Mechanical stability of two-step chemically deposited hydroxyapatite

coating on Ti substrate: Effects of various surface pretreatments

Alexander V. Zavgorodniy,1 Oscar Borrero-López,2,3 Mark Hoffman,3

Racquel Z. LeGeros,4 Ramin Rohanizadeh1
1
Faculty of Pharmacy, University of Sydney, NSW 2006, Australia
2
Departamento de Ingenierı́a Mecánica, Energética y de los Materiales, Escuela de Ingenierı́as Industriales,
]Universidad de Extremadura, Avenida de Elvas, s/n, Badajoz 06071, Spain
3
Department of Materials Science and Engineering, University of New South Wales, NSW 2052, Australia
4
College of Dentistry, New York University, New York, New York 10100, U.S.A.

Received 10 January 2011; revised 4 April 2011; accepted 5 April 2011

Published online 28 June 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.b.31872

Abstract: The success of implants in orthopaedic and dental coating on Ti substrate in acidic condition at 75
C and the
load-bearing applications crucially depends on the initial bio- second step was the hydrolysis of the monetite coating to
logical fixation of implants in surrounding bone tissues. HA. Coatings were characterized by scanning electron micro-
Using hydroxyapatite (HA) coating on Ti implant as carrier scope (SEM), energy-dispersive X-ray spectroscopy (EDS),
for bone morphogenetic proteins (BMPs) may promote the and X-ray diffraction (XRD). The roughness of substrates and
osteointegration of implants; therefore, reduce the risk of coatings was measured using profilometry technique. The
implant failure. The goal of this study was to develop an HA mechanical stability of the coatings deposited on the pre-
coating method in conditions allowing the incorporation of treated substrates was assessed using scratch test. The coat-
protein-based drugs into the coating materials, while achiev- ings deposited on the grit-blasted Ti surface demonstrated
ing a mechanical stable coating on Ti implant. HA coatings superior adhesive properties with critical shearing stress
were deposited on six different groups of Ti surfaces: control 131.6 6 0.2 MPa. VC 2011 Wiley Periodicals, Inc. J Biomed Mater Res

(no pretreatment), pretreated with alkali, acid, heat at 800
C, Part B: Appl Biomater 99B: 58–69, 2011.
grit blasted with Al2O3, and grit blasted followed by heat
treatment. HA coating was prepared using a two-step proce- Key Words: hydroxyapatite coatings, titanium implants,
dure. First step was the chemical deposition of a monetite scratch tests

How to cite this article: Zavgorodniy AV, Borrero-López O, Hoffman M, LeGeros RZ, Rohanizadeh R. 2011. Mechanical stability of
two-step chemically deposited hydroxyapatite coating on Ti substrate: Effects of various surface pretreatments. J Biomed Mater
Res Part B 2011:99B:58–69.

INTRODUCTION include biomimetic processes,3–5 sol–gel,6,7 ion-beam-

Titanium and titanium alloys orthopaedic and dental assisted deposition,8 electrochemical deposition,9,10 chemical
implants are widely used in load-bearing applications due deposition,11,12 and other deposition methods. Among them,
to the high strength and a relatively low elastic modulus of plasma-sprayed HA coating has received commercial success
Ti (100 GPa), which is a closer match to the elastic modulus and has been widely used in clinical situations.13 However,
of bone (around 30 GPa), than conventional implant materi- still 17.5% of hip implants and 8.2% of knee implants14
als, such as 316-L stainless steel (210 GPa) or cobalt–chro- coated with HA plasma-sprayed fail and require revision
mium alloys (240 GPa).1 Titanium is a biocompatible but operation.15 In plasma spraying technique, HA powder is
bioinert material. Coating of Ti orthopaedic and dental heated to extreme temperatures, 6000–10,000
C, and par-
implants with bioactive materials such as calcium phos- tially molten particles are deposited on a metal substrate.
phates will improve implant’s biological performances and The high temperature causes partial dehydration and
integration into the surrounding bony tissues. Hydroxyapa- decomposition of HA and, in turn, results in the formation
tite (HA) is an osteoconductive material with a chemical of variation of calcium phosphate phases, such as calcium
structure similar to mineral phase of bone and tooth. HA oxide, a- and b-tricalcium phosphates, tetracalcium phos-
coating on titanium implants has been demonstrated to phate, and amorphous calcium phosphate. An HA plasma-
enhance peri-implant bone tissue formation and improved sprayed particle is composed of a solid core and a liquid
implant osseointegration.2 The advances in HA coating dep- shell. The rapid cooling of crystal phases in the solid core at
osition techniques have been described in the literature and impact with the substrate results in conservation of high

Correspondence to: A. V. Zavgorodniy; e-mail: [email protected]

Contract grant sponsor: Australian Research Council Discovery; contract grant number: DP0986230

