Bone bonding behavior of titanium and its alloys

when coated with titanium oxide (TiO,) and titanium

silicate (Ti&)

Toshiaki Kitsugi,' Takashi Nakamura,l.* Masanori Oka? Wei-Qi Yan,' Tatsuya Goto; Takehiro Shibuya;
Tadashi K ~ k u b oand
, ~ Shiro Miyaji4
'Department of Orthopaedic Surgery, Faculty of Medicine; 2Reseavch Center for Biomedical Engineering; 3Depurtment of
Applied Chemistry, Faculty of EngineerinR,
-
Kyoto
.
University, Kyoto, Japan; 4Nippon Electric Glass Co. Lfd., Otsu,
Shigu, lapun

It has been proposed that the essential requirement for artifi- *

cial materials to bond to living bone is the formation of bone-
like apatite on their surfaces in the body. Recent studies have
shown that titanium hydrogel and silica gel induce apatite
formation on their surface in a simulated body fluid. In this
study, the influence of titanium oxide and titanium silicate
on the bonding of titanium alloys to bone was studied.
Rectangular implants (15 X 10 X 2.2 mm) of titanium, Ti-6Al-
4V, Ti-bAl-2Nb-Ta,Ti-6Al-4V coated with TiO,, and Ti-6A1-4V
coated withTi& were implanted into the tibia1metaphyses of
mature rabbits. At 8 and 24 weeks after implantation, the tibiae
containing the implants were dissected out and subjected to
a detaching testing. The failure load for titanium, Ti-6Al-
4V, Ti-6Al-ZNb-Ta, Ti-6A1-4V coated with TiOz, and Ti-
6A1-4V coated with Ti& were, respectively, 0.68 ? 0.48,
0.22 -t 0.46, 0.67 0.59, 2.18 2 0.71 and 2.03 ? 0.41 kgf at
* *
8 weeks, and 2.7 0.91, 2.58 1.29, 2.38 ? 0.41, 3.79 ? 1.7,
and 2.79 ? 0.87 kgf at 24 weeks after implantation. Histologi-
cal examination by Giemsa surface staining, CMR, and SEM-
EPMA revealed the coated titanium alloy implants directly
bonded to bone tissue during early implantation. A Ca-P
layer was observed at the interface of the coated implants
and the bone. The results of this study indicated that TiOz
and Ti&, can enhance the early bonding of titanium alloys
to bone by inducing a Ca-P layer (chemical apatite) on the
surface of titanium alloys. It also is suggested that the direct
bone contact occurs in relation to the calcium and phospho-
rus adsorption onto the surface of the titanium passive layer
formed during long-term implantation. 0 1996 John Wiley &
Sons, lnc.

INTRODUCTION taining phosphorus, calcium, and hydroxyl groups,

Titanium and its alloys have been used successfully
for dental and orthopedic implants,'-3 because they
have good biocompatibility. When titanium is im-
planted into human bone, it forms a direct contact
with bone tissue, a process termed o~seointegration.~,~ceramics. Silica is a principal component of the bioac-
Ultrastructural analysis has revealed a dense amor-
phous layer separating the titanium implants from
bone An oxide passive layer on the surface of
titanium and its alloy implants grows slowly during
the implantati~n,',~-'~ and a titanium hydrogel is
formed on the surface of the passive layer. Phosphate
ions and calcium ions are incorporated into the oxide
layer and the titanium hydrogel. Oxides on titanium
and titanium alloys change into complex layers con-
*Towhom correspondence should be addressed at Depart- nium oxide (TiOz) and titanium silicate (Ti&) and
ment of Orthopaedic Surgery, Faculty of Medicine, Kyoto
University, 54 Kawaharacho Shogoin, Sakyo-ku, Kyoto
606, Japan.
which then bond to the apatite of bone t i s ~ u e . ' , ~ - ' ~
Bioactive glasses or glass-ceramics, especially glass-
ceramics containing apatite and ~ollastonite,"-'~ have
been used in orthopedic surgery as vertebral bodies
or iliac bone spacers. An intermediate Ca-P apatite
layer is formed at the interface of bone and glass-
tive glasses and glass-ceramics.'h-'RThus it is thought
that hydrated silica plays a primary role in forming
apatite on the surface of bioactive glass-ceramics. Sil-
ica gel can act as a stimulant to introduce hydroxyapa-
tite heterogeneous nucleation,"-** after which sponta-
neous growth is effected by taking calcium ions and
phosphate ions from the surrounding body fluid. As
shown in the above-mentioned reports, titanium hy-
drogel and silica gel induce apatite nucleation on
their
In this study, we coated a titanium alloy with tita-
examined whether or not these coatings enhanced the
bonding of the titanium alloy to bone.

