formation of nanotubular tio2 bioactive oxide layer on ...laporan akhirskim geran.penyeudikanasas...

48
UNlVERSm SAINS MALAYSIA Laporan Akhir Projek Penyelidikan Jangka Pendek Formation of Nanotubular TiOa Bioactive Oxide Layer on Titanium and Cell Response Studies by Assoc. Prof. Dr. Srimala Sreekantan Prof. Zainal Arifin Ahmad Mr. Ishak Mat Norehan Mokhtar 2011

Upload: others

Post on 04-Mar-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

UNlVERSm SAINS MALAYSIA

Laporan Akhir Projek Penyelidikan Jangka Pendek

Formation of Nanotubular TiOa Bioactive

Oxide Layer on Titanium and CellResponse Studies

byAssoc. Prof. Dr. Srimala Sreekantan

Prof. Zainal Arifin Ahmad

Mr. Ishak Mat

Norehan Mokhtar

2011

%

/i, bUKANU hKUa - P31K)

B

FINAL REPORTFUNDAMENTAL RESEARCH GRANT SCHEME (FRGS)

Laporan Akhir Skim Geran.PenyeUdikan Asas (FRGS) IPTPlndaam1/2010

RESEARCH TITLE

Tajuk Penyelldikan

PROJECT LEADERKetua Projek

PROJECT MEMBERS

(including GRA)

Ahli Projek

: Formation of NanotubularTi02Bioactive oxide layer on titanium and ceil response

m

Srimaia Sreekantan

1. Zainai Arifin Ahmad

2. Ishak Mat

3. Norehan Mokhtar

""'msi'P/mi• r.

I.;.-, s'•

ACHIEVEMENT PERCENTAGE

Project progress according tomilestones achieved up to thisperiod

Percentage (based on the revisedobjectives and'tnilestone)

0-50% 61 -75%

RESEARCH OUTPUT

76 -100%

Number of articles/ manuscripts/books

(Please attach the First Page ofPublication)

indexed Journal Non-Indexed Journal

1 (Journal of Nanomaterials)Prove Attachedin Appendix A

0

Conference Proceeding(Please attach the First Page ofPublication)

International National

0 0

intellectual Property(Please specify) : .-.uS... .

t - ••• • • "^zr.a

HUMAN CAPiTAI^^OfeVELOPMENT'

Human CapitalNumber . Others

(please specify)On-£loing Graduated

Citizen MalaysianNan

MalaysianMalaysian

Non

Malaysian

PhD Student -

Master Student

Undergraduate Student 2

Total 2

Budget Approved (Peruntukan diluluskan)Amount Spent (Jumlah Perbelanjaan)

Balance (Baki)Percentage of Amount Spent(Peratusan Belanja)

Account Statement is attached in Appendix B

RM 24,000.00RM 31,766.53

RIVI -7.766.53

132.00.%

ang. kepa^. pembaRayjja^^

1 internationalActivity Date (Month. Year) Organizer

(e.g : Course/ Seminar/ Symposium/Conference/ Workshop/ SiteVisit)

0

1 NationalActivity Date (Month, Year) Organizer

(e.g : Course/ Seminar/ Symposium/Conference/ Workshop/ SiteVisit)

0

We only manage to complete a portion of the proposed projecLfor.thebudget approved for this work not sufficienttocomplete ailthe objectives. The cell interaction is notinvestigatedfrrdetailfor the budget approved only 24, 000 and it iscertainly not sufficient for such studies. Therefore we jusf:manageta work on the formation of Ti02 nanotubes on Tialloys which supplied by Straits Orthopaedics.

'^Rehambahiiaikkh)- ''

Hope thebudget allocated is adequeat to conduct comprehensive work on such studies. This will allow us to understandthe fundamental knowledge lies in the project an propose mechanistic model of such cell interaction on nanotubularstructure. This ultimateoutput will contribute to the field of health science.

^AABSfRAGTf^ot More ^ 20W6x^s-(lAbsfraR:'̂ ^eMiiianM

This work focus on the formation of titanium oxide {Ti02) nanotubes on theTi-6AI-4V alloy by anodization method inorganic ethylene giycoi electrolyte (EG). The dimension of nancAubes could betuned by changing the electrochemicalparameters such as NH4F content, voltage applied; and anodizationtime. The average nanoporous or nonotubediameter and length were found to increase with increasing ambunt of fluoride, anodization voltage and anodizationtime. Minimum of 0.5 g NH4F is required for growth of nanotubes.:TT02 nanotubes with average diameter of 110 nmand 3.1 pm lengths were obtained in EG containing 0.7 g NH4F. Upon annealing from temperature 400°C to 600°C inargon atmosphere shows crystallization of the nanotubes to anatase phase exists at 400°C while rutile dominantly existsat eOO^C. Anodization in the organic electrolyte resulted in homogeneous structure unlike the one reported in aqueousacidified fluoride solution which resulted in inhomogeneous stnjcture due to the severe attack of the p-phase. For thecase of behavior of cell interaction using human foreskin fibrobfast cells line (HS27) studies shown that amorphous Ti02nanotubes structure has better cellular interaction as compare to anatase and rutile crystal structure.

The Full technical report is attached in Appendix C

24 Aug 2011 Project Leader's Signature;Tandatangan Ketua Projek

H

91^^ Nu <!hctab. UAeigwii^ 1

Name:

Nama:

Date:

Tarikh:

Signature: iTandatangan:

uo|)i>oi|qAj

IIKipudddu

('

).I

Hindawi

EQuj^^ ;/C -v',-:'-g^f

239289.V3 (Research Article)

Journal

Additional Files

Manuscript Number

Resubmitted On

Authors

Characterization and photocatalytic activity ofenhanced copper-silica loaded titania prepared viahydrothermal method

Journal of Nanomaterlals

Regular

Reply to review reports

239289.v3 (Research Article)

2011-06-23

t. '± Ramarao Poliahj t: 5 Srimala Sreekanta

t_ -5 Edward Andrew Payzant (Assigned on 2011-06-23)

Editorial Recommendation Publish Unaltered (Made On 2011-07-13)

Editorial Decision Accepted (Made On 2011-07-13)

Cha

ract

eriz

atio

nan

dph

otoc

atal

ytic

activ

ityof

enha

nced

copp

er-s

ilica

load

ed

tita

nia

prep

ared

via

hydr

othe

nnal

met

hod

Ram

arao

Polia

h',S

rim

alaS

reek

anta

n^*

^Sch

ool

ofM

ater

ials

and

Min

eral

Res

ourc

esEn

gine

erin

g,U

nive

rsiti

Sain

sM

alc^

sia,

Engi

neer

ing

Cam

pus,

Nib

ongT

ebai

,Seb

eran

gPe

raiS

elat

an,P

ulau

Pina

ng14

300,

Mal

aysia

.

Em

ail:

ram

arao

2609

@gm

ail.c

om.s

nmal

a@en

g.us

m.m

y

*Cor

resp

ondi

ngau

thor

.

—Sc

hool

ofM

ater

ials

and

Min

eral

Res

ourc

esEn

gine

erin

g,U

nive

rsiti

Sain

sM

alay

sia,

Eng

inee

ring

Cam

pus,

Nib

ong

Teba

l,Se

bera

ngPe

raiS

elat

an,P

ulau

Pina

ng14

300,

Mal

aysi

a.

Tel

:+

604

5995

255:

Fax:

.+60

459

4101

1

E-m

ail:

srim

ala@

.eng

.usm

.my

(S.S

reek

antp

n).

Ab

str

act

TiO

ina

nopo

wde

r,lo

aded

with

SiO

zan

dC

u-Si

Oz,

was

prep

ared

unde

rbo

thac

idic

and

basi

c

envi

ronm

ents

via

the

hydr

othe

imal

met

hod.

The

mor

phol

ogy

and

stru

ctur

eo

fTiO

zw

ere

stud

ied

byX

RD

,TE

Man

dFT

-IR

.The

phot

ocat

alyt

icac

tivity

ofsa

mpl

esw

asst

udie

dby

mon

itori

ngth

e

degr

adat

ion

ofm

ethy

lora

nge,

usin

ga

UV

-vis

ible

spec

trop

hoto

met

er.T

heef

fect

ofT

i/Sir

atio

,

pH,a

ndad

ditio

non

the

form

atio

nof

TiO

zand

itsph

otoc

atal

ytic

activ

ityw

ere

inve

stig

ated

inde

tail.

The

resu

ltssh

owth

ata

larg

esu

rfac

ear

eaan

da

high

sur&

ceac

idity

wer

eim

port

ant

fact

orst

oac

hiev

ego

odT

iOzp

erfo

rman

ce.

The

pres

ence

ofT

i-O

-Sib

ondi

ngen

hanc

edsu

rfec

e

acid

ity,w

hich

impr

oved

itsab

ility

toad

sorb

mor

ehy

drox

ylra

dica

lsan

din

crea

sed

itssu

rfac

e

area

.A

dditi

onof

0.1m

ol%

conc

entr

atio

nof

Cu^

*an

d2S

mol

%Si

Oz

inT

iOz

indu

ced

the

form

atio

no

fnew

stat

escl

ose

toth

eco

nduc

tion

band

,whi

chna

rrow

edth

eba

ndga

pen

ergy

and

enha

nced

the

phot

odeg

rada

tion

effi

cien

py.

:,•

.

Keyw

ords

:Tita

nia;S

ilica

;Cop

per;

Hytjf

btherm

a!;Ph

otopa

talyti

pacti

vity;

pH

1.0 Introduction

In recent years, cases of air and water pollution have been increasing due to the

continuous rise inpopulation Mdurbanization. Improper waste management has been identified

as one of the prime factors that contribute to the pollution. In 2003, the average amount of

municipal solid waste (MSW) generated in Malaysia was 0.5-0.8 kg/person/day, which has

further increased to 1.7 kg/person/day in major cities [1]. Solid waste management continues to

bea major challenge throughout the world, particularly in rapidly growing cities and towns [2].