58 V
C 2011 WILEY PERIODICALS, INC.
 ORIGINAL RESEARCH REPORT

temperature phases, whereas the liquid shell is transformed

into an amorphous phase. It is difficult, using plasma spray-
ing coating method, to control phase composition variation,
morphology, uniformity, and mechanical stability of the coat-
ings. The uneven degradation of the plasma-sprayed HA
coating in body is likely to result in nonhomogenous bone
bonding or bone growth around the implant and/or delami-
nation and separation of coating fragments, which can cause
the disintegration of the coating materials and severe com-
plications in implant integration in the body.16 Therefore,
the initial biological fixation of the implants and the me-
chanical stability of the HA coating of the implant are cru-
cial for the long-term life span and success of orthopaedic
and dental implants.
One of the strategies to enhance bone healing around FIGURE 1. Ti disk is positioned vertically in the electrolyte coating
implant, and therefore implant fixation and osseointegration, solution. [Color figure can be viewed in the online issue, which is
is to use the HA coating on the implant as a carrier for in available at wileyonlinelibrary.com.]
situ delivering of drugs, such as bone morphogenetic pro-
teins (BMP).17 The incorporation of the drug during the allow to incorporate protein-based drugs into the HA coat-
process of the coating formation will ensure a sustained ing. The study determined the effects of substrates pretreat-
release of the drug into the surrounding bony tissue after ments, i.e., machine ground, grit blasted, heat treated, grit
the joint replacement surgery. Plasma spraying method can- blasted followed by heat treatment, acid- and alkali-treated
not be used for incorporation of drugs into the coating Ti substrates on coating physico–chemical and mechanical
materials due to the extreme temperatures of this tech- characteristics.
nique. Chemical deposition of HA coating in mild conditions,
coating temperature below 100
C, may allow the incorpora-
MATERIALS AND METHODS
tion of heat-sensitive protein-based drugs into the HA coat-
Pretreatments
ing. The interfacial integrity of the HA coating and Ti sub-
The surface of commercially pure Ti (cp-Ti) disks (15 mm
strate is essential during the implant life span and the
in diameter, 1 mm thickness) were machine ground to 800
requirements for mechanical properties of coatings in ortho-
grit with a silicon carbide (SiC) paper and then cleaned in a
paedic load-bearing applications have been documented and
sequence of ultrasonic baths: ultrapure water, acetone, ultra-
specified in the Food and Drug Administration guidelines18
pure water, ethanol, and finally, ultrapure water, 10 min in
as well as in the ISO standards.19
each bath. The disks, which received no further treatment,
The adhesive strength of the coatings is the combination
were used as a control group. For the acid and alkali treat-
of chemical bonding and mechanical interlocking between the
ments, Ti disks were soaked in, respectively, 2M H3PO4 and
coating and the substrate. It has been reported that apatite
2M NaOH for 24 h at room temperature and then gently
has affinity to rutile (one of the TiO2 crystalline phases) due
washed in ultrapure water, and then air dried for 24 h. Heat
to the close match between the O–O distance in the (110)
treatment was carried out in a furnace at 800
C for 1 h.
plane of rutile and the Ca–Ca distance in the (100) plane of
Grit blasting was performed with 120/þ140 mesh Al2O3
the carbonated apatite, which resulted in a highly adhesive
abrasive with the particle size range 125 lm to 106 lm.
apatite coatings on rutile.20,21 Another chemical modification
The blast distance was 15 cm, air pressure 30 psi at room
of Ti surface is acid treatment. It has been reported that acid
temperature for 10–15 s. The blast was set to be perpendic-
treatment of Ti surface increases the positive charge of the Ti
ular to the specimen surface. After the grit-blasting proce-
oxide layer on Ti substrate and induces the formation of HA
dure, the specimens were cleaned ultrasonically in 20%
coating on Ti surface in simulated body fluid (SBF).22 Fur-
HNO3, ultrapure water, ethanol, and ultrapure water for 20
thermore, acid treatment may form small pits on the surface
min in each bath and then air dried on the clean bench for
of Ti substrate, which leads to a greater mechanical interlock
24 h at room temperature. Some grit-blasted specimens
between the coating materials and the substrate. Alkali treat-
were then subjected to heat treatment in a furnace at 800
C
ment of Ti surface has also been reported to increase the
for 1 h.
bonding strength between HA and Ti by formation of a Ti ox-
ide layer on substrate.23 Mechanical modification of Ti sub-
strate using grit blasting, results in mechanical interlocking HA coating and characterization
between the coating and the surface of the substrate. It has The prepared Ti disks were vertically placed in a solution
been reported Al2O3 grit blasting of Ti alloy substrate (Figure 1), 0.23M NaH2PO4H2O, 0.23M CaCO3, 0.55M H3PO4
resulted in a higher bonding strength of HA coatings com- at pH 2.70 and 75
C. After 24 h, the specimens were
pared with what obtained using SiO2 grit.24 removed from the coating solution and gently washed in
This study was designed to synthesize and characterize ultrapure water. The specimens were then hydrothermally
chemically deposited HA coating in conditions, which may treated in alkali solution, 0.2M NaOH, pH 12.82, at 75
C for

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | OCT 2011 VOL 99B, ISSUE 1 59
FIGURE 2. A: EDS of 800 grit ground Ti substrate showing the presence of small amount of Si on the surface of substrates; B: XRD of the 800
grit ground Ti substrate, indexed according to JCPDS 44-1294; C: EDS of the Al2O3 grit-blasted Ti substrate showing the presence of Al on the
surface of Ti substrates; D: XRD of the grit-blasted Ti substrate showing the presence of Al2O3 on the surface of Ti substrates.