Journal of Biomedical Materials Research, Vol. 32, 149-156 (1996)

0 1996 John Wiley & Sons, Inc. CCC 0021-9304/96/020149-08
150 KITSUGI ET AL.

fii
MATERIALS AND METHODS Detaching T e s t
7

Implant preparation

Titanium, Ti-6A1-4V, Ti-GAl-2Nb-Ta (Kobe Steel co.,

Ltd., Kobe, Japan), Ti-6A1-4V coated with TiOz,and Ti-
15mm
6A1-4V coated with TiSSil were used. Compositions
of the metals used in this study are listed in Table
I. The surfaces of titanium, Ti-6A1-4V, and Ti-6A1-
2Nb-Ta were sandblasted with #60 corundum (A1203
powder, average diameter 250 pm). The surface rough- 2 mm
nesses of titanium, Ti-6A1-4V, Ti-6Al-2Nb-Ta, Ti-6A1-
4V coated with TiOz,and Ti-6A1-4V coated with Ti,Si3
were Ra = 1.59, 1.61, 1.41, 1.07, and 1.61 pm, re-
spectively, and Rmax = 7.70, 15.38, 15.38, 13.23, and
16.26 pm, respectively. Surface roughness was exam-
ined by a surface texture measuring instrument (Surf-
com 558AD, Tokyo Seimitsu). J
Titanium alloys were prepared to make Ti02 and
Ti5Si3(Nippon Electric Glass Co. Ltd., Otsu, Shiga, Ja-
pan). Ti02 was made by surface oxidation of Ti-6Al-
4V. The titanium alloys were kept at 800°C for one h Figure 1. Operative method and a detaching test
under atmosphere. Titanium silicate was made with a
coating of glass paste. Glass paste (glass powder mixed
with an acrylic binder and isobutyl methacrylate) was tests, as previously described (Fig. 1).During the de-
spread on both sides of Ti-6Al-4V with a thickness of taching test, each segment was held by a hook con-
50 to 100 pm and sintered at 900°C for 2 min. Then nected to an Instron-type autograph (Model s-500, Shi-
the surfaces of the titanium alloys coated with glass madzu Seisakusho Co., Ltd., Kyoto). Traction was
paste were soaked in HC1 for 12 h. Next, surface finish- applied vertically to the implanted plate at a crosshead
ing of the glass paste was performed and the layer of speed of 3.5 cm /min. The failure load when an implant
silicate exposed. The shear strength of the coating film became detached from the bone or when the bone itself
was 20 MPa. The coated samples were cleaned and broke was then measured. Data were analyzed using
defatted with water, acetone, and alcohol. a one-way analysis of variance and a two-tailed Stu-
dent’s f test.

Bone bonding examination Histological examination

Rectangular implants (15 mm X 10 mm X 2.2 mm)
were implanted into the metaphyses of the bilateral
tibiae of mature male rabbits weighing between 2.5 and
3.0 kg, according to the method previously described.I2
The rabbits were divided into 10 groups of seven ani-
mals each and sacrificed with an overdose of Nembutal
at 8 and 24 weeks after implantation. Tibia1 segments
containing the implant were prepared for detaching
After the detachment test, the specimens in which
the bone did not become detached were fixed in 1OY0
phosphate-buffered formalin for 7 days and then were
dehydrated in Yo%, 80%,90%, and 100% ethanol, for 3
days at each concentration, and embedded in polyester
resin. Sections 1 mm thick were cut on a milling ma-
chine with a diamond disk (Crystal Cutter, Marto Ltd.,
Tokyo). Using a speed lap (Marto Ltd., Tokyo), some

