Such an

improvement was to be expected, given the dense and uniform coating structure, as

well as the tight bonding of the TiO2 to both the HA and the Ti substrate. The

favorable chemical affinity of TiO2 with respect to HA as well as to Ti, i.e. its tight

bonding to both HA and Ti, greatly contributed to the observed improvement in

bonding strength.36

4. Bioactivity test

The formation of the apatite layer can be simulated in an in vitro environment by

using a simulated body fluid (SBF).42 The SBF solution, which is a metastable

calcium phosphate solution supersaturated with respect to apatite, contains inorganic

ions in concentrations corresponding to those in human blood plasma. The formed

apatite has several characteristics similar to the apatite in bone tissue and is thought to

be formed by an inorganic chemical reaction in vitro similar to that occurring in the

bone tissue. This in vitro bioactivity test generally is accepted as giving an indication

of the in vivo bioactivity.42

Bioactivity of the coatings with various compositions was evaluated by immersion

of specimens in SBF for 1 and 2 weeks, and the change of the coating surface was

observed. Figure 16 shows SEM images of the coating surfaces on the Ti substrate

after immersion for 1 and 2 weeks. For 7 days, much more bone-like apatite was

deposited on PH coatings than others. However, after immersion for 14 days, we

observed bone-like apatite with needle shape on all of the PH, HT30, and HT70

coatings. It was reported that after soaking for 7 days, bone-like apatite was observed

on composite coatings, which were prepared by HA powder and TiO2 sol40, but there

has been little study on bioactivity of the composite films prepared by both HA and

TiO2 sols. In this study, we observed apatite with needle shape not only on the pure

HA coatings but also on the hybrid coatings. It means that composite coatings made

５９

by sol-gel process show excellent bioactivity.

The process and kinetics of apatite formation on HA could be affected by bulk

factors such as density and surface area as well as by surface factors such as

composition and structure.73 The first surface structural change that the HA undergo

on soaking in SBF is the formation of a Ca-rich amorphous calcium phosphate on

their surfaces. In view of change in Ca/P ratio, the formation of Ca-rich amorphous

calcium phosphate is a consequence of interaction of the HA surface specifically with

the calcium ions in the SBF. The second surface structural change is the formation of

Ca-poor amorphous calcium phosphate, for which the HA appear to use the Ca-rich

amorphous calcium phosphate on their surfaces to interact with the phosphate ion in

the fluid. The third surface structural change is the formation of apatite. The Ca-poor

amorphous calcium phosphate on the HA appears to gradually crystallize into

bonelike apatite, through which the HA appear to stabilize their surfaces in SBF. In

addition, the bioactivity of titanium dioxide coatings has been studied in vitro and in

vivo. In the in vitro studies, Ti samples coated with sol-gel TiO2 have been soaked for

periods between 1 and 15 days in a simulated body fluid. Li et al. have shown the

precipitation of bonelike apatite at the surface of the coating. The thickness of the

layer, presenting a crystallinity very similar to that of bone apatite, may reach up to

then micrometers after two weeks in SBF.74 However, in this study, no apatite was

observed on the surface of TiO2 coatings after immersion in SBF.

5. Biological performance

Cell adhesion and cell spreading is an important parameter for implant engineering.

As the attachment of osteogenic cells onto the implant surface is the prerequisite for

osteogenic gene expression, alkaline phosphatase activity, mineralization and

consequent new bone formation at the interface, it is important to understand the

influence of materials surface properties on cell attachment and growth. Low

biomaterial efficiency is often due to poor integration of the implant with surrounding

tissue.75 The in vitro cellular responses to the sol-gel-derived coatings on Ti substrate

were assessed in terms of cell attachment and proliferation.

In practice, the effect of surface properties of biomaterials (Ti/Ti alloys,

６０

biodegradable polymers, and dense ceramics) on the cellular responses has been

studied extensively.76-77 Based on these studies, the physical properties (surface

roughness and morphology) and chemical status (crystallinity and solubility) affected

cellular responses in vitro, such as cell attachment, proliferation, and differentiation as

well as in vivo. Moreov