15 h, washed in ultrapure water, and air dried at room tem- with the total length of 16 mm. The sliding speed of the in-
perature for 24 h. denter was 10 mm/min and the loading rate was 10N/min.
The average surface roughness (Ra) of the pretreated The mechanical events occurred along the scratch tracks
substrates and HA coating was measured according to the were evaluated under SEM (XL30, Philips) to locate the
ISO 428725 using profilometer (Surftest SV-600, Mitutoyo, position of the coating failure. The corresponding normal
Japan), equipped with a spherically tipped conical stylus of and tangential forces were then identified from the curves
radius 5 lm. The profilometer was calibrated using a stand- measured by the scratch tester. The critical shearing stress
ard reference specimen. The crystal structure of the sub- for coating removal was calculated according to Benjamin
strates and coatings was determined using X-ray diffraction and Weaver theory,27–29 which pertains to the soft coatings
(XRD, S6000, Schimadzu, Japan) with a step size of 0.02
at on a harder substrate case:
a scanning rate of 1
min1. Surface morphology and the
chemical composition of the pretreated Ti substrates and kAH
sc ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (1)
coatings were analyzed using scanning electron microscope R2  A2
(SEM, Ultra plus, Zeiss, Germany) coupled with energy-dis- qffiffiffiffiffi
Lc
persive spectrometer (EDS, Bruker, Germany). Scratch tests where A ¼ pH is the radius of contact, Lc is the critical
of the HA coatings deposited on pretreated Ti substrates load, H ¼ 4.04 GPa is hardness of the substrate material,
were performed according to the test protocol outlined in which was measured using an ultra-microindentation
ASTM C1624-05.26 A Revetest scratch tester (CSM Instru- system equipped with a Berkovich indenter (UMIS-2000,
ments, Switzerland) equipped with a Rockwell C diamond Fischer-Cripps Labs, Australia) according to the Oliver and
indenter, 200 lm radius, was used to perform all the Pharr method.30 R is the radius of the diamond tip, and k ¼
scratch tests. Based on the coating adhesion in different 0.2.28
groups, normal loads applied to the indenter were raised
from 1 to 6N over a distance of 5 mm, from 4 to 11N over RESULTS
a distance of 7 mm, and from 10 to 17N over a distance of Chemical analysis of the cp-Ti substrates prepared by
7 mm. The dimensions of the specimens precluded the real- machine grinding with SiC paper showed that even after
ization of a single scratch covering the whole load range cleaning procedures, traces of Si remained present on the

60 ZAVGORODNIY ET AL. MECHANICAL STABILITY OF HYDROXYAPATITE COATING ON TI SUBSTRATE

 ORIGINAL RESEARCH REPORT

FIGURE 3. A and B: Ti substrate ground to 800 grit; C and D: grit-blasted Ti substrate, arrows indicate Al2O3 particles embedded into the Ti sur-
face; E: back-scattered electrons image of the grit-blasted Ti substrate, darker areas on the image correspond to Al2O3 particles embedded into
the Ti surface.

surface of the control specimens [Figure 2(A)]. Structural [Figure 3(A)], were observed. The average surface rough-
analysis, which was performed using XRD, showed only Ti ness (Ra) of this control group was 0.426 6 0.049 lm.
peaks with no traces of crystalline Si compounds [Figure SEM observations of the grit-blasted surface revealed
2(B)]. Al2O3 particles embedded into the Ti substrate [Figure
EDS of the Al2O3 grit-blasted Ti substrates showed the 3(C,D), arrows]. Back-scattered electron (BSE) imaging
presence of small amount of Al [Figure 2(C)]. The analysis revealed that despite of the thorough cleaning protocol,
of the XRD profiles of the grit-blasted Ti substrates con- which is a protocol accepted as a standard method for the
firmed the presence of Al in the form of crystalline Al2O3 Ti orthopaedic prosthesis, some Al2O3 particles were still
particles [Figure 2(D)], which were embedded into the sur- embedded into Ti surface [Figure 3(E)]. The average surface
face of Ti during the grit-blasting procedure and were not roughness of this group was 1.121 6 0.092 lm.
completely removed by ultrasonically cleaning in 20% The XRD analysis of the heat-treated Ti revealed the for-
HNO3. mation of rutile on the surface of Ti [Figure 4(A)] in the
The morphology of the machine-ground Ti substrates form of a dense coating of rutile crystals [Figure 4(B,C)]
showed that the roughness of the surface was mainly due to and the average surface roughness was 0.595 6 0.049 lm.
the scratches left on the surface by SiC paper [Figure A dense rutile coating also formed on the surface of the
3(A,B)], no SiC particles, which were detected using EDS grit-blasted and then heat-treated Ti specimens [Figure

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | OCT 2011 VOL 99B, ISSUE 1 61
FIGURE 4. A: XRD of the Ti substrate treated at 800
C for 1 h. R—rutile, indexed according to JCPDS 21-1276. (B and C) SEM images of the mor-
phology of the Ti substrate treated at 800
C for 1 h.

FIGURE 5. A–C: SEM images of the grit blasted and then heat-treated Ti substrate. D: BSE SEM image demonstrating the morphology of the grit
blasted and subsequently heat-treated Ti substrate. Arrows show Al2O3 particles embedded into the surface of the Ti substrate.

62 ZAVGORODNIY ET AL. MECHANICAL STABILITY OF HYDROXYAPATITE COATING ON TI SUBSTRATE

 ORIGINAL RESEARCH REPORT

FIGURE 6. BSE (A) and SE (B) SEM images of the acid-treated Ti substrate for 24 h. BSE (C) and SE (D) SEM images of the alkali-treated Ti sub-
strate for 24 h.

5(A–D)]. Figure 5(B,C) (arrows) as well as the BSE image of of Ti was partly dissolved and pits were formed on the sur-
the grit-blasted heat-treated Ti substrate [Figure 5(D), face [Figure 6(A) and 5(B)]. The average surface roughness
arrow] revealed that rutile did not cover Al2O3 particles, of the acid pretreated Ti substrates was 0.453 6 0.061 lm.
embedded into the surface of Ti after the grit-blasted proce- BSE images of the Ti substrate treated in sodium hydroxide
dure. The average surface roughness of this group was solution show contrast variation, which could be attributed
1.421 6 0.034 lm. to the inclusion of the Naþ ions into titanate hydrogel layer
The morphological observations of the acid pretreated [Figure 6(C)]. Higher magnification image of the alkali pre-
Ti substrates revealed that under the acid attack the surface treated Ti surface [Figure 6(D)] showed honey-comb

FIGURE 7. Typical EDS (A) and XRD (B) profile (indexed according to JCPDS 09-0432) of the HA coatings deposited on control and pretreated Ti
substrates. The EDS and XRD profile shown obtained from the HA coating deposited on a control Ti substrate.