TABLE I
Compositions of Metals Used in the Present Study
Composition/ Wt%

Metal Ti A1 Mo Zr V Nt, Ta Fe C N 0 H
Ti Bal. - - - - - - 0.06 - 0.003 0.074 0.0014
Ti-6A1-4V Bal. 6.45 - - 4.16 - - 0 18 - 0.004 0.068 0.0117
Ti-6Al-2Nb-Ta Bal. 3.50 15.17 5.13 - - - 0.03 0.006 0.009 0.100 0.0075
Specimens and Compositional data from Kobe Steel Co., Ltd., Kobe 651, Japan.
BONE BONDING BEHAVIOR OF Ti AND Ti ALLOYS
Figure 2. Photograph of bone and implant after the detach-
ing test. (a) Surface of titanium alloy coated with TiOZ.
(b) Surface of titanium alloy coated with Ti&. (a) and (b) The
breakage of coating layer is observed after the detaching test.
were then ground to about 200 pm in thickness for
Giemsa surface stainingz5and contact microradiogra-
phy. The bone-implant interface was observed with
light microscopy.
The surface was prepared for scanning electron mi-
croscopy-electron probe microanalysis (SEM-EPMA)
by polishing the 2 mm sections with diamond paper.
Prior to analysis, these samples were coated with a
thin layer of carbon. They were then analyzed using
a scanning electron microscope (Hitachi S-800, Tokyo)
connected to an energy-dispersive X-ray microana-
lyzer (Horiba EMAX-3000, Kyoto). X-ray intensities for
silicon, calcium, phosphorus, and titanium were ana-
lyzed across the implant/bone interface. An electron
beam (0.04 p m in diameter and at 3 X amp) was
used, and the X-ray intensities in counts per second
(cps) were recorded. The accelerating voltage was
20 kv (full scale Ca-P, 500 cps; Si, 500 cps; Ti, 1000 cps).
RESULTS
Evaluation of bonding ability and failure load
Some of the tested noncoated samples (titanium and
titanium were detached fromthebone before the
detaching test. Therefore the failure load was defined
as 0 kgf. Macroscopically, the coating layer of TiOzand
151
Ti& was partially broken after the detaching test both
8 and 24 weeks after implantation (Fig. 2). At 24 weeks
after implantation, detachment occurred on one side but
never occurred simultaneously on both sides.
At 8 weeks after implantation, the failure loads
for Titanium, Ti-6A1-4V, Ti-GAl-2Nb-Ta, Ti-6A1-4V
coated with TiOz, and Ti-6A1-4V coated with Ti&
were 0.68 ? 0.48 kgf, 0.22 ? 0.46 kgf, 0.67 ? 0.59 kgf,
2.18 2 0.71 kgf, and 2.03 ? 0.41 kgf, respectively. And
at 24 weeks after implantation, the failure loads were
2.70 ? 0.91 kgf, 2.58 ? 1.29 kgf, 2.38 ? 0.41 kgf,
3.79 ? 1.57kgf, and 2.79 2 0.87 kgf, respectively. Films
of TiOz and Ti& enhanced early bonding of the tita-
nium alloy (Table 11). The failure loads of the coated
samples differed significantly from those of noncoated
titanium and its alloys ( p < 0.01, Student’s t test). The
failure loads of titanium and titanium alloys increased
with the passage of time ( p < 0.01). In Ti-6A1-4V coated
with Ti&, there was no significant difference in the
failure load with the passage of time. At 24 weeks after
implantation, the failure loads of Ti-6A1-4V coated with
TiOz differed significantly from those of Ti-6Al-2Nb-
Ta ( p < 0.05). There was no significant difference
among the uncoated samples at 24 weeks after implan-
tation.
Evaluation of the bone-implant interface
(Figs. 3-5)
In titanium, Ti-6A1-4V, and Ti-6Al-2Nb-1Ta at 8
weeks after implantation, a gap between the implant
and the bone tissue was observed and stained with
Giemsa lipid at the interface of the bone and the im-
plant. No inflammation or foreign body reaction was
observed around the metal implants. At 24 weeks after
TABLE I1
Detaching Failure Loads (Kgf)
8 Weeks 24 Weeks
Ti 0.68 5 0.48 ( n = 12) 2.70 f 0.91 ( n = 14)
Ti-6A1-4V 0.22 2 0.46 (n = 10) 2.58 t 1.29 (n = 10)
Ti-6Al-2Nb-Ta 0.67 t 0.59 ( n = 14) 2.38 ? 0.41 ( n = 10)
Ti02 2.18 t 0.71 ( n = 12) 3.79 t 1.75 ( n = 8)
Ti& 2.03 L 0.67 ( n = 8) 2.79 ? 0.87 ( n = 10)
n = sample numbers; mean value t- standard deviation.
Ti (8 weeks) vs. Ti (24 weeks): p < 0.01; Ti-6A1-4VTi (8 weeks)
vs. Ti-6A1-4VTi (24 weeks): p < 0.01; Ti-6Al-2Nb-Ta (8 weeks)
vs. Ti-6Al-2Nb-Ta (24 weeks): p < 0.01; TiOz (8 weeks) vs.
TiO, (24 weeks): p < 0.01; Ti (8 weeks) vs. Ti-6A1-4V (8
weeks): p < 0.01; Ti (8 weeks) vs. Ti02 (8 weeks): p < 0.01;
Ti (8 weeks) vs. Ti& (8 weeks): p < 0.01; Ti-6A1-4V (8 weeks)
vs. TiOz (8 weeks): p < 0.01; Ti-6A1-4V (8 weeks) vs. T& (8
weeks): p < 0.01; Ti-bAl-zNb-Ta (8 weeks) vs, TiOz (8 weeks):
< 0.01; Ti-bAl-zm-Ta ( 8 weeks) vs. Ti5Si3(8 weeks): p <
0.01; Ti-6Al-2Nb-Ta (24 weeks) vs. TiO, (24 weeks): p < 0.01.
152 KITSUGI ET AL.