Land filling and incineration are two of the most common methods of solid waste

disposal. Unlike land filling, incineration requires minimum land, reduces thevolume of solid

waste by half, and can be operated in any weather. Nevertheless, emission of pollutants,

especially dioxin and furan, is themajor drawback of incineration. Among thevarious solutions

available, photocatalysis is considered to be a green technique that has great potential to

decompose contaminants witiiout leaving bagful intermediates [3-5].Choiet al. reported that

the polychlorinated dibenzo-p-dioxins such as mono-, tetra-, h®Pta-, and octachlorinated

congeners weresuccessfully degraded bya TiOifilm qnder UV or So|arlightirradiation [6].

Titaniumdioxide is the most promisingphotocatalyst due to its superior properties such

as low cost, environmental compatibility, and long term photochemical stability [7-10].

However, its wide band gap energy (3.0 for rutile; 3.2 for anatase) limits its use as a

photocatalyst in various applications. The large surface area,high crystallinity, low crystallite

size, and crystal structureare important properties that influence the photccatalytic activityof

TiOj. In recent years, much efibrt has been directed towardTiOj modification to enhance its

photccatalytic activity [11-15]. For instance, TiOa loaded with various secondary oxides such

asZr02 [16], WO3, Mn02, CuO, VzOj, A^Cb [17] and transition metals saltsuch as Cu^* [18]

and Fe '̂̂ [19,20], has been reported to be a more efficient photocatalyst than pureTi02. The

composite Ti02-Si02 has attracted great interest because of its ability to prevent the grain3

growthof TiOi particlesand enhancethe thermalstabilityfor the phasetransformation ofTi02

from anatase to rutile [21-27]. Xu et al. [24] studied the effect of a wide range of SiCb levels

added to the microstructure and the photccatalytic activity of TiO: powder; preparation

involved the sol-gel methodand calcination at various temperatures. Addition of SiOz resulted

in a marked reduction in grain size and an increase in the surface area of the catalyst. Results of

other studies are summarized in Table 1. Although many reports describe the effect ofSi02 on

the photccatalytic activity of Ti02, the results do not agree with one another, possibly due to

differences in the processing methods or experimental conditions. On the other hand, Chen et al.

[18] has been reported that 0.1wt% of Cu^"*" doped Ti02-Si02 showed higher photccatalytic

activity than that of undoped Ti02-Si02. The effects of pH value on pureTiOz formation have

been reported by several groups [28,29].Yu et al. preparedTIO2 powderat differentpH values

throughthe hydrothermal method[28].They found that the pH valueof the startingsolutionhas

a significant effect on the crystallinity, crystallite size, phase structure, and photocatalytic

activity ofthe synthesized Ti02 powder. They proposer) that basic conditions are favorable for

the formation of pure anatase. Samples preparerl at pl^ 9rlispjay higjier phojocatalytic activity.

However, this result contradicts those obtained by ICaraiTii eta|. [29], who concluded that Ti02

prepared under acidic conditions (plf 3) threngh the so|-ge| method has higher photocatalytic

activity. ....

Although the sol-gel method is widely used due to its simplicity and low cost,

calcination in air is requiredfor the formation of the anatase phase from amorphous TiCb. Li et

al. [26] did a comparative study of the sol-gel and hydrothermal methods for the synthesis of

Ti02-SiO2 compositenanoparticles, and found that samples prepared through the hydrothermal

routestill possess a stableanatase phase, a largespecific surface area,a small particle size,and

a highphotocatalytic activity, evenwhen calcined at 1000 °C. However, they did notreporttlie

effectof pH on the formation of Ti02-Si02 nanoparticles. In the present snidy, the effectof pH

values on photocatalytic activity andtheoptimum Ti/Siratiofor photocatalytic activity of Ti02

powderpreparedbyhydrothermalmetliodwereexamined.TheeffectofCu-SiC)2additionon

thephasestructureandthephotocatalyticactivityofTi02werealsostudied.

Table1PhotocatalyticactivityofTiOzwiflivariousSiOicontent

Ti/SiRatioPreparationCrytallitePhasePhotocatalytic(%)MethodSize(nm)StructureActivity(%)

1:1(50%)[18]5-8

1:1(50%)[19]9h)'00-120A:100%95(RhodamineB.2h)

19:1(5%)[20]fsoo^ocih)0.7A:1Q0%73(MO.1h)3*1^25%^1211Sol-gel4.5(A) j.l(25/0)IZIJ2h)9.2(R)

7:3(30%)[22]2h)

3:2(40%)(23]7

_3.^(25%)[24]

A:100%90(MB,2h)

A;43.3%65(MO,2h)

A:|00%9;i(MO,5h)

A:100%95(MB,1h)

A:100%80(MB,8h)

Aanatase,R=rutile,MB=methylblue.MO»methylorange

2.0ExperimentalProcedure

2.1Preparation

Initially,theoxidepowderTiOi-SiOzwithvariousTi/SiratiosatpH3and9were

prepared,andthephotocatalyticperformancewasevaluatedtoselecttheoptimumTi/Siratio.

TheTi/SiratiosandpHvaluesintheexperimentaresummarizedinTable2.

Tetrabutylorthotitanate(TBOT)withanormalpurityof97%(Aldrich,US)and

tetraethylorthosilicate(TEOS)withanormalpurityof98%(Merck,Germany)wereusedasthe

sourceoftitaniumandsilicon,respectively.Bothweredissolvedinequalvolumesofethanol,

afterwhichafewdropsofhydrochloricacjdwereaddedtoTEOSascatalyst.Afterward,the

mixturewaskeptinawaterbathmaintainedat70"Cfor2h.TheTEOSandTBOTprecursors

werethenmixedandstirredfor15minutesbeforeacldingthedesiredamountofcopper(ll)

nitrate.ThepHvalueswereadjustedbyadding{lydfocjijprjcacidfortheacidiccondition,and

sodiumhydroxideforthebasiccondition.Thissolutionwasstirredfor1hatroomtemperature

andplacedinaTeflon-linedstainlesssteelautoclave,inwhichitwasheatedat150°Cfor24h,

andthencooledtoroomtemperatpre.filesatppieobtainedwaswashedandthendriedat100

"C.Apowderwasobtained.

2.2Characterizationandphotocatalyticdegradation

ThepowdersampleswerecharacterizedbyX-raydifftaction(XRD)usingaBrukerD8

powderdifftactometeremployingCuKaradiation.Theacceleratingvoltageandtheapplied

currentwere40kVand40mA,respectively.TheaveragecrystallitesizeoftheTiOi-SiOi

oxidepowderwascalculatedusingX-raylinebroadeningmethodsbasedontheScherrer

formula.Additionally,themorphologyofthepowderwasobservedbytransmissionelectron

microscopy(TEM)usingaPhilips420T.Thechemicalstructureinformationoftheparticles

was obtained by Fourier transform infrared spectroscopy (Spectrum One, Perkin Elmer, US).

Photocatalytic degradation studies were performed using 30 mg/L methyl orange (MO)

solution. A total of 0.1 g TiOi-SiOj oxide powder was added to 30 mL ofMO solution and

photoirradiated for 1h at room temperature using aTUV 18W UV-C Germicidal light The

concentration of the degraded MO was determined using a UV-visible spectrophotometer

(Varian, Gary 50 Cone).

Table2 Crystal properties ofTi02-Si02 oxide powder

Sample Ti/Si RatioSi02 Final Crystallite size/nm

SAO I: 0 0 3 6.30

SA5 19 : 1 5 3 6.63

SAID 9 1 10 3 7.06

SA16 5 1 16.67 3 8.91

SA20 4 1 2P 3 9.54

"SA25 3 1 25 3 6.88

SA2S-Cu(0.05) 3 1 25 3 10.8

SA25-Cu(0.1) 3 I 25 3 9.48

SA2S-Cu(0.15) 3 1 25 3 9.62

SA2S-Cu(0.2) 3 1 25 3 10.11

SA25-Cu(0.5) 3 1 25 3 8.21

SA25-5 3 1 25 5 8.63

SN25-7 3 1 25 7 9.98

SA30 7 3 30 3 11.78

_SA50 1 1 50 3 11.76

SBO I 0 0 9 3.75

SB16 5 1 16.67 9 -

SB20 4 1 20 9 -

SB2S 3 1 25 9 -

SB2S-12 3 1 25 12 -

SB25-14 3 1 25 14 4.57

SB30 7 3 30 9 -

SBSO I 1 50 9 -

—SA-= Acid, SB = Basic, SN = Neutral

3.0 Results and discussion

3.1 Effect ofSi02 content

XRD was used to analyze the formation of the crystalline phase of the Ti02 powder

prepared with various amounts of Si02 in both acidic and basic environments. Figure 1

illustrates the XRD patterns of Ti02 prepared at pH 3 with various Ti/Si ratios. The Ti02

prepared at various Ti/Si ratios led to formation of the anatase phase. Patterns of samples

without Si02 (SAO) shows peak for anatase and a small peak for the brookite phase at a

diffiraction angle of~20''. Therelative intensities oftheanatase diffraction peaks ineach sample

aredifferent; this suggests thatthe degree ofcrystallinity isaffected bythe Si02 content. A high

crystallinity of Ti02 was obtained with addition of 30 mol% Si02 (SA30), whereas low

crystallinity was observed insamples with 25 ino|% SjOa {SA25). No Si02 crystal phase was

identified in all samples. This indjcates that SjOj existerj as an amorphous phase in the Ti02-

Si02 composite.