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | OCT 2011 VOL 99B, ISSUE 1 63
FIGURE 8. Typical microstructural morphology of HA coatings deposited on the control and pretreated Ti substrates. The geometry of the apa-
tite agglomerates conformed to the profile of the precursor monetite crystals. Images shown were obtained from the control group.

morphology of the titanium substrate and the average sur- 9(C), Ra ¼ 6.417 6 0.351 lm]. The morphological homoge-
face roughness was 0.434 6 0.052. neity of the coatings deposited on the alkali pretreated sub-
EDS analysis of the coatings deposited on control and strates was poor and the deposited HA particles did not
pretreated Ti substrates showed that the major chemical cover the entire metal surface of the substrate, with the
constituents of the coatings were calcium, phosphorus, and presence of voids between the coating materials [Figure
oxygen [Figure 7(A)], and the Ca/P ratio of the coating ma- 9(D), arrows]. These observations correspond with the high-
terial was 1.15. Traces of sodium, silica, and titanium est values of the average surface roughness of the HA coat-
were also detected. ings deposited on alkali pretreated Ti substrates, Ra ¼
The structural analysis using XRD confirmed that the 12.603 6 1.839 lm, compared with Ra values of the other
calcium phosphate deposited on Ti substrate in all groups groups.
was hydroxyapatite [Figure 7(B)]. Higher magnification mor- HA coatings deposited on heat-treated substrates (rutile)
phological observations of the coatings deposited on the showed numerous cracks through the thickness of the coat-
control and pretreated Ti substrates showed that the coat- ings [Figure 9(E), arrows], Ra ¼ 9.085 6 0.391 lm, whereas
ings consisted of agglomerates [Figure 8(A,B)] of bundles of coatings deposited on grit blasted followed by heat-treat-
needle-like apatite crystals, which formed complex three- ment substrates showed a dense and homogenous coatings
dimensional interwoven pattern within the agglomerates without visible cracks, Ra ¼ 5.629 6 0.546 lm.
[Figure 8(C,D)]. The geometry of the agglomerates con- The adhesive strength of the HA coating to the pre-
formed to the plate-like profile of the precursor monetite treated Ti substrates was investigated by scratch tests. Each
crystals. scratch was examined under the SEM using the BSE detec-
Lower magnification BSE SEM images of the HA coatings tor to reveal mechanical failures of the coatings, such as
deposited on the control and pretreated Ti substrates areas of coating removals and cracks (Figure 10).
revealed variation of the coating morphologies (Figure 9). Coatings deposited on control and all pretreated sub-
HA particles deposited on the control Ti substrate [Figure strates demonstrated ductile failure modes in scratch tests,
9(A), Ra ¼ 4.472 6 0.276 lm] appeared to be less packed consistent with failure modes typical to soft coatings on
than those deposited on the grit-blasted Ti substrates [Fig- harder substrates.29 The apatite coatings deposited on the
ure 9(B), Ra ¼ 5.547 6 0.508 lm] but comparable with the grit-blasted Ti substrates exhibited the first catastrophic
coatings deposited onto acid pretreated Ti substrate [Figure failure [Figure 10(A), arrow 1A], i.e., localized removal of

64 ZAVGORODNIY ET AL. MECHANICAL STABILITY OF HYDROXYAPATITE COATING ON TI SUBSTRATE

 ORIGINAL RESEARCH REPORT

FIGURE 9. HA coatings morphologies deposited on control and pretreated Ti substrates. A: Control group; B: grit-blasted Ti substrates; C: acid-
treated Ti substrates; D: alkali-treated Ti substrates, notice voids in the HA coating (arrows); E: heat-treated Ti substrates, notice cracks through
the thickness of the HA coating (arrows); and F: grit blasted followed by heat treatment Ti substrates.

the coating, at normal loads of 13.13N (Table I), with a tests of the coatings deposited on the acid-treated Ti sub-
corresponding friction force of 0.5N. No cracks were visi- strates [Figure 10(C), arrow 1C] was recorded at 4.17N,
ble within the scratch tracks before the localized detach- which was followed by the catastrophic failure of the coat-
ment of the coating [Figure 10(A)]. ing [Figure 10(C), arrow 2C] at typical values of 4.89N with
Scratch experiments on the apatite coatings deposited the corresponding friction force of 1.2N. The onset of the
on the grit blasted and then heat-treated Ti substrates catastrophic failure of the apatite coatings deposited on the
showed the first crack through the thickness of the coating alkali-treated substrates [Figure 10(D), arrow 1D] was
[Figure 10(B), arrow 1B] typically at normal loads of measured at 4.17N, and the corresponding friction force
10.24N (Table I), followed by the first detachment event was 0.9N. Coatings deposited on the control Ti substrates
[Figure 10(B), arrow 2B] at normal loads 10.34N (Table I). started to show first cracks [Figure 10(E), arrow 1E] at nor-
The corresponding friction force for the delamination was mal loads of the scratching indenter of about 3.88N,
0.4N. The onset of the crack appearance under scratch whereas the onset of delamination of the coatings [Figure

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | OCT 2011 VOL 99B, ISSUE 1 65
FIGURE 10. Scratches on HA coatings deposited on a (A) grit blasted, (B) grit blasted and heat treated, the initial load of the indenter was 10N
and the final load was 17N. (C) Acid-treated, (D) alkali-treated Ti substrates, the initial load of the indenter was 4N and the final load was 11N.
Scratches on HA coatings deposited on a (E) ground to 800 grit (control) and (F) ground to 800 grit followed by heat treatment Ti substrates, the
initial load of the indenter was 1N and the final load was 6N. Arrows show an onset of a crack or a delamination event along scratch tracks of
the coatings.