Figure 3. Contact microradiogram (original magnification X20) of bone (*) and ceramic (*). (a) Surface of Ti-6A1-4V at 8
weeks after implantation. Bone ingrowth is observed at the implant surface. There is a gap between the implant and the
bone. (b) Surface of titanium alloy coated with Ti& at 8 weeks after implantation. (c) Surface of titanium alloy coated with
TiOzat 24 weeks after implantation. (b) and (c) Bone is seen in direct contact with the implant surface without any fibrous tissue.

Figure 4. Giemsa surface staining (original magnification x20) of bone (*) and metal (*). (a) Surface of titanium at 8 weeks
after implantation. Soft tissue is observed and stained with the blue color of Geimsa liquid at the interface of bone and
implant. (b) Surface of titanium alloy coated with TiOz at 8 weeks after implantation. (c) Surface of titanium alloy coated
with Ti& a t 24 weeks after implantation. (b) and (c) Many osteocytes with canaliculi near the boundary to the implant. No
inflammation or foreign body reaction is observed at the interface.
BONE BONDING BEHAVIOR OF Ti AND Ti ALLOYS 153

Figure 5. Scanning electron micrograph and electron probe micro analysis (SEM-EPMA, original magnification X2000,
[---I = 15 pm): bone y), metal (*), P (b),Si (*), Ca (+), Ti (*). Full scale Ti, 2000 cps; Ca, P, Si, 500 cps. (a)
Interface of TiOp and bone tissue at 8 weeks after implantation: 1, Si and P; 2, Ca and Ti. (b) Interface of TiSi? and
bone tissue at 8 weeks after implantation: 1, Si and P; 2, Ca and Ti. (c) Interface of Ti5Si3and bone tissue at 24 weeks
after implantation: 1, Si and P, 2, Ca and Ti. SEM-EPMA reveals a decrease in Si and an increase in P, but no change
in Ca across the implant/bone interface.