The formation ofbrookite phpsc jn san]p|c SAO (^cidic condition) could be explained as

follows [28]: Ti(OH)^OC4H9)+« a r®?Mlt ofjiydfolysis reaction between TBOT and

HCI where * was related to therH vajue ofstarting sojution. Since the ligand field strength of

Cr ions was larger than that of butojq' group inHCI solution, thus the CI* ions could substitute

tlie butoxy group in the Ti(OH)x(OC4H9)4.r complex and led to the formation of complex

Ti(OH)2Cl2. The complex Ti(OH)2Cl2 actually existed in the form of Ti(0H)2Cl2(H20)2 in a

solution due to theTi(IV)(3d°) complex ions are all octahedrally coordinated in solution and

crystal. It is reported that Ti(0H)2Cl2(H20)2 could be the precursor of brookite. This

mechanism was favored to explaintheoccurrence of brookitein thesample.

The XRDpatterns of Ti02 prepared at pH 9 with various Ti/Si ratios (notshown here)

shows samples prepared without Si02 (SBO) produced peaks characteristic of anatase at 26 of

24,34,39, and 48°, and thepeak ofmtile at 28°. However, thepeak intensity of this sample is8

very low compared to those produced by samples prepared under acidic conditions; this

indicates that the phase transformation to anatase or rutile was not fully achieved at basic

conditions. In contrast, samples prepared with SiOr generallyshowed amorphous Ti02. In basic

conditions,Na^ attacks tire Ti-O-Ti bond, which results in the formationofa two-dimensional

layered structure with dangling bonds that contributeto the formationof the amorphous phase.

This-is in good agreement with the TEM analysiswhichis discussedlater.

The crystallite size of the Ti02-Si0z powders prepared in acidic condition was

calculated using Scherrer formula, and plottedagainstSiOzcontent Figure2 shows a non-linear

relationship between the crystallitesize and the amountof added SiOz. The trend is contrary to

results in the literature wherein SiOz addition was found to retard grain growth of TiOz and

therefore reduce its crystallite size [23,24,26]. In the present study, the lowest crystallite size

was 6.3 nm for pure TiOz (SAO), and increased gradually wifh further addition of SiOz.

However, the crystallite size ofTiOz for sample SA2S decreased drastically as the SiOz amount

reached 25 moi%; this may be attributed to the large surface atea of the sample. Binary metal

oxides areknown to induce surface acidity [30]. TheFTTR spectra of theTiOz-SiOz powder at

various SiOz contents are shown in Figure 3. When the SiOz corttent reached 25 moI%, a small

peak at 960 cm*' corresponding to theTi-O-Si bond was observed. Thepresence of thisoxide

miglit have increased the surface acidity of the sample. Higher sur&ce acidity led to a higher

degree of adsorption of theOH radicals (1640 cm*') and resulted ina larger sur&ce area. The

sur&ce area was inversely proportional to grain size. The Ti-O-Si band intensity increased with

addition of 30-50 mol% SiOz, and the intensity of bands for the hydro}tyl group decreased.

These indicate that large amounts of SiOz(30-50 mol%) in TiOzcannot effectively improve the

surface acidity and prevent the growth of TiOz grains. Therefore, in sample SA25, SiOz

effectively enhanced the surface acidity, suppressed &e grain growth of TiOz,. und produced

"smaller crystallite size. Tlie peak fortheasymmetric stretching of Si-O-Si at 1060 cm*' clearly

increased with addition of SiOz, whereas theband intensity at 400-700 cm*' (vibration ofTi-0

9

bonds in Ti-O-Ti bonding)decreased.Therefore, the intensity of the peak for the Si-O-Si bond

was inverselyproportional to the intensityof the peak for the Ti-O-Ti bond.

40 50

Oegroe (2theta)

Figure 1 XRD patterns ofthe TiOz povvder prepared with different amounts ofSiOz at

pH 3; (a) 0 mol%, (b) 5 mol%, (c) 10 mo|%, (d) 16 mo|%, (e) 20 mol%, (0 25 mol%,

(g) 30 moI%, (h) 50 mol%. (A: anatase, B; brookite)

10 20 30

Si02 content (mol%)

Figure 2 Relationship betweenthe crystallitesize and amount of SiOzon TiOzpowder

prepared at pH 3.

10

OH

gro

uS8

0rn

-o-T

i)

35

00

30

00

25

00

20

00

15

00

10

00

50

0

Wav

enu

mb

er(c

m'M

Figu

re3

FTIR

spec

traof

TiOi

-SiO

zoxi

depo

wder

cont

ainin

gSiC

h:(a

)0in

ol%

,(b

)5m

ol%,(

c)10

mol%

,(d)

16m

ol%,(

e)20

mo!%

,(f)

25m

ol%,(

g)30

mol%

,(h)

50m

ol%.

Ain

dica

tesab

sorp

tion

band

sat9

60cm

"',w

hich

isch

arac

teris

ticof

Ti-O

-Sib

ondi

ng.

3.2

Eff

ect

ofp

Hva

lue

Figu

re4

show

sth

eXR

Dpa

ttern

ofTi

Ozpo

wder

with

25m

ol%

SiO

i(Ti

02-2

5ino

l%

Si02

)atv

ario

uslev

elsof

pH.I

nth

epre

senc

eofS

iO:,

anac

idic

cond

ition

wasf

evor

able

tofo

rm

anat

ase

struc

ture

.A

sthe

syste

mis

shift

edfro

mac

idic

tone

utra

l,th

ein

tens

ityof

thea

nata

se

peak

(20

=25

°)an

dth

ecr

ystal

lite

size

incr

ease

d.Th

us,p

Hva

lue

also

affe

cted

the

degr

eeof

crys

tallin

ityde

spite

the

pres

ence

ofSi

Oj.

AtpH

12,t

hecr

ystal

struc

ture

ofsa

mpl

eSB

25-1

2

coul

dno

tcor

resp

ond

toan

atase

,ru

tile

orbr

ooki

te.It

was

simila

rto

that

ofN

a207

Ti3[

31],

-H2n

024S

i3Ti4

and

H4Na

40i6S

i4Ti2

prob

ably

due

toth

eir

sam

ela

yere

dtit

anat

efa

mily

[32]

.

Furth

erin

crea

sein

basic

cond

ition

(pH

14)r

esul

ted

info

rmat

ion

ofth

epar

tiala

nata

seph

ase,

as

seen

inth

eTEM

imag

esof

TiC)

2-25

mol%

Si02

pow

ders

prep

ared

atva

rious

pHva

lues

inFi

gure

5.Sp

heric

allik

epa

rticle

s(F

igur

e5a

and

b)we

reob

serv

edin

sam

ples

prep

ared

inac

idic

and

11

neut

ralc

ondi

tions

.Atp

H12

,ala

mel

lars

truct

ure

(two-

dim

ensio

nal

laye

red

struc

ture

s)fo

rmed

asa

resu

ltof

ther

eact

ion

betw

een

thes

ampl

ean

dNaO

H(F

igur

e5c

),w

hich

invo

lved

atta

ckof

Na"*"

onth

eTi

-O-T

ibo

nd.

The

edge

sof

the

lam

ella

rst

ruct

ure

mig

htha

vem

any

atom

sw

ith

dang

ling

bond

sw

ithen

ough

ener

gyto

dest

abili

zeth

etw

o-di

men

siona

lstru

ctur

es[3

3].A

sthe

syste

msh

ifted

tow

ard

high

pH,t

hela

mel

larT

i02

defo

rmed

tosa

tura

teth

eda

nglin

gbo

nds.

At

pH14

,the

trans

ition

thre

e--»

•tw

o-on

edi

men

sion

was

alm

ostc

ompl

ete,

whi

leth

elam

ella

r

struc

ture

form

edtu

bes

(Fig

ure

5d);

tliis

resu

lted

ina

parti

alan

atas

eph

ase.

Thus

,th

epH

appe

ared

toha

vea

sign

ifica

ntef

fect

onTi

02ph

ase

stru

ctur

ean

dm

orph

olog

y.

Figu

re6

illus

trate

sth

eFT

IRtra

nsm

issio

nsp

ectra

ofTi

Or2

5m

ol%

Si02

pow

der

prep

ared

atva

rious

pHva

lues

and

auto

clave

dat

150

°Cfo

r24

h.Th

esp

ectru

mof

pure

Ti02

is

also

includ

edas

refere

nce.

The

broa

dpe

aks

at34

00cm

"'an

dthe

peak

sat

1640

cm'i

nall

spec

traar

eatt

ribut

edto

surfa

ce-a

dsor

bed

wate

rand

tlie

bend

ing

mod

eofh

ydro

xylg

roup

s.Th

e

surfa

ce-a

dsor

bed

water

and

hydr

oxyl

grou

pde

creas

edsji

gl]tly

asthe

pHinc

rease

d;thi

sis

poss

ibly

due

toth

ere

duct

ion

insii

rface

afc^

ofth

esa

mpl

e,w

hich

prev

ented

furth

erw

ater

vapo

r

abso

rption

.This

isco

nsist

entv

vj|lit

heres

pjtsw

herei

nsam

plesp

mparc

dat

pH3

had

highe

r

phot

ocata

lytic

activ

ityco

mpa

red

with

thos

epr

epar

edat

pH7.

The

peak

s,wh

ichwe

reab

sent

in

pure

Ti02

,app

eared

at96

0cm

''an

d10

40-1

070

cm*'

due

tothe

Ti-O

-Sis

tretch

ingan

dthe

asym

metr

icstr

etch

ing

vibr

ation

ofSi

-OSi

,res

pecti

vely

.The

sepe

aks

grad

ually

decr

ease

dan

d

even

tual

lydi

sapp

eare

dat

high

erpH

.Thi

sim

plies

that

the

Ti-O

-Sib

onds

wea

kene

das

the

pH

increa

sed.

Thes

tretch

ing

vibrat

ionof

Ti-0

bond

sinT

i-O-T

ican

beob

serv

edat

400-

700

cm"'.

Peak

sre

lated

toca

rbox

ylgr

oups

were

obse

rved

atthe

1340

-147

0cm

"'ran

ge.T

heca

rbox

yl

grou

psm

ight

bea

resu

ltof

oxid

ation

ofor

gani

cspe

ciesd

urin

ghy

drot

herm

altre

atmen

t[28

].A

s

the

prep

arati

onco

nditi

onss

hifte

dfro

mac

idic

toba

sic,t

heox

idati

onpr

oces

swas

inhi

bited

;thi

s

caus

edth

epe

aks

todi

sapp

earc

ompl

etel

y.