10(E), arrow 2E] was observed at 5.89N, corresponding fric- higher pH on the conformation of BMP 2 molecules have
tion force was 1.8N. Apatite coatings deposited on the heat not yet been carried out.
pretreated Ti substrates demonstrated catastrophic failure The result of the EDS analysis of the coating [Figure
[Figure 10(F), arrow 1F] at the loads of 1.14N, the corre- 7(A)] showed that the prepared coatings were calcium defi-
sponding friction force was 0.1N. The critical shearing cient, compared with the stoichiometric HA and therefore
strength values corresponding to the critical loads of the the prepared apatite coatings resembled the calcium defi-
coatings were calculated according to the Eq. (1) and the cient apatite structure of inorganic phase of bone or den-
results are summarized in Table I. tine. Although the microstructural morphology of the coat-
ings deposited on the different pretreated and control Ti
substrates was similar (Figure 8), the coating morphologies
DISCUSSION
at a larger scale were relatively different among various
The XRD analysis [Figure 7(B)] confirmed that HA coatings
groups (Figure 9). The mechanical performance of the coat-
on pretreated Ti substrates were successfully prepared in
ings deposited on pretreated Ti substrates also demon-
mild conditions, at 75
C. Such a relatively low temperature
strated a significant variation (Figure 10, Table I).
may allow to employ such HA coatings as a carrier and
It is known that in air atmosphere and room tempera-
delivery device for bone forming drugs. Incorporating drugs
ture titanium spontaneously reacts with oxygen and forms a
into the HA coating during the coating deposition process
passive surface oxide layer in the order of a few nano-
may ensure a sustained and prolonged drug release into the
meters. This layer mainly consists of amorphous titanium
bony tissues surrounding the implant. Although BMP 2 is a
dioxide (TiO2) but other forms of Ti oxides, such as Ti2O3,
heat-sensitive drug, it has been demonstrated that BMP 2
may also be present. In general, TiO2 has a polymorphic
does not denature when subjected to heat treatment at
characteristic and can be formed in three crystalline phases,
70
C for 2 h.31 Also, it has been recently suggested that
anatase, rutile, and brookite, or amorphous structure. It has
changes in pH may play a role in the conformational
been previously reported that in comparison with crystal-
changes of BMP 2, however so far only the 4.5–6.5 pH range
line rutile phase of titania, the amorphous phase has
was investigated,32 and the studies on the influence of

TABLE I. Critical Load Measured During the Scratch Tests and the Calculated Failure Stress

Pretreatment Method Critical Load (N) Critical Shearing Stress (MPa)

Grit blasted 13.13 6 0.04 131.6 6 0.2

Grit blasted followed by heat treatment 10.34 6 0.04 116.5 6 0.2
Acid treated 4.89 6 0.06 79.7 6 0.5
Alkali treated 4.17 6 0.03 73.5 6 0.2
Machine ground to 800 grit 5.89 6 0.02 87.6 6 0.2
Machine ground to 800 grit followed by heat treatment 1.14 6 0.04 38.3 6 0.7

66 ZAVGORODNIY ET AL. MECHANICAL STABILITY OF HYDROXYAPATITE COATING ON TI SUBSTRATE

 ORIGINAL RESEARCH REPORT

inferior capability to induce apatite formation on Ti sub- with the machine-ground Ti substrate (control), resulted in
strate when it is incubated in simulated body fluid (SBF).33 a greater mechanical interlocking between the coating and
In our study, crystalline titania may form when the Ti sub- the substrate, which reflected in the substantial increase of
strates were incubated in acidic coating solution (during the critical shearing stress of the coating to the value of
deposition of monetite), and therefore, the relatively high 131.60 MPa (Table I) for the grit-blasted samples in com-
value of the critical shearing stress of the apatite coating de- parison to the control group. In the same time, in spite of
posited on Ti control group, 87.61 MPa (Table I), could be the cleaning procedure, the contamination of Ti surface was
attributed to the presence of crystalline phases of titania still clearly noticeable [Figures 2(C,D) and 3(C–E)]. The
and their affinity to apatite. implant surface contamination with Al2O3 particles may
Morphological observations of the apatite coatings de- have impact on bone healing in the peri-implant region
posited on rough grit-blasted Ti substrates [Figures 9(B)] through competition between the aluminum and calcium
demonstrated that the coatings appeared to be finer, than ions.35 Also the potential toxic effect of Al ions has been
those deposited on control substrates (Figure 9). The influ- well documented and cannot be dismissed.36 Further in
ence of the surface roughness on the energy barrier for the vitro and in vivo studies are needed to address these issues.
heterogeneous nucleation can be qualitatively evaluated in Remaining traces of Si [Figure 2(A)], on the other hand,
terms of the classical nucleation theory (CNT). The Gibbs may have positive effect on bone growth around implant.
free energy change (DG) due to the formation of a solid Although it has not yet been well established, Si is contrib-
phase on the rough surface is as follows: uting to the architecture of the calcified tissues such as
bone as a constituent element of some mucopolysacchar-
DG ¼ NDl þ rNL ANL þ ðrNS  rLS ÞANS (2) ides.37 Various hypotheses indicate that Si could be a cross-
linking agent for acid mucopolysaccharides, collagen, and
where N is the number of particles forming the solid phase elastin and therefore important for holding the organic ma-
cluster, Dl is the change in chemical potential of a single trix together in a structured way in connective tissues.37,38
particle on moving from a liquid to a solid phase (when Rutile film was formed on titanium by heat treatment (at
phase transition is favored, Dl is negative),rNL is the nu- 800
C, 1 h), which was confirmed by the XRD structural anal-
cleus-liquid interfacial energy, ANL is the nucleus-liquid ysis [Figure 4(A)]. In water-based solutions, the hydroxyl
interfacial area,rNS is the nucleus-substrate interfacial ener- groups attach to the Ti ion of TiO2 forming Ti-OH groups. The
gy,rLS is the liquid-substrate interfacial energy, and ANS is point of zero charge (PZC) of rutile is 5.9,39 which was
the nucleus-substrate interfacial area. In the real situation higher than pH 2.77 of the coating solution. Hence, the surface
of the rough surface, the Young’s equation: of the rutile film was protonised in contact with the coating
solution (positively charged groups were formed), which in
rLS  rNS ¼ rNL cos hY (3) turn attracted negatively charged phosphate ions (PO3 4 ,