implantation, a contact microradiogram showed that
bone ingrowth occurred on the rough surface of the
implant. At high magnification, a very thin layer of
Giemsa-stained tissue was seen in some instances at
the interfaces of metal implants and trabeculae.
In titanium alloys coated with TiO, and Ti&, bone
could be seen in intimate contact with the surfaces of
implants. Bone ingrowth was observed at the bone-
implant interface. Many osteocytes with canaliculi
were observed near the boundary to the implant.
Giemsa surface staining showed no intervening soft
tissue. No inflammation or foreign body reaction could
be detected at the interface.
SEM-EPMA showed a Ca-P layer at the interface of
Ti-6A1-4V coated with TiOz or Ti-6A1-4V coated with
Ti5Si3and bone tissue. Line analysis across the interface
revealed that X-ray intensities of Ca and 'I were in-
creased, whereas those of Si were decreased. There
was no breakage, detachment, or adsorption of the
coating layer of the TiO, and Ti5Si3.No interposition
of soft tissue was observed at the coating layer-bone
interface.
154
DISCUSSION
In this study, the coating of Ti02 and Ti& en-
hanced the bone bonding ability of titanium alloys
in an early-term implantation as compared with the
noncoated titanium and its alloys. The results of
SEM-EPMA also indicated precipitation of calcium
and phosphate on the surfaces of titanium alloys. It
is posited that Ti02 and TiSi3 changed to gel-derived
titanium, which may be an inducer for early hydroxy-
apatite formation because of its abundant TiOH
groups. It also was noted that in long-term implanta-
tion, the bone bonding ability of titanium and its
alloys increased. The observation of a passive film
formed during long-term implantation provides fur-
ther insight into the unique osseointegration of tita-
nium. Both an increase in oxide thickness and the
incorporation of elements from the extracellular fluid
(P, Ca, Si) into the oxide on the surface of pure
titanium and titanium alloys may occur as a function
of implantation time. Titanium and its alloys may
be bioactive (the formation of the Ca-P layer at the
interface of bone and metal) when their hydrolysis
can be accelerated in long-term implantation.
It is reported that silica glass and quartz do not
induce apatite formation in a simulated body fluid.22
One of the main differences among silica gel, silica
glass, and quartz is the density of silanol groups (SiOH)
on their surfaces. The silanol groups abundant on the
surface of silica gel might be responsible for apatite
formation. In our previous in vivo failure loads
in silica glass and the mother glass of AW-GC at 10
weeks after implantation were 0.18 t- 0.24 and 3.06 t
1.29 kgf, respectively, and at 25 weeks after implanta-
tion they were 1.30 1.18 and 3.88 i 1.06 kgf, respec-
tively. SEM-EPMA observation showed the formation
of a Si-rich layer and a Ca-P layer between bone and
the mother glass of AW-GC. It is thought that the Si
layer is a vitreous silica-like gel. The Si layer may be
a weak zone for bonding between bone and the mother
glass of AW-CC, therefore making failure loads for the
mother glass of AW-GC weak. Failure loads for silica
glass were low because silica glass is one of the bioin-
ert materials."
From the results of this present study, it is considered
that Ti& forms a silica gel layer in early-term implan-
tation. The silica gel layer plays a role in hydroxyapa-
tite nucleation and crystallization. However, the failure