12

Degree (2 theta)

Figure 4 XRD pattern of 25 mol%SiOj-loadedTiOzpowderat various pH values:

(a) pH 14, (b)pH 12, (c)pH9,(d)pH7,(e)pH5,(f)pH3.(A; anai^e)

(c) r ...m

....

m.m

.t>t'V^

SOnm

Figure 5 TEM Imagesof Ti02-25% SiOi oxide powderpreparedat variouspHvalues:

(a) pH 3, (b) pH 7, (c) pH 12, (d) pH 14.

f&'Sll

OH

grou

p

3S

00

30

00

25

00

20

00

1500

Wav

en

um

bers

10

00

50

0

Figu

re6

FTIR

spec

traof

TiO

r25%

Si02

oxide

powd

erpr

epare

datv

ariou

spH

value

s:(a

)pur

eTi

Oi(

pH3)

.(b)p

H3,

(c)p

H5.

(d)p

H7,

(e)p

H9,

(f)p

H12

.A

indi

cates

abso

rptio

nba

ndsa

t960

cm"',

whi

chis

char

acter

istic

ofTi

-O-S

ibo

ndin

g.

3.3

Effe

ctof

Cu^"^

addi

tion

inTi

Oi-S

102

pwcj

ef

The

XRD

patte

rnso

fCu^

*-lo

aded

Ti02

-Si0

2ar

esho

wn

inFi

gure

7.Th

eTi

Oi-S

iCb

load

edwi

thva

rious

amou

ntso

fCu^

^m

aintai

ned

anan

atase

phas

e!Th

epre

senc

eofC

u^*w

as

hard

lyde

tect

edby

XRD

due

toits

low

cont

ent.

How

ever

,the

rela

tive

inte

nsiti

esof

the

anat

ase

peak

vary

amon

gsam

ples.

High

TiOj

ciysta

llinity

waso

btaine

dwith

addit

ionof

0.2mo

l%Cu

^*

_and

,low

erde

gree

sofc

rysta

llini

tyw

ere

obse

rved

insa

mpl

esco

ntai

ning

0.5m

ol%

copp

er,w

hich

wer

edu

eto

the

poor

distr

ibut

ion

ofCu

^^in

theT

i02

matr

ix.T

heTi

Oiw

ithou

tCu

prod

uced

relat

ively

small

diffr

actio

npe

akso

fana

tase

at25

°(10

1),i

nco

ntra

stto

the

patte

rnso

fthe

sample

swith

Cu^"^

.TTie

refore

,pha

setra

nsfor

matio

nwas

notf

ullya

chiev

edwi

thout

Cu^"*"

,asthe

Ti0

2re

tain

edpo

rtio

nsof

the

inac

tive

amor

phou

spha

se.

15

i

40

so

Dag

rae

(2th

«ta

)

Figu

re7

XR

Dpa

ttern

of25

%Si

02-lo

aded

Ti0

2pow

derw

ithva

rious

copp

erco

nten

ts:

(a)0

mol

%,(

b)0.

05m

ol%

,(c)

0.1

mol

%,(

d)0.

15m

ol%

,(e)

0.2

mol

%,(

f)0.

5m

ol%

.

(A:

anat

ase)

3.4

Phot

ocat

alyt

icac

tivity

The

photo

catal

ytic

activ

ityof

JjieTj

Qi-S

iOjp

qwde

r,pr

epare

dun

dera

cidic

and

basic

cond

ition

sw

asev

alua

ted

thro

ugh

thed

egra

datio

nof

tpef

hylo

rang

e(M

O)u

nder

ultra

viol

etlig

ht

irrad

iatio

nfo

r1h

atro

omte

mpe

ratu

re.E

valu

atio

nw

asre

peat

edfo

r3tim

esan

dave

rage

valu

es

wer

epl

otte

d.Sa

mpl

espr

epar

edin

acid

icco

nditi

ons

disp

laye

dph

otoc

atal

ytic

activ

ityhi

gher

than

that

ofsa

mpl

espr

epar

edin

basic

cond

ition

s(Fi

gure

8).T

his

isat

tribu

ted

toth

epre

senc

eof

anat

ase

inth

esam

ples

prep

ared

unde

raci

dic

cond

ition

s.Ph

otoc

atal

ytic

activ

ities

ofth

esam

ples

with

diff

eren

tTi/S

irat

iova

ry.

Pure

Ti0

2(SA

O)

prep

ared

inac

idic

cond

ition

sre

sulte

din

high

phot

ocat

alyt

icac

tivity

.H

owev

er,a

dditi

onof

25an

d50

mol

%of

Si02

inTi

Oi

prod

uced

the

high

est

phot

ocat

alyt

icac

tivity

com

pare

dw

ithpu

reTi

02.

Thus

,th

epr

esen

ceof

SiO

aw

as

esse

ntia

lto

high

erph

otoc

atal

ytic

activ

ity.A

sthe

Si02

pres

enti

nTi

02pr

omot

eda

high

degr

ee

ofan

atas

eci

ysta

llini

ty,a

low

ercr

ysta

llite

size

was

prod

uced

due

tosu

ppre

ssio

nof

TiO

agr

ain

16

growth. This might have contributed to die larger sur&ce area of the Ti02 particles. The

addition of SiOz in TiOz also enhanced the acidity of the binaiy oxide. A model has been

proposed to explain this increase in acidity [34]. In this model, the silicon cation enters the

lattice of the host oxide, TiOz. and retains its original coordination number. Since the silicon

cation is still bonded to same number of oxygen atoms despite coordination changes in the

oxygen atoms, a charge imbalance is created. The charge imbalance must be satisfied. Thus,

Lewis sitesare expected to form dueto the positive charge in the TiOj-SiOz. Thesurface with

improved acidity canadsorb more OHradicals, thus resulted ina larger surfece area. Thismight

enhanced the photocatalytic activity and ledto complete degradation of MO. This is consistent

with the FTIR results of the sample (pH 3) possessed high intensity of hydroxyl group. The

result also indicates that sample prepared atbasic conditions ^howpd the poor h40 degradation,

which is in agreement with theobservations of Yuet a|. [23]. Tj^ey revealed that amorphous

""TiOz has a lower photocataljrtic activity compared with ciystalline TiOz- Hence, SiOrloaded

TiOz powder prepared under basic conditions isinefficient fpthjg]) photodegradation.

To understand the degradation of MO with time, samples were irradiated and collected

every IS min.PureTiOz showed rapiddegradation at thebeginningwhich eventually remained

constant after 15 min (Figure 9). This is due to the low anatase poncentration of TiOz, and its

inherent structure and low surface area. The TiOz with 25 and 50 mol% SiOz had low

^degradation rate during the first 15 min. Afterward, degradation markedly increased; both

samples completely degraded MO within 1 h. It is worth to note that the complete

mineralization of MO in sample SA25 was faster than that of sample SA50. The TiOz with 25

mol% SiOz exhibited the highest photocatalytic activity, which can be attributed to the

combination of several factors including large sur&ce area, high degree ofanatase ciystallinity,

and improved sur&ce acidity. Hence, an optimumamountof SiOz(25 moI%)added to TiOz

systems may be essential to enhance the photocatalyticactivity.

17

Since SA25 was found to be the best sample for MO degradation, this was used to

investigate the effect of pH on photocatalyticactivity. The TiOz-SiOz powder prepared at pH 3

and pH 5 showed the highest photocatalytic activity (Figure 10). Rapid degradation rates were

observed and resulted in nearly 100% degradation of MO. In contrast, the photocatalytic

activity ofsamples prepared at pH 7,9, and 12 were very low. Samples from neutral conditions

(pH 7) had the anatase structure and high ciystallinity, but showed lower photocatalyticactivity.

As the pH increased from 3 to 7, the crystallite size increased while the surfece area decreased,

which resulted in lower photocatalytic activity. These observations also indicate that a neutral

environment for the synthesis could not enhance the photocatalytic activity. Overall, TiOz

powder loaded with 25 mol% SiOz prepared at pH 3 was the sample with the highest

photocatalyticactivity.

To investigate the effect of op photocatalytic activity, SA25 was loaded with

vaiying amounts ofCu^*. The photocatalytic activity was piarkedly enhanced in the presence of

Cu^* (Figure 11). About 97% ofMO was mineralized with the addition of0.1 mol% ofCu^^

This is mainly due to tlie high degree qfciystallinity ofthe samples and the capability ofCu^* to

induce the formation of new states close to the conduction band, which leads to reduction of the

band gap energy and an increase in the photodegradation efficiency. However, as the Cu^*^

increased, the photocatalytic activity decreased rapidly. The photocatalytic activitydropped to

~21% when the Cu '̂̂ content reached 0.5mol%. Thismaybe attributed tothe lowcrystallinity

and the behavior ofCu^"^ at high concentrations, wherein it becomes a recombination center for

the photo-induced electrons and holes thereby inhibits photocatalysis [20]. Hence, an optimum

amount of Cu^"^ inTiOz-SiOz isessential toimprove itsphotodegradation efficiency.

18

r

10

01

40

20

10IS

20

253

03

5

Sil

ica

con

ten

t(ni

ol%

)

40

45

50

Figu

re8

Phot

odeg

rada

tion

ofM

O(1

hr)o

nTi

Oz-

SiO

ipow

derp

repa

red

at(a

)pH

3,

(b)p

H9

.

01

53

04

5

Deg

rada

tion

time

(min

utes

)

Figu

re9

Phot

odeg

rada

tion

ofM

Oon

Ti0

2-Si

02w

ithva

rious

Sico

nten

t,pre

pare

dat

pH3:

(a)

0%Si

.(b)

16%

Si.(

c)25

%Si

,(d)

50%

Si. 1

9

*-n

40

M2

0-

15

30

45

Deg

rada

tion

tim

e(m

inut

es)

Figu

re10

Phot

odeg

rada

tion

ofM

Oon

Ti0

2-25

mo!