HPO24 , H2PO4 ) from the coating solution that bonded to the
where yY is Young’s contact angle, has to be generalized: surface of the substrate and formed a negatively charged layer.
This negatively charged layer in turn attracts positively
rLS  rNS ¼ rNL cos h (4) charged highly mobile Ca2þ ions, which bind to the negatively
charged phosphate sites and, consequently, a metastable phase
where y is the apparent contact angle. According to Wenzel of amorphous calcium phosphate was formed. According to
equation: the calcium phosphates solubility diagram,40 monetite is the
only thermodynamically stable phase of calcium phosphates at
cos h ¼ rW cos hY (5)
pH 2.70. Therefore, the layer of amorphous calcium phosphate
(ACP) was then crystallized into the monetite phase of calcium
where rW is the roughness factor, defined as the ratio of the
phosphate and served as a precursor for the formation of the
actual surface area and the flat surface area34:
monetite coating on Ti substrate. Hydrothermal treatment of
ANS the prepared monetite coating in the alkali solution (pH
rW ¼ >1 (6) 12.80) converted the monetite crystals to apatite, which was
Aflat
confirmed by the XRD [Figure 7(B)]. According to the JCPDS
Taking into account Eqs. (4) and (5), Eq. (2) becomes: 21-1276 of the standard rutile XRD spectra, the relative inten-
sity of the XRD peaks corresponding to (101) and (110)
DG ¼ NDl þ rNL ANL  rNL ANS  rW cos hY (7) planes of the randomly oriented rutile powder is 0.6, whereas
the ratio of the XRD intensity peaks of rutile films synthesized
Equation (5) shows that the energy barrier for the het- in our study after the heat treatment of the Ti substrate was
erogeneous nucleation is reduced for rougher substrates 0.3 [Figure 4(A)]. It follows that rutile crystals, which formed
(higher rW). Additionally, the mechanical modification of the the titania film, had preferred orientation of the (110) crystal
surface by grit blasting substantially increased the specific plane, which has been demonstrated to have a good lattice
surface area of the substrate and subsequently increased match to apatite.20 This results in the reduction of stresses on
nucleation rate of calcium phosphate crystals. An increase the distribution boundaries between apatite and rutile, and
in the surface roughness after grit blasting, in comparison therefore increasing the adhesion between apatite and rutile