loads were similar to those of the mother glass of AW-
GC. Probably the silica gel also was a weak zone for
bone bonding.
It has been known that a Ca-P-rich (apatite) layer
formed on the surface of glass-ceramics plays an essen-
tial role in chemical bonding between glass-ceramics
and bone tissue. The Ca-P-rich layer consists of a
carbonate-containing hydroxyapatite of small crystal-
lites and/or defective structures.I8These compositional
KITSUGI ET AL.
and structural characteristics of the surface apatite are
common to those of the bone apatite. It is thought
that some sort of covalent or electrostatic interaction
between valence shell electrons occurs between the Ca-
P-rich layer and the apatite of bone tissue. In this study,
bone bonding to titanium alloys coated with TiOl and
Ti&, may have been associated with formation of the
Ca-P layer, which, in turn, accelerated the bone bond-
ing behavior of titanium alloys. In the SEM-EPMA
observations of the present study, it is possible to detect
a Ca-P layer at the coating layer and bone tissue. The
coating of TiO, and Ti& interconnected the bone with
the metal implant.
From a mechanical point of view, failure load is not
the ultimate tensile strength because of the difficulty
in measuring the bone/implant contact area, which is
necessary for calculating the tensile strength.28Since
the procedure for the detaching test is simple and con-
sistent, the results are clear and easy to compare. Thus
the detaching test is considered especially useful as
an initial test for measuring implant/bone bonding
strength.'' 28 Our published data of failure loads are
shown in Table III.'2262y ' O Nakamura'*'" reported the
failure loads for AW-GC, Bioglass 45S5, and alumina
ceramics as follows: AW-GC was 7.43 t 1.19 kgf at 8
weeks after implantation and 8.15 ? 3.93 kgf at 25
weeks after implantation; Bioglass was 2.75 -+ 1.8 kgf
at 8 weeks after implantation; and alumina was
0.18 & 0.02 kgf at 8 weeks after implantation. Another
previously published failure load for AW-GC was
7.12 ? 1.37 kgf at 10 weeks and 7.14 2 2.28 kgf at 25
weeks after implantation.I7Under load-bearing condi-
tions, the failure loads in AW-GC and alumina 12
weeks after implantation were 2.06 -t 0.72 kgf and 0 2
0 kgf, respectively ( p < 0.01). Those of AW-GC and
alumina at 25 weeks after implantation were 12.9 I
3.32 kgf and 2.00 t- 1.41 kgf, respectively ( p < 0.01).2R
Takatsuka" reported that uncoated Ti alloy plates min-
imally bonded almost natatall to bone at 4 and 8 weeks,
but showed slight bonding at 25 weeks. The failure
loads for titanium were 0.0291 t 0.0403 kgf at 8 weeks
after implantation and 2.85 2 1.07 kgf at 25 weeks
after implantation.
TABLE 111
Failure Loads of Various Ceramics (Kgf)
Dense sintered alumina'2,'" 0.18 2 0.02
Bioglass 45S5-type g l a ~ s ' ~ , ~ " 2.75 t 1.80
SiOz glass at 10 weeks26 0.18 +- 0.24
SiOzglass at 25 weeks26 1.30 5 1.18
Mother glass of AW-GS at 10 weeks" 2.94 I1.77
Mother glass of AW-GC at 25 weeks2" 3.55 t 1.51
AW-GC",'" at 8 weeks 7.43 5 1.19
AW-GC",'" at 25 weeks 8.15 +- 3.93
Dense sintered k~ydroxyapatite'~,'~ 6.28 i 1.58
Dense sintered 13-3CaO . P20512,3n 7.58 I1.97
BONE BONDING BEHAVIOR OF Ti AND Ti ALLOYS
In the present study, the failure loads of titanium
and titanium alloys were lower than those of AW-GC
and calcium phosphate (hydroxyapatite and tricalcium
phosphate) and similar to those of alumina and Bio-
glass. They do not appear to have bioactive characteris-
tics in a mechanical bonding ability to bone tissue.