%Si

Oip

owde

rpre

pare

dat

vari

ous

pHva

lues

:a)

pH3,

b)pH

5,c)

pH7,

d)pH

9.e)

pH12

.

00.

10

.2O

J

00^*

*^co

nten

t(m

ol%

)

Figu

re11

Phot

odeg

rada

tion

ofM

O(1

h)on

Cu'̂̂-

loade

dTi

02-2

5m

oI%

Si02

pow

der

prep

ared

atpH

3.

20

4.0 Conclusion

The efifect of SiOz content and pH value on the TiOi photocataljrtic activity was

investigated. Addition of SiOzinto the TiOzstrongly affected the degree of crystallinityof the

composite.The phase structure was dependenton pH value in the presenceofSiOj. An acidic

environment led to anatase phase formation, and samplesprepared in basicconditionsexhibited

the amorphous phase. Addition of SiOa improved the surfece area of TiOz particles

enhancing tlie surface acidity; this led to high photocatalyticactivity compared widi pure TiOj.

A higher degree of MO degradation was produced by Ti02 saixtples with 25 mol% SiOi

prepared at pH 3. The Ti02 loaded with 0.1 moi% 00'"^ and 25 mol% SiOz exhibited

photocatalytic activity higher tlian that without Cu^^ Therefore, the optimum amount of SiOz

and €0^"^ and the pHduring preparation were essential fo ^cjiievjpg high photocatalytic TiOz

activity.

Acknowledgments

The authors would like to thank the Universiti Sains Malaysia Fundamental Research Grant

Scheme (Grant No. 6071195) and die USM Fellowship for their sponsorship.

21

References

[1] S. Kathirvale, M.N.MuhdYunus,K. Sopian,A.H. Samsuddin, Renewable Energy.29,

559(2003)

PJ L.A. ManaC M.AA. Samah,N.I.M.Zukki, Waste Management. 29,2902 (2009)

P] J.L. Graham,C.B.AlmquisI; S. Kumar,S. Sidhu,Catai Today. 88,73 (2003)

[4] Y. Ide, K. Kashiwabara, S. Okada,T. Mori, M. Hara, Chemosphere 32, 189(1996)

[5] R. Weber, T. Sakurai, H. Hagenmaien, Appl. Catai B: Environmentai 20,249 (1999)

[6] W. Choi, S.J. Hong, Y.S. Chang, Y. Cho, Environ^SctTechnol. 34,4810 (2000)

[7] A.L.Linsebigler, G. Lu,J.T. YatesJr., Chem. Rev. 95,735 (1995)

[8] K. Hashimoto, H.Irie, A.Fujishima, Jpn^. J. Appi Phys. 44,8269 (2005)

[9] X. Chen, S.S. Mao, C/rem. J?ev. 107,2891 (2007)

b.- ...

[10] X. Chen, CA/rt. J. CatoA 30,839 (2009)

[11] M. Bellardita,M. Addamo,^ D| Paoja,G. Nlarcj, (w. Falmisano, j^. Gassar,M. Borsa,

J. Hazardous Materials (^IQ) -

[12] J. Bennani, R. Dillert, T.M. Gesing, D. Bahnemann,Separation and Purification

Technology 61,173 (2009)

[13] Y. Arai, K. Tanaka, A.L. Klilaifat,J. MoL Catai A 243,85 (2006)

[14] RA. Aziz, I. Sopyan, Indian J. Chem.48,951 (2009)

[15] RN. Viswanath,S. Ramasamy,ColloidsandSurfaces. A 133,49 (1998)

[16] K.Y. Jung, S.B. Park, Mater. Utt. 58,2897 (2004)

22

[17] S.R. Kumar, C.Suresh, A.K. Vasudevan, N.R. Suja, P.Mukundan, K.G.K. Warner,

Mater. LetL 38,161 (1999)

[18] R.F. Chen, C.X. Zhang, J. Deng, G.Q. Song, InL J. Minerals, Metallurgy and Materials

16,220(2009)

[19] H. Chun,T. Yuchao, T. Hongxiao, Catal. Today. 90,325 (2004)

[20] D. Zhang, TransitionMet. Chem. 35,933 (2010)

[21] C.H.Kwon, J.H Kim, I.S.Jung, H. Shin,K.H. Yoon K H, Ceram. Inter.29,851 (2003)

[22] L.Zhou,S. Yan, B. Tian, J. Zhang, M. Anpo, Mater.Lett 60,396 (2006)

[23] J.G. Yu, J.C. Yu, X. Zhao, J. Sol-GelSci. 24,95 (2002) • n .

[24] G.Xu,Z.Zheng,Y. Wu,N.p?ngN,'Cera/»./«ter.35, l(2QQ9)

[25] P. Cheng, M. Zheng, Y. Jin, Q. Huang, M. Gu, Mater.Left.57,2989 (2003)

[26] Z. Li, B. Hou, Y. Xu, D. Wu, Y. Sun, W. Hu, F- Pang F, J. Solid Siate Chem. 178, 1395

(2005) ' •-

[27] M.Hirano, K. Ota, J. Am, Ceram. Soc. 87, 1567(2004)

[28] J.G. Yu, Y. Su, B. Cheng, M. Zhou, J. Mol. Catal. A 258,104 (2006)

[29]^ A. Karami, J. Iran, Chem. Soc. 7,154 (2010)

[30] K. Guan, B. Lu, Y. Yin, Surface and Coating Technology 173,219 (2003)

[31] J.G. Yu, H. Yu, B. Cheng, C.Trapalis, J. Mol. Catal. A 249, 135 (2006)

[32] J.G. Yu, H. Yu, B. Cheng, X. Zhao, Q. Zhang, y. Photochem. Photobiol. A 182,121

(2006)

23

[33] Y.Q. Wang, G.Q. Hu, X.F. Duan, H.L. Sun, Q.K. Xue, Chem. Phys. Lett. 365,427 (2002)

[34] K. Guan,Surface and CoatingTechnology 191,155(2005)

24

: (

j • i:.-

7 . •

flppendiK B

flceounl Kcilemenl

'eboo* OA I1 a

Purchase Requlsttton • Purchase Order

UseiCoda: HAIZAN / USMKCTUVE / PBAHAN

Current Date: 26/08/20U 11:13:39 AM

P Suppliers > Maintenance

Proeratn Code: Votebook9100

Version: 13.92r Updated at 30/05/2011

Rnandals P Coda Info P Reports

Current Program: Votebook (Header)

Admin

DB: 13.00, 09/18/2010 VB: 13.01, 03/14/2011 Switch Language: English /Malay

Wlidcaid: eg. Uke 100%, Like10%1, Uke %1

Hement i: Element 2: % Element 4: Ipbahan

Elements: 6071195 Yean 2011

Detail Excel 116

116

Detail Excel 117

Ddall Excel 117

Detail Excel 117

117

9999

BudgetControl

Account Description Budget Account Code Rollover Budget Cash Received Advanced . Commit Actual Available Percentage

T Penyefidikan Rindamentals (FGRS) 203.111.0.PBAHAN.6071195 3,859.95 0.00 0.00 0.00 0.00 0.00 3,859.95 0.00%

T SubTotal 3,859.95 0.00 0.00 0.00 0.00 0.00 3,859.95 0.00%

T Penyelidikan Fundamentals(FGRS) 203.22L0.PBAHAN.6071195 -532.00 0.00 aoo 0.00 0.00 0.00 •532.00 0.00%

T Penyeiidikan Fundamentals (FGRS) 203.227.0.PSAHAN.6071195 -3,892.40 0.00 0.00 0.00 6,378.50 5,244.00 -15,514.90 0.00%

T Penyelidikan Fundamentals (FGRS) 203,^9,0.PBAHAN.6O71195 4,870.42 0.00 0.00 0.00 0.00 450.00 4,420.42 0.00%

T SubTotal 446.02 0.00 0.00 0.00 M7i5p 5,0M.ap -11,626.48 0.00%

GrandTot^ 4,305.97 0.00 0.00 0,00 6,378,50 5,691.pp -7,766.53 0.00%

http://10203.18.3/eprocurenient.kctA^otebook9100A_LIST.asi«K

flppendiH C

Tochniecil Reporl

Introduction

Final Technical Report FRGS(1 Mei 2010-Aug 2011)

• Titanium (Ti) and its alloys have be^. wiWy. used; as important biomaterials formany years. In particular, Ti-6A1-4V is widely iis^ as,a;bi6material for artificial hip or dentalimplants because ofits special mechanical, cHemicai-^d biological properties. Itcan providedirect physical bonding with adjacent bone snrfece without forming a fibrous tissue interfacelayer. Although the success rates ofimplant are vei^ high,: implants occasionally loosen andfail. PureTi metal lacks desirable bioactive (bonerg!:owth),properties, a thin TiOz passivationlayer forms on Ti surface to impart bioactivity and chemicalilionding to bone. Such a layer,however, is composed ofsmooth and dense Ti02 layer and" is>susceptible to the formation offibrous tissue that prohibits osteoblastic cells from firmly attaching onto the surface, and cancause the loosening of implant and inflammation (Yu et al., 2010). Thus surfacemodifications, altering topography and chemical composition-of these materials, are requiredto enhance direct structural and functional anchoring of the prostliesis to the living bone(osseointegration). The most common surface treatments of commercially available implantsand prosthesis encompass pickling, sandblasting, plasma spraying, anodizing, and micro-arcoxidation. For decades, until recently, cell interactions and bone response were studied onimplant surfaces with micrometer-scale topography. It have shown that nanoscale topographyas well as the order of the nanofeatures organisation is a critical factor romoting early cellresponses to the implant surface (Das et al., 2009). Hence the fabrication of highly orderedTi02 nanotube films for biomedical applications is attracting much attention. Among theseprocesses, electrochemical anodization of titanium in fluorinated electrolytes is a relativelysimple method to synthesize porous or tubular structures. In this work, anodization processwas performed in ethylene glycol containing fluoride and de-ionized water to produce thenanotubes structure. Then heat treatment was conducted to produce various crystallinestructure which are anatase and rutile phase ofTi02. The final stage was on testing the cellinteraction on the different crystallinesTi02 nanotubes.