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | OCT 2011 VOL 99B, ISSUE 1 67
layers. On the contrarily, scratch tests of the HA coating depos- than that femur undergoes during gait (<31 MPa) and
ited on rutile substrate, prepared using heat-treatment, [Figure nearly six times higher than the value required for the
10(F)] demonstrated a catastrophic failure of the coatings at orthopaedic prosthesis (>22 MPa).
much lower loads than HA coatings deposited on the other
substrates (Table I), even though HA has a high affinity to ru- ACKNOWLEDGMENTS
tile. The rutile layer produced by thermal oxidation of the Ti The authors thank Dr. Avi Bendavid, CSIRO, for the preparation
substrate at 800
C or above has been reported to be mechani- of the grit-blasted Ti specimens. The assistance from the
cally unstable due to the large difference in coefficient of ther- ACMM, University of Sydney, is acknowledged.
mal expansion, large lattice mismatch, and large volume ratio
of rutile to Ti (1.73) phases.41 All these factors may explain
REFERENCES
the catastrophic delamination of the HA coatings deposited on 1. Geetha M, Singh AK, Asokamani R, Gogia AK. Ti based biomateri-
rutile films at relatively low loads under scratch tests. als, the ultimate choice for orthopaedic implants—A review. Prog
HA coatings deposited on the phosphoric acid treated Ti Mater Sci 2009;54:397–425.
substrate delaminated catastrophically [Figure 10(C)] at 2. Dorozhkin SV. Bioceramics of calcium orthophosphates. Biomate-
rials 2010;31:1465–1485.
normal loads of about 4.90N (Table I), which was signifi- 3. Arafat MT, Lam CXF, Ekaputra AK, Wong SY, Li X, Gibson I. Bio-
cantly lower, than that obtained for the control group mimetic composite coating on rapid prototyped scaffolds for
(5.90N). Using thin-film XRD (TF-XRD), it has been demon- bone tissue engineering. Acta Biomater. 2011;7:809–820.
4. Park IS, Choi UJ, Yi HK, Park BK, Lee MH, Bae TS. Biomimetic
strated that phosphoric acid treatment of Ti produced an
apatite formation and biocompatibility on chemically treated Ti-
amorphous layer of titania and it has been reported that 6Al-7Nb alloy. Surf Interface Anal 2008;40:37–42.
such layer does not induce the formation of apatite.33 The 5. Pattanayak DK, Matsushita T, Doi K, Takadama H, Nakamura T,
thick amorphous layer of titania formed by phosphoric acid Kokubo T. Effects of oxygen content of porous titanium metal on
its apatite-forming ability and compressive strength. Mater Sci
treatment may reduce the number of nucleation sites result- Eng C 2009;29:1974–1978.
ing in the weaker HA coating adhesion. 6. Wen CE, Xu W, Hu WY, Hodgson PD. Hydroxyapatite/titania sol-
HA coating deposited on the alkali pretreated Ti sub- gel coatings on titanium-zirconium alloy for biomedical applica-
tions. Acta Biomater 2007;3:403–410.
strate demonstrated poor coating coverage [Figure 10(D)].
7. Carradò A, Viart N. Nanocrystalline spin coated sol-gel hydroxy-
These observations indicated that the changes in the chemi- apatite thin films on Ti substrate: Towards potential applications
cal composition of the substrate after alkali treatment had for implants. Solid State Sci 2010;12:1047–1050.
reduced the ability to induce monetite coating nucleation 8. Yoon H-J, Song J-E, Um Y-J, Chae GJ, Chung S-M, Lee I-S, Jung
U-W, Kim C-S, Choi S-H. Effects of calcium phosphate coating to
and deposition in acidic conditions. After the conversion of SLA surface implants by the ion-beam-assisted deposition
the monetite coating into HA coating by hydrothermal treat- method on self-contained coronal defect healing in dogs. Biomed
ment in alkali conditions, the catastrophic delamination of Mater 2009;4:044107.
the HA coating under scratch test was recorded at 4.17N 9. Hu R, Lin C, Wang H, Tao T. Modulation effects of collagen I on
the structure of electrochemically deposited hydroxyapatite coat-
(Table I), which was significantly lower than that in the con- ing. Mater Lett 2010;64:915–917.
trol group (Table I). As a result of the alkali treatment, a 10. Yang G-L, He F-M, Hu J-A, Wang X-X, Zhao S-F. Biomechanical
layer of sodium hydrogen titanate forms on the surface of comparison of biomimetically and electrochemically deposited
hydroxyapatite-coated porous titanium implants. J Oral Maxillo-
the Ti substrate,42 which is negatively charged. After the
fac Surg 2010;68:420–427
protonisation in the acidic conditions during the formation 11. Zavgorodniy AV, Borrero-López O, Hoffman M, LeGeros RZ, Roha-
of the monetite coating, the surface free energy of the sub- nizadeh R. Characterization of the chemically deposited hydroxy-
strate was significantly reduced, which in turn resulted in apatite coating on a titanium substrate. J Mater Sci Mater Med
2011;1:1–19.
the decreased nucleation ability of Ti substrate43 for mone- 12. Rohanizadeh R, LeGeros RZ, Harsono M, Bendavid A. Adherent
tite and consequently for apatite crystals. apatite coating on titanium substrate using chemical deposition.
J Biomed Mater Res A 2005;72A:428–438.
CONCLUSIONS 13. Sun L, Berndt CC, Gross KA, Kucuk A. Material fundamentals and
clinical performance of plasma-sprayed hydroxyapatite coatings:
Apatite coatings were prepared on pretreated Ti substrates A review. J Biomed Mater Res B 2001;58:570–592.
in mild conditions, at 75
C, which opens the possibility to 14. Kurtz S, Mowat F, Ong K, Chan N, Lau E, Halpern M. Prevalence
incorporate temperature sensitive drugs, such as BMP-2, of primary and revision total hip and knee arthroplasty in the
into the apatite coating during the coating deposition pro- United States from 1990 through 2002. J Bone Joint Surg 2005;
87:1487–1497.
cess. Using apatite coating on orthopaedic or dental prosthe- 15. Chung Y-Y, Ki S-C, So K-Y, Kim D-H, Park KH, Lee YS. High revi-
sis as drug carrier can reduce healing time after the implan- sion rate of hydroxyapatite-coated ABG-I prosthesis. J Orthop Sci
tation surgery, promoting the performance and the stability 2009;14:543–547.
of implants. Chemistry and morphology of the pretreated Ti 16. Vidalain J-P. Twenty-year results of the cementless Corail stem.
Int Orthop 2011;35:189–194.
substrates play crucial role in the induction and formation 17. Schnettler R, Knöß PD, Heiss C, Stahl J-P, Meyer C, Kilian O,
of calcium phosphate coatings and its adhesion to Ti sub- Wenisch S, Alt V. Enhancement of bone formation in hydroxyapa-
strate. In conclusion, the apatite coating on Ti substrate tite implants by rhBMP-2 coating. J Biomed Mater Res B 2009;
using two-step chemically deposition method had highest 90B:75–81.
18. FDA. Calcium Phosphate (Ca-P) Coating Draft Guidance for Prepa-
adhesion when the Ti substrates were grit blasted. The criti- ration of FDA Submissions for Orthopedic and Dental Endosseous
cal shearing stress of apatite coating on grit-blasted Ti sub- Implants. Washington, DC: Food and Drug Administration; 1992.
strate was 131 MPa, which is more than four times higher pp 1–14.