Especially in early-term implantations, titanium and
its alloys have lower abilities to bond to bone tissue
and are similar to bioinert materials. When implanted
for the long term, there is a possibility of calcium phos-
phate precipitation on the surfaces of titanium and its
alloys and of bone bonding ability. Although titanium
and its alloys showed good osseointegration under
light microscopy, they had no high mechanical ability
to bond to bone. In a clinical application, it is necessary
to obtain early bonding to bone tissue. Coating of TiO,
and Ti& on metal implants may be one of the methods
for obtaining stability in a short period of time when
using in load-bearing conditions.
References
1. P. Tengvall and I. Lundstrom, ”Physico-chemical con-
siderations of titanium as a biomaterial,” Clin. Mater.,
9, 115-134 (1992).
2. R. Van Nookt, ”Titanium: The implant material of to-
day,“ J. Mater. Res., 22, 3801-3811 (1987).
3. P. I. Branemark, B. 0. Hansson, R. Adell, U. Breine, J.
Lindstrom, 0.Hallen, and A. Ohman, ”Osseointegrated
implants in the treatment of the edentulous jaw. Experi-
ence from a 10-year period,” Scand. J. Plast. Reconstr.
S U U ~ .1, 1, 1-132 (1977).
4. L. Carlsson, T. Rostlund, B. Albrektsson, and P. I.
Branemark, ”Osseointegration of titanium implants,”
Acta Orthop. Scand., 57, 285-289 (1986).
5. J. E. Sundgren, P. Bodo, and 1. Lundstrom, ”Auger elec-
tron spectroscopic studies of the interface between hu-
man tissue and implants of titanium and stainless steel,”
J. Colloid Intevface Sci., 110, 9-20 (1986).
6. L. Seniierby, P. Thomsen, and L. E. Ericson, ”Ultrastruc-
ture of the bone-titanium interface in rabbits,“ /. Mnter.
Sci.:Mater. in Med., 3, 262-271 (1992).
7. P. Li, C. Ohtsuki, T. Kokubo, K. Nakanishi, N. Soga,
and K. de Groot, ”The role of hydrated silica, titanium,
and alumina in inducing apatite on implants,” J. Biomed.
Mateu. Res., 28, 7-15 (1994).
8. P. Li and K. de Groot, ”Calcium phosphate formation
within sol-gel prepared titania in vitro and in vim,”
J. Biomed. Mater. Res., 27, 1495-1500 (1993).
9. T. Hanawa, ”Titanium and its oxide film: A substratum
for formation of apatite,” in The Bone-Biomaterial Inter-
face, An International State of the Aut Workshop, Confer-
ence proceedings, University of Toronto Press, Toronto,
December 1990, pp. 49-61 and 3-4.
10. K. E. Hearly and P. Ducheyne, ”Hydration and prefer-
ential molecular adsorption on titanium in vitro,” Bioma-
tevials, 13, 553-561 (1992).
11. T. Kokubo, M. Shigemitsu, Y. Nagashima, M. Tashiro,
T. Nakamura, T. Yamamuro, and S. Higashi, “Apatite-
and wollastonite-containing glass-ceramic for pros-
thetic application,” Bull. Inst. Chem. Res, Kyoto Univ., 60,
260-268 (1982).
12. T. Nakamura, T. Yamamuro, S. Higashi, T. Kokubo,
and S. Ito, “A new glass-ceramic for bone replacement:
155
Evaluation of its bonding ability to bone tissue,“ J. Bio-
med. Mater. Res., 19, 685-698 (1985).
13. T. Kitsugi, T. Yamamuro, T. Nakamura, S. Higashi, Y.
Kakutani, K. Hyakuna, S. Ito, T. Kokubo, and T. Shi-
buya, “Bone bonding behavior of three kinds of apatite-
containing glass-ceramic,” J. Biomed. Mater. Res., 20,
1295-1307 (1986).
14. T. Yamamuro, T. Nakamura, S. Higashi, R. Kasai, Y.
Kakutani, T. Kitsugi, and T. Kokubo, “An artificial bone
for use as a bone prosthesis,” in Progress in Artificial
Organs--1983, Vol. 2, K. Atsumi, M. Maekawa, and
K. Ota (eds.), ISAO Press, No. 204, Cleveland, 1984,
pp. 810-814.
15. T. Kitsugi, T. Yamamuro, T. Nakamura, T. Kokubo, T.