Problem statement

Ti-6A1-4V alloys possessed two-phase of alpha and beta structures which arealuminum and vanadium, respectively. It has been reported that this alloy associated withlong term health problem due to toxicity ofaluminum and vanadium released. Titaniumbased alloy tends to undergo severe wearwhen rubbed ^th dther metals, it is also stated thatthe two-phase equiaxed microstructure more susceptible to corrosion as the compositionaldifference across the grain boundaries increases which leads to the galvanic cell formation(Geetha et al., 2009). Thus, surface treatment modification such as formation of thick Ti02nanotubes that protects the surface of implants as well as giving good adhesion with cell hasbeen applied on titanium and titanium alloys to increase the longevity of implant. Surfacemicrostructure is an important factor in modulating cell response at the implant/boneinterface. In the formation ofTi02 nanotubes highly dependent on the electrolyte compositionand electrochemical conditions. There are not many literatures reports on the formation ofTi02 nanotubes of Ti alloys in organic electrolytes especially in ethylene glycol. It isanticipated that ifthe morphology is well defined, the cell interaction study will be improveddue to the improvement of the interaction of nanotubes with cells. Therefore in this workcomprehensive investigation on the formation of nanotubes on Ti alloy was reported.

Objectives of this project

• The objectives of this research are as follows:• To grow TiO: nanotubes on Ti-6A1-4V alloy via anodization method.• To study the effect of NH4F content, applied voltage, and anodization time on

the formation of TiOi nanotubes on Ti-6A1-4V alloy.• To characterize crystal structure phase and morphology ofTi02 nanotubes.• To study the response of human foreskin fibroblast (HS27) cell on optimized

TIO2 nanotubes with different crystalline structure.

Results and Discussionsy

Morphology of Ti-6Al-4V

Prior to anodization experiment, Ti alloysheet was chemically cleaned by solution of1 M HNO3 and 0.45 M NH4F to remove native oxide layer on the surface of Ti-6A1-4V.Figure 1 shows FESEM images ofTi-6A1-4V after the cleaning process. As seen, the surfaceof the Ti alloy is rough and irregular with two different distinct regions. Certain regioncomposed of small grain while the others are large. EDX analysis was used to test thedifference within this two regions and it wasfound the small grains are enriched with V whilethe big grains are composed of Al.

0-phases

tsvaui

a-phoses

Mags 3.OQKX ^ 'iVB • fl-DTi EMf • SXO.Vv SgnsIA • InLeni

C-aie 8 iai 2011 Tri» 1747 09

Figure 1: FESEM Image showing the TI6Ai4V alloy sheet ofafter the cleaning process (a andp phases are an-noted with arrows; a-phases are dark; p-phases are bright). The imageacquired after ultrasonlcally cleaned in IM HNO3 and 0.45 M NH4F for 1 minute. The pphases are enriched In V.Anodization ofTi-6Al-4Vin ethylene glycol

Effect ofNH4F content onthe anodization of Ti alloys inethylene glycol

In this part, the experiment was conducted to investigate the effect of differentamount of NH4F on the morphologies of the nanotubes. Hence optimum weight of NH4F toproduce self organized nanotubes can be determined. All samples are anodized in 99 ml

ethylene glycol and 1ml H2O at 60V for 1hour and were annealed at400 °C. This parameterwas selected based on our preliminary studies (Sreekantan et a)., 2010). The content of NH4Fwas varied from 0.1 g, 0.3 g, 0.5 g and 0.7 g. The surface morphology of Ti-6A!-4V sheetanodized in ethylene glycol with different fluoride content lead to different morphologies(Figure 2). Anodization of Ti alloy in electrolyte containing O.lg NH4F gives an irregularnanoporous structure (Figure 2a). There are areas covered with small pits while other havelarger pores. The pits or pores are formed due. to the different in the degree of localizeddissolution of the oxide layer. (Figure 2b) shows the Tr alloys anodized at 0.3 g NH4F. Asseen the pit and pores which are seen for 0.1 g.^4F have been converted into sponge likestructure. The inset shows length of sponge.like structuresiofabout 163.8 nm. It could alsobe found that the entire metallic surface: was,-not uniformly covered with nanoporousstructure.

For 0.5 g and 0.7 g NH4F, selforganized TiOznanotubes-aitay is formed (Figure 2 cand d). The average diameter ofthe nanotubes is: approximately 81 and 110 nm for 0.5 g andfor 0.7 g, respectively. Generally the pores orthe nanotube diameter are found to be largerwith increasing fluoride content in the electrolyte.. This •is due to enhanced chemicaldissolution makes the barrier layer at the bottom relatively thin, which in turn lead to easiertransport of the fluoride ions in oxide layer. Such condition increases the electrical fieldintensity resulting in further pore or nanotube growth.

200nm WD= emm EHT=5.00kV Signal A= InLensMag - 30.00 KX[—_j Dal6:ajan2011 Time ;17:50:48

Figure 2(a): FESEM images ofTiOa nanotubes anodized in 99 ml ethylene glycol + 1ml HjOwith 0.1 gNH4F, at 60V for 60 minutes.

Mag = 30 00 KX EHTaS.OOkV Sign8l.A= InlensI I 0alo:8Jart20H TImo iT ^O O?

t'. ^.T-'V u' io-1

Figure 2(b): FESEM images ofTi02 nanotubes anodized in 99 m! ethylene glycol + 1ml H2Owith 0.3g of NH4F, at 60V for 60 minutes. Inset showsthe cross sectional view.

>

m

200nm WD= 6 mm EHT=5.00kV SignalA = InLensMag = 30.00 K X TkKC^

Date ;8 Jan 2011 Time :17;34;3S

Figure 2(c): FESEM images ofTi02 nanotubes anodized in 99 ml ethylene glycol + 1ml H2Owith-0.5 g of NH4F, at 60V for 60 minutes. >

200nm WD= 5 mm EHT=5.00kV Signal A = InLensMag = 30.00 KX Date :31 Dec 2010 Time :11:01:40 ZEISS

Figure 2(d): FESEM images ofTiOi nanotubes anodized in 99 ml ethylene glycol + 1 ml HjOwith 0.7 g of NH4F, at 60V for 60 minutes.

A 11' X a*TiA R-Tl; '- A •a-Ti (e) ;

• ''' ''V

• J| 1. A

28 (degree)

Figure 3; XRD pattern of the annealed Ti-6AI-4V alloy sheets anodized In EG at 60V for 60minutes with (a) as-anodized at 30 V (b) 0.1 g, (c) 0.3 g, (d) 0.5 g, and (e) 0.7 g weight ofNH4F. A = Anatase, a-Ti = alphaTi and P-Ti = BetaTi.

The corresponding XRD pattern of the Ti alloy sheet anodized in 0.1 g, 0.3 g, 0.5 g and 0.7 gNH4F is shown in Figure 3. After annealing at 400 °C, the diffraction peak appears at 20 of25° representing (101) plane of anatase Ti02. In addition, diffraction pattern of a-Ti at 20 =35.18°, 40.26°, 53.08° and 63.19° corresponding to the (100), (101), (102), (110) plane and p-Ti peak at38.75° corresponding at(002) plane were detected. These peaks are originated fromthe substrate of the Ti alloy. It is worth nothingthe peakrintensityof a and p Ti decreased asthe content of fluoride .increased.5This.Ts-jli.kely due? ta larger Sickness of the oxide layerformed on the Ti alloy.

Effect ofvoltage on anodization ofIT alloy in ethyleneglycol

Experiment was continued in sfmilarelectrolyte with 0.7 g NH4F to investigate theeffect of voltage on the Ti alloy. The voltajge was varied fronx 10 V, 20 V, 30 V, 40 V, and 60V for 1 hour. Different voltage in anodization process leads to different surface structure ofTi02. At low anodization voltage (lOV) in Figure 4a, small pits are observed on the surfaceof the oxide layer. At this stage the dissolution of titanium had just occurred. As the voltageincreased to 20 V in Figure 4b, the pits nucleate to form large porous structure. However, notall the surface is covered with porous Ti02. There are still areas covered with oxide layer.This is due to the insufficient voltage to induce complete dissolution and breakdovm ofbarrierlayers to dissolve the oxide into tubes.

When voltage increasedto 30V up to 60V, selforganizednanotube arrays are formed.Howeverthe lengths of the nanotubes in certain area are relatively low as compared to others(insets in Figure 4c). This finding is different compared to few authors who claimed theexistence of two kinds of the nanostructure on the surface of the Ti alloy. One was selforganized nanotubearrays and the other was irregularnanoporous structure.Accordingto fewauthors (Luo et al., 2008 and Matykinaa et al., 2011), the self organized nanotubes arrays areproven to be grown on the a phase region and nanoporous structure on P phase region. The aphase are enriched with Al whereas P phase region enriched with V. Because of the differentchemistries of these phases, the formation of the nanotubular oxide layer is not uniform as pphases get etched preferentially by the electrolyte. Different results were obtained in thiswork probably due to different electrolyte used. Most of the aforementioned studies utilizeaqueous solution which subjected highchemical dissolution rate. In here the use of ethyleneglycol with pH 6.5 could avoid the significant attack on P phase and produce homogeneousmicrostructure. However, the length, of the nanotubes are different within the a and p phaseregion, beingshort at p phase regionas the, splubility ofrV;is fasterthanthe Al.