68 ZAVGORODNIY ET AL. MECHANICAL STABILITY OF HYDROXYAPATITE COATING ON TI SUBSTRATE

 ORIGINAL RESEARCH REPORT

19. ISO 13779-2. Implants for surgery - Hydroxyapatite - Part 2: Coat- 31. Yano K, Hoshino M, Ohta Y, Manaka T, Naka Y, Imai Y, Sebald W,
ings of hydroxyapatite. International Organization for Standariza- Takaoka K. Osteoinductive capacity and heat stability of recombi-
tion; Geneve, Switzerland, 2008. nant human bone morphogenetic protein-2 produced by Esche-
20. Wu J-M, Hayakawa S, Tsuru K, Osaka A. Low-temperature prepa- richia coli and dimerized by biochemical processing. J Bone
ration of anatase and rutile layers on titanium substrates and Miner Metab 2009;27:355–363.
their ability to induce in vitro apatite deposition. J Mech Behav 32. Luca L, Capelle MAH, Machaidze G, Arvinte T, Jordan O, Gurny R.
Biomed 2004;87:1635–1642. Physical instability, aggregation and conformational changes of
21. Rohanizadeh R, Al-Sadeq M, LeGeros RZ. Preparation of different recombinant human bone morphogenetic protein-2 (rhBMP-2). Int
forms of titanium oxide on titanium surface: Effects on apatite J Pharm 2010;391:48–54.
deposition. J Biomed Mater Res A 2004;71A:343–352. 33. Cui X, Kim H-M, Kawashita M, Wang L, Xiong T, Kokubo T,
22. Kokubo T, Pattanayak DK, Matsushita T, Takadama H, Nakamura Nakamura T. Preparation of bioactive titania films on titanium
T. Positively charged bioactive titanium oxide formed on Ti metal metal via anodic oxidation. Dent Mater 2009;25:80–86.
by acid and heat treatments. Bioceramics 22: The 22nd Interna- 34. Wenzel RN. Resistance of solid surfaces to wetting by water. Ind
tional Symposium on Ceramics in Medicine; Daegu, South Korea, Eng Chem 1936;28:988–994.
2009. pp 249–252. 35. Esposito M, Hirsch J-M, Lekholm U, Thomsen P. Biological fac-
23. Kokubo T, Miyaji F, Kim H-M, Nakamura T. Spontaneous forma- tors contributing to failures of osseointegrated oral implants, (II).
tion of bonelike apatite layer on chemically treated titanium met- Etiopathogenesis. Eur J Oral Sci 1998;106:721–764.
als. J Am Ceram Soc 1996;79:1127–1129. 36. El-Rahman SSA. Neuropathology of aluminum toxicity in rats
24. Mohammadi Z, Ziaei-Moayyed AA, Mesgar AS-M. Grit blasting of (glutamate and GABA impairment). Pharmacol Res 2003;47:
Ti-6Al-4V alloy: Optimization and its effect on adhesion strength 189–194.
of plasma-sprayed hydroxyapatite coatings. J Mater Process Tech 37. Gibson IR, Best SM, WB. Chemical characterization of silicon sub-
2007;194:15–23. stituted hydroxyapatite. J Biomed Mater Res 1999;44:422– 428.
25. ISO_4287. Geometrical Product Specifications (GPS)—Surface 38. Sprio S, Tampieri A, Landi E, Sandri M, Martorana S, Celotti G,
Texture: Profile Method-Terms, Definitions and Surface Texture Logroscino G. Physico-chemical properties and solubility behav-
Parameters. International Organization for Standarization; Gen- iour of multi-substituted hydroxyapatite powders containing sili-
eve, Switzerland, 1997. con. Mater Sci Eng C 2008;28:179–187.
26. ASTM C1624–05. Standard Test Method for Adhesion Strength 39. Kosmulski M. The significance of the difference in the point of
and Mechanical Failure Modes of Ceramic Coatings by zero charge between rutile and anatase. Adv Colloid Interface Sci
Quantitative Single Point Scratch Testing. ASTM International, 2002;99:255–264.
West Conshohocken, PA, USA, 2005. 40. Brown PW. Phase relationships in the ternary system CaOAP2O5AH2O
27. Benjamin P, Weaver C. Measurement of adhesion of thin films. P at 25
C. J Am Ceram Soc 1992;75:17–22.
R Soc A 1960;254:163–176. 41. Siva Rama Krishna D, Brama YL, Sun Y. Thick rutile layer on
28. Weaver C. Adhesion of thin films. J Vac Sci Technol 1975;12: titanium for tribological applications. Tribol Int 2007;40:329–334.
18–25. 42. Morgado E Jr, de Abreu MAS, Pravia ORC, Marinkovic BA, Jardim
29. Bull SJ, Berasetegui EG. An overview of the potential of quantita- PM, Rizzo FC, Araújo AS. A study on the structure and thermal
tive coating adhesion measurement by scratch testing. Tribol Int stability of titanate nanotubes as a function of sodium content.
2006;39:99–114. Solid State Sci 2006;8:888–900.
30. Oliver WC, Pharr GM. An improved technique for determining 43. Wang XJ, Li YC, Lin LG, Hodgson PD, Wen CE. Apatite-inducing
hardness and elastic modulus using load and displacement sens- ability of titanium oxide layer on titanium surface: The effect of
ing indentation experiments. J Mater Res 1992;7:1564–1583. surface energy. J Mater Res 2008;23:1682–1688.

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | OCT 2011 VOL 99B, ISSUE 1 69