Shibuya, and M. Takagi, ”SEM-EPMA observation of
three types of apatite-containing glass-ceramics im-
planted in bone: The variance of a Ca-P-rich layer,”
J. Biomed. Mnter. Res., 21, 1255-1271 (1987).
16. T. Kokubo, H. Kushitani, S. Sakka, T. Kitsugi, and T.
Yamamuro, ”Solutions able to reproduce in vivo sur-
face-structure changes in bioactive glass-ceramic A-
W,“ /. Bionied. Mateu. Res., 24, 721-734 (1990).
17. T. Kokubo, S. Ito, Z. T. Huang, T. Hayashi, S. Sakka, T.
Kitsugi, and T. Yamamuro, ”Ca-P-rich layer formed on
high-strength bioactive glass-ceramic A-W,” I. Biorried.
Mater. Res., 24, 331-343 (1990).
18. T. Kokubo, “Surface chemistry of bioactive glass-
ceramics,” ]. Noncrystalline Solids, 120, 138-151 (1990).
19. M. M. Pereira, A. E. Clark, and L. L. Hench, ”Calcium
phosphate formation on sol-gel-derived bioactive
glasses in vitru,” I. Biomrd. Mnter. Res., 228, 693-698
(1994).
20. P. Li, C. Ohtsuki, T. Kokubo, K. Nakanishi, T. Naka-
mura, and T. Yamamuro, ”Process of formation of bone-
like apatite layer on silica gel,” I. Mafrr. Sci.: Maler. in
Med., 4, 123-131 (1993).
21. P. Li, C. Ohtsuki,T. Kokubo,K. Nakanishi, T. Nakanishi,
N. Soga, T. Nakamura, and T. Yamamuro, ”Apatite
formation induced by silica gel in a simulated body
fluid,” J. Am. Ceram. Soc., 75, 2094-2097 (1992).
22. P. Li, C. Ohtsuki, T. Kokubo, K. Nakanishi, N. Soga,
T. Nakamura, and T. Yamamuro, “Effects of ions in
aqueous media on hydroxyapatite induction by silica
gel and its relevance to bioactivity of bioactive glasses
and glass-ceramics,” J. Appl. Bioniateu., 4, 221 -229
(1993).
23. K. E. Healy and P. Ducheyne, ”The mechanism of pas-
sive dissolution of titanium in a model physiological
environment,” J. Biomed. Mater. Res., 26,319-338 (1992).
24. T. Hanawa and M. Ohta, ”Calcium phosphate naturally
formed on titanium in electrolyte solution,” Bioninterials,
12, 767-774 (1991).
25. U. M. Gross and V. Strunz, ”Surface staining of sawed
sections of undecalcified bone containing alloplastics
implants,” Stain Technol., 52, 217-219 (1977).
26. T. Kitsugi, T. Yamamuro, T. Nakamura, and T. Kokubo,
”Bonding behavior of MgO-CaO-Si02-P205-CaF2 glass
(mother glass of A. W.-glass ceramics),” /. Biomed.
Muter. Res., 23, 631-648 (1989).
27. T. Kitsugi, T. Yamamuro, T. Nakamura, M. Oka, and T.
Kokubo, ”Influence of disodium (1-hydroxythylidene)
diphosphonate on bonding between glass-ceramics
containing appatite and wollastonite and mature rabbit
bone,” Calcq. Tissue Int., 52, 378-385 (1993).
28. T. Kitsugi, T. Yamamuro, and T. Kokubo, “Bonding
behavior of a glass-ceramic containing apatite and wol-
lastonite in segmental replacement of the rabbit tibia
156 KITSUGI ET AL.

under load-bearing conditions,” ]. Bone It. Surg., 71-A,
264-272 (1989).
29. T. Kitsugi, T. Yamamuro, T. Nakamura, S. Kotani, T.
Kokubo, and H. Takeuchi, “Four calcium phosphate
ceramics as bone substitutes for non-weight-bearing,”
Biomaterials, 14, 216-224 (1993).
30. T. Nakamura, T. Yamamuro, S. Higashi, Y. Kakutani,
T. Kitsugi, T. Kokubo, and S. Ito, “A new bioactive
glass-ceramic for artificial bone,” in Treatise on Biomedi-
cal Materials, Vol. I , T. Yamamuro, (ed.),Research Center
for Medical Polymers and Biomateriais, Kyoto University,
Kyoto, 1985, pp. 109-117.
31. K. Takatsuka, T. Yamamuro, T. Kitsugi, T. Nakamura,
T. Shibuya, and T. Goto, ”A new bioactive glass-
ceramic as a coating material on titanium alloy,” ]. Appl.
Biomater., 4, 317-329 (1993).
Received November 10, 1994
Accepted December 7, 1995