The average diameter and the length ofT the. nanotubes formed with variousanodization voltages are shown in Figure 5. Generally itwas found that raising the appliedvoltage leadsto larger diameter and length. This is dueto the high applied electricfield whichleads to enhance polarization of Ti-O bond and thus weakened to promote dissolution of themetal oxide. This lead to greater driving force for ionic transport through the barrier layeratthe bottom of the nanotubes favouring formation of [TiFe]" complex that will eventuallyaccelerate the pore growth. At the. same time, the faster movement of the Ti/Ti02 interfaceinto the metal (Shankar et. al, 2007) resulted in the formation of long nanotubes. The standarddeviation of nanotube diameter at different voltages is almost constant, in the range of (5-lOnm) but for length showed significant difference due to high solubility of V which reducesthe length of the nanotubes in certain region. The side view of the nanotubes is relativelysmooth as seen in Figure 6. However, TEM analysis showed the presence of small ridges atthe wall. It is also found that the nanotubes are uniform with hollow nature of tube opening.The thickness ofthe wall was approximately 20 nm.

ijkr, < »»•

S|S> j

mmmm^ WmBmm200nfn WD= 5 mm EHT=5.00kV Signal A = InLens

Mag - 50.00 KX [— —| DatB:7Feb2011 111118:9:53:37

Figure 4(a): FESEM images ofTi02 nanotubes anodized in 99 ml ethylene glycol + 1ml H2O+ 0.7 g NH4Ffor 60 minutes at 10 V.

m

g*'\^

m71-r.:

.i . I K'

1

lOOnm WD= 5 mm EKT= S.OOkV SignaJA=]nL6ns

H Date :7 Feb 2011 Time :9:47:39Mag = 50.00 KX [_^

Compact

"tj"structure

m

Figure 4(b): FESEM images ofTiO? nanotubes anodized in 99 ml ethylene glycol + I ml H2O+ 0.7 gNH4F for 60 minutes at 20 V.

.00 nm

i. - w.%7"-'iff?/

w A ">*- * i*.' ' '• *

MMmm

Mag = 60.00 K X

If «' «,4^-i

200nm WD= 5mm EHT=5.00kV Signal A = lnLens

I ! Date :7Feb 2011 Time :9:40:55

Figure 4(c): FESEM images ofTI02 nanotubes anodized in 99 ml ethylene glycol + 1ml HjO+ 0.7 gNH4F for 60 minutes at 30 V. Inset showsthe cross sectional view.

Af* »'l' \ *

f

rj / '

-.K V, 200nm WD= 6 mm EHT=5.00kV Signal A= InLensMag - 50.00 KX |—_—| Date;8Jan2011 Time:17;44:02

Figure 4(d): FESEM images ofTiOi nanotubes anodized in 99 ml ethylene glycol + 1ml H2O+ 0.7 gNH4F for 60 minutes at 40 y. 1

1um WD= 4rrBTr EHT=5.0QW SignalA = InLensMag= 15.00 KX I 1 Date:18Feb201l Time:16:21:47

Figure 6: FESEM image ofside view with T1O2 nanotubes

c:\edax32\genesis\genmap$.spc 07-Feb-201110:53:22LSecs: 101

478-,

Figure 8: EDX spectra of the oxide residua! found on the Ti alloysurfaceafter anodized in 99ml ethylene giyco! + I ml H2O + 0.7 g NH4F, at 40 V

The chemical composition of the Ti02 nanotubes prepared at various voltages wasdetermined using EDX analysis. The representative spectrum of the Ti alloy anodized atvarious voltages is shown in Figure 8. The result clearly shows the presence of C, O, Al, Tiand V. EDX spectrum of the other sample are almost similar (not shown), only slightvariation on the At % of all the elements. However it is interesting to note that thecomposition of the vanadium slightly reduced from 2,3% to 1.76 % with increase voltagefrom 10 to 60 V. This is probably due to-preferenti^-etching of V at high voltage ascompared to other phases.

Table 4.1: EDX spectrum atvarious:Woltagesi'

Element At % at different voltai

10 V 40 V i 60 V

C 03.42 05.13 04.58

0 47.10 39.86 1 43.68

Al 04.68 05.50 04.19

Ti 39.53 44.20 42.77

V 02.30 02.16 01.76

X-ray Photoelectron Spectroscopy (XPS)

The chemical composition present at the surface of Ti-6A1-4V nanotube subjected toannealing at 400 °C for 4 hours under Argon atmosphere was investigate using XPS analysis,and the result is shown in Figure 9. Based on the result, the binding energies of Ti2p3/2 andTi2pi/2 were found at 456.5 eV and 463.3 eV, respectively (Figure 9a). This is likely due tothe presence of Ti"^ oxidation state. The Ti''̂ ions reacted with O^* ions during anodization,and consequently formed Ti02. This was confirmed by the presence of Ti-0 (527.5-527.7eV) in 01s (Figure 9b). Besides, the presence of OH' ions and absorbed H2O was detected at528.9 eV and 531.9 eV. Other than the information of Ti02, binding energy of A12p wastraced at 72.1 eV, 72.2 eV and 73.9 eV (Figure 9c). This is attributed to the binding energy ofAl-0, which in agreement with that report by Li et al, (2008). In addition, the V2p3/2 peakcorresponding to the metallic V appears at 515.5 eV and 518.2 eV(Figure 9d). This is duetothe fact that the affinity of Al and Ti with 0 is stronger than that of V. Al and Ti diffusedoutward, bringing on implanted layer outward growing, therefore V is deficient relatively inthe outmost layer. However, the existing of oxidized state of Al, and the leaching of Vcontribute to the reduce cellular toxicily of Al and V:of Ti-6A1-4V as biomaterial (Li et al.,2008). • • ' Vi

0

CQ

CO

d

(a) (Ti2p3;2)A456.5 eV W Ti2p

1 41 4i 41 4I A

(Ti2pi/,)463.3 e¥

AF *

m •

1

445 450 455 460 465 470Binding energy (eV)

Figure 9(a): XPS spectraof Ti2p region

475

27.7 eV

527.5 eV

524 526 528 530 . 532 534Binding energy (eV)

Figure 9 (b): XPS spectraof01s region

536

67

S

A

d0)

(d)

505

72.2 eV

69 71 73 75

Binding energy (eV)

Figure 9(c): XPS spectra of A12p region

518.2 eV

515.5 eV

510 515 520

Binding energy (eV)

Figure 9(d): XPS spectra ofV2p.region

77

V2p

525

Effect ofanodization time on anodlzed Ti alloys in ethylene glycol

This set of experiments was done to monitor the growth of nanotubes with increasinganodization time. The experiment was carried out in similar electrolyte with 0.7 g F at 60V.The growth of the nanotubes was monitored by taking FESEM images of various timeintervals (20 seconds, 1 minute, 5 minutes, 10 minutes, 15 minutes, 30 minutes, 1 hour and 6hours). Table 4.2 shows the diameter and length of nanotubes!Ti-6Al-4V alloy formed in thevariation of anodization time. ^

i ' •j

Table 4.2: Diameter and Length ofTi02 nanotubes in diffefentanodization time

Time (minutes]20 sec

Tube Diameter (nm)' Tube Lent MorphologyCompact structurePits formation

Porous structure

nanotubesnanotubesnanotubesnanotubes

1; -

5 64

i:o 68

70

30 100

1 hour 110

6 hours 159

0.4 • •

XA1.7

^,83.1

nanotubes

Figure 10a shows the image of the surface of 20 seconds. An oxide layers are formedon the surface of the ofTi alloy due to the interaction within O^" and OH". After 1minute theoxide layer has been etched in certain area, thus forming irregular pits due to localizeddissolution (Figure 10b). After 5 minutes of anodization, the pits are converted to largerpores(Figure 10c). However mostof the areas arestill covered with oxide layer. At 10minutes, thepore density is increased and the cross sectional view (Figure lOd) confirms the formation ofnanotube structure. After 15 minutes, the surface is completely filled with self organizednanotubes arrays (Figure lOe). Prolonging the anodization time to 30 minutes (Figure lOf), 1hour (Figure 2d) and 6 hours (Figure lOg) resulted in larger diameter and tube length. Theoverall formation of the nanotubes on Ti alloy (Ti-6A1-4V) is similar to the Ti-8Mn reportedby Mohapatra et al.

Mags 50.00KX

• - ...... ..

V Y-. ». • ... .1. .». • . t. A— ........ . - .

WQs {mm SHT: &<|l>kV

I Doto'ldUar^lQIISgn^A^inlenf

Tin-^

Figure 10a: Top view of FESEM images ofTi02 nanotubes anodlzed at 20seconds.I

IVlag= 50.00 KX200nm WD= 5 mm EHT=5.00IW Signal A = InLens

I I Date;14Mar2011 Tlme:8:59:34

Figure 10b: Top view of FESEM images ofTiOz nanotubes anodized at 1 minute. The insetshows cross sectional view of the sample

200nm WD= 6mm EHT=5.00kV SignalA= InLensIVlag= 50.00 KX | [ Date:4Mar2011 Time :16;59:06

1

Figure 10c: Top view of FESEM images of TiOz nanotubes anodized at 5 minutes.

%

f

I If-

M

M

Map fin nn Ky WD= Smm EHT= 5.00kV Signal A=InLens^ I Date:4Mar2011 Time :16;52:55

'V

I

Figure lOd: Topview of FESEM images ofTiOi nanotubes anodized at 10 minutes.

Wlag = 50.00 K X 2Q0nm WD= 5 mm •EHT=5.00kV SignalA = InLens

Date :4 Mar 2011 Time :16:56:16

Figute lOe: Top view of FESEM images ofTiOz nanotubes anodized at 15 minutes, The insetshows cross sectional view.

i." "ft.V' K - . 4.

.» 't'.'i'* t.7 ^ •'

;.:KP- - (ffrlVrl l'.. ^

•jf'fli ;! '̂ •I'bll J.;T^ ^

Mag = 50.00 K X200nni WD = 5 mm EHT=5.00kV Signal A = lnLens

Date ;4 Mar 2011 Time :16:43:02

Figure lOf: Top view of FESEM images of Ti02 nanotubes anodized at 30 minutes The insetshows cross sectional view.

Mag = 50,00 K X2KI'ifri 4trmT EHT= G.CO kV StgnoiA= jptena

Date 20 Feb 2011 Tni#'22-3ri0

•r^.nk

Figure lOg: Top view of FESEM Images of Ti02 nanotubes anodized at 6 hours. The insetshows cross sectional view of the respective samples

0.03

0.02

I 0.01

a

0

0 60 120 180

Time (inin)

Figure 11: Current density profileofTi alloy

Figure 11 shows the current density profile recorded during anodization. Three stageof current density were observed in ethylene glycol. Stage 1, represent formation of oxidelayer, stage 2 signified nucleation of nanoporous and stage 3 denoted formations ofnanotubes. Growth of nanotubes structure occurred at stage 3 where by the length of thenanotubes increased from 2.8 to 6.4 pm.

Effect ofannealing temperature

Figure 12 shows the XRD paittem of the as anodized and annealedTiOa nanotubes atdifferent temperatures. The result clearly shows the crystal structure of TiOi depends on theannealing treatment. The XRD of the as-anodized saniple indicate amorphous structure asonly a and p-Ti peaks were pb^rvecL.Annealing.at'400;i?G haspromoted the crystallization ofanatase phase at 20 of 25° and 48°j corresponding 01) and>'(200) plane, respectively.Dominant structure of the sample annealed at-^600 °C M'rutile, implies completetransformation from anatase to rutile at this temperatures

The morphology of the as anodized Ti02 without heat treatment and annealed at500°C, 550°C as well as 600°Gare shown in Figure B. a,b,c and d, respectively. The sampleheat treated at 400°C is discussed previously in section 4.1. Overall, the results show therobustness of Ti02 nanotubes formed on Ti alloy is teiriperature dependent. The nanotubesstructure is sustained till 400°C. At 500°C and 550°C, thickening of titania nanotubes walloccurred due to mass transport involving Ti''̂ diffusion at the bottom and wall of the Ti02nanotubes. Finally, at 600°C, tubular structure is completely damaged and nanorod likestructure is formed. The schematic diagram explaining the morphologies changes withtemperature are illustrated in Figure 14.

Nanotubular

structure

initia

barrier

layer

Growth of wall thickness

and barrier layer

Dense nanorod

like strcuture

Figure 14: Schematic illustration to show the growth of the nanotubes during thermalannealing

Surface Roughness

AFM topographic maps of as anodized andannealed Ti02 are shown in Figure 15 a-c.The surface roughness parameter Ra, Rq and Z range of each sample were examined andsummarized in Table 3. It can be seen that higher annealing temperature induces higherroughness level as a consequence of an increased in densification of the nanotubes.

Table 3: Theroughness parameter obtained upon annealing temperature variation

"" :—-.-Annealing temperature (®C) I ^ •'

Roughness Parameters (nm)f~~~^^ ^ •"

Ra (average)RMSZ range (max.-min.)

48.65^

60.49

292.30

. 67.25

84.56

521.10

95.97

122.60

646.90

Stxrrr :

Figure 15: AFM topography and 3D morphology ofTi02 nanotube (a) 0°C (b) 400°C and (c)600°C

Cell interaction

In this part, the behavior of cellular interaction on three different types of anatase,amorphous and rutlle TiOz using fibroblast cells line was done. MTS assay was performed tostudy the cells viability. The cellular activities of cell viability for fibroblast cells on threedifferent types of crystal structures are .shown in Figure 16. Cell proliferation is theattachment of the cells on the material surface. The results shows no significant valueobtained for 24 and 48 hours of cell culture for ai! three crystal structures. At 72 hours,amorphous structure has significant value; this reveals increase in proliferation HS27 cells onas-anodized TiOi nanotubes surface TRi^^sftow HS27 fibroblast cells has better cellularinteraction on amorphous Ti02 nanotubes as compared to anatase or rutile. It also indicatesthat fibroblast cells like surface with lower surface rou^ness of about 49 to 60 nm.

Amorphous Anatase

•24h

B48h

a72h

Rutile

Figure 16: MTS assay data showing the optical density (OD) of reaction product of the MTSworking solution with HS27 fibrobiast cells cultured as-anodized, annealed at 400°C and 600°C amorphous Ti02 nanotubes, and heat-treated (anatase) TiO, nanotubes after 24 h, 48 h, and72 h of incubation. The error bars in the figure represent the standard deviation forVoursamples for each data.

Conclusion

^The TiOs nanotubes were fabricated on Ti-6AI-4V alloy by anodization process.Optimized condition for formation of TiOs on Ti alloy are: electrolyte with 99 ml ethyleneglycol +1ml H2O +0.7 gNH4F, voltage 60 Vand time ofanodization is 60 minutes. Ti02nanotubes with average diameter of110 nm and 3.1 pm length are obtained. Cell interactionstudy on Ti alloy with various crystalline structure om fibrobiast cells line showed thatamorphous structure possess good proliferation, followeAby,anatase and rutile phase.

References •

Bose, Bandyopadhyay, (2009) Enhanced "Fatigue"'Ferformance of Porous CoatedTi6A14V Biomedical Alloy, 225. ; -Li, Xiao, Liu. (2008) Synthesis And Bioactivity Of Highty Ordered Ti02 Nanotube ArraysApplied SurfaceScience, 255. 365-367Luo,Yang, Liu, Fu, Sun, Yuan, Zhang ,Liu. (2008).; Fabrication And Characterization OfSelf-Organized Mixed Oxide Nanotube Arrays By Electrochemical Anodization OfTi-6A1-4V AlloyMaterials Letter,s 62.4512-4515Malykinaa, &J. M. (2011). Morphologies Of Nanostructured Tio2 Doped With F On Ti-6A1-4V Alloy, Electrochimica Acta, 56, 2221-2229. • :Shankar, Mor, Prakasam, Yoriya, Paulose, Varghese, Grimes. (2007) Highly-ordered Ti02nanotube arrays up to 220 pm in length: use in water photoelectrolysis and dye-sensitizedsolarcells Nanotechnology 18, 1-11

Sreekantan, Saharudin, Lockman, & Tzu. (2010). Fast-Rate Formation Of Ti02 NanotubeArrays In An Organic Bath And Their Applications In Photocatalysis. Nanotechnology, 21Yu, Jiang, Zhang & Xu, L. (2010). The Effect Of Anatase Ti02 Nanotube Layers OnMC3T3-B1 Preosteoblast Adhesion, Proliferation, And Differentiation. 1012-1021.

4>

• •• ^'}'0'O.''''yr-f Vv- -Va

:. ^ 'y <\ M-^VV

. v .'j-w-^r:'w-o-*^viii v 'A..f \...v: '^Vr:"I,.';•••-•.- ••» , .S),'"f,5*V'i.k't'

>'•-'iv '••'Tr- V >•!»• .tiv .i .̂t.r .\ -l-!•>>:• •""Y V

^^®^|gp^llll^i^p|l|l|l^'

A.

B.

C.

1.

2.

3.

4.

5.

6.

7.

8.

TITLE OF RESEARCH:Tajukpenyelidikan:

Formation ofNanotuhular Ti02 Bioactive OxideLayer onTitanium and Cell Response Studies

PERSONAL PARTICULARS OF RESEARCHER / MaJdumat Penyelidik:

(i)Name of Research Leader:

Nama Ketua Penyelidik:Dr. Sreemala Sreekantan

(ii)School/Institute/Centre/Unit:

Pusat Pengajian /Institut/Pusat/Unit:Pusat Pengajian Kejuruteraan Bahan &SumberMineral

(Hi)

(iv)

(V)

Phone No. (Office):No. Tel (Pejabat):

Hand Phone No:

No. TelefonBimbit:

Fax. No:

No. Faks:

SUMMARY OF ASSESSMENT

(Tick (ylj the appropriate box. Also, provide additionalcomments in Section F)

Achievement of Project Objectives

Quality ofOutput

Quality of Organisational outcomes

Quality of sectoraPnational impacts

TechnologyTransfer / Commercialization Potential

Quality and Intensity of Collaboration

Overall Financial Expenditure

Overall A^essment of Beneflts

Inadequate Acceptable Very Good

^ 1

H

ZTI

Z1

3

H

Office ofResearch Platform - Universiti Sains Malaysia - February 2011

D OUTPUT SUMMARY

1 Achievement Percentage

Project progress according tomilestones achieved up to this period

25% 50% 75% 100%

2 Research Findings

ISl JournalInternational

JournalNational Journal

Number of articles/ manuscripts/books \

International National

Paper presentations

\

Others

(Please specify)

3 Human Capital Development

Number

On-going Graduated

M'sian Non- M'sian M'sian Non- M'sian

Ph D Student

M Sc Student

Undergraduate FinalYearProject z

Temporary Research Officer

Temporary Research Assistant

Others (Please specify)

Total

Office of Research Platform - Universiti Sains Malaysia - February 2011

E.

F.

G*

H.

I.

ACTION TO BE TAKEN (Tick (^) the appropriate boxes)

]/Project file to beclosed. (Forfinancial andadministrationpurposes)

• Project to be Reviewed Once Additionni InforniBtion Has Been Obtained by the Project Leader(Please see section E)

Forward to Innovation Office for Consideration:

CH Patent ED CommercialisationI IOthers (Please specify):

Forward to Division of Industry & Community Network(DICN)

Others (Pleasespecify):(E.g.: Formations ofteams or cluster etc.)

Technology Transfer

Additional information to beprovided bytheProject Leader [Please provide details ofthe specific informationbeing requested from the project Leader inthe areas identified in Section E]

Comments Regarding Assessment (Please provide below anexplanation of any assessment made inSection Cshowing a ratingbelow"acceptable"]

Overall Comments

Name of Panel: Signature

Date

Endorsement & Comments/Suggestions of Research Dean:

Signature: Date

Officeof Research Platform - Universiti SainsMalaysia - February 2011