Materials Transactions, Vol. 43, No. 2 (2002) pp.

261 to 266
c 2002 The Japan Institute of Metals
Surface Characterization of Amorphous ZrAl(Ni, Cu) Alloys Immersed
in Cell-Culture Medium
Sachiko Hiromoto
1
, Katsuhiko Asami
2
, An Pang Tsai
1
and Takao Hanawa
1
1
National Institute for Materials Science, Tsukuba 305-0047, Japan
2
Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan
For the evaluation of zirconium-base amorphous alloys as biomaterials, the surface compositions of amorphous ZrAlNiCu and ZrAl
Cu alloys were characterized using X-ray photoelectron spectroscopy. The alloys were polished, autoclaved, and immersed in Hanks solution
or Eagles minimum essential medium containing fetal bovine serum, MEM + FBS. Aluminum was enriched in the surface oxide lm and
the substrate just under the lm of the alloys polished in water. After autoclaving, aluminum and copper were enriched in the substrate while
zirconium was preferentially oxidized and incorporated into the surface oxide lm. In Hanks solution, copper and nickel decreased in the
substrate and surface oxide lm, resulting in the enrichment of aluminum in the substrate. In MEM+FBS, zirconium preferentially decreased
by the effects of amino acids and proteins while copper was enriched in the substrate. The surface composition of zirconium-base amorphous
alloys was much inuenced by amino acids and proteins in MEM+FBS.
(Received September 14, 2001; Accepted Decemebr 17, 2001)
Keywords: zirconium-base amorphous alloy, surface composition, cell-culture medium, biomaterial
1. Introduction
The biomedical use of zirconium-base amorphous alloys
is attempted to utilize the advantages of amorphous struc-
ture in biomaterials, such as bone plate, wires, and dental
implants. The zirconium-base amorphous alloys show high
tensile strength over 1500 MPa and low Youngs modulus un-
der 100 GPa,
1)
which can improve mechanical biocompatibil-
ity of the biomaterials. On the other hand, all materials ap-
plied in human body must show high corrosion resistance as
well as high strength and low Youngs modulus. Therefore,
anodic polarization behavior in various biological environ-
ments and repassivation behavior of the amorphous ZrAl
NiCu alloy were examined to evaluate the corrosion behav-
ior of the alloy.
26)
The anodic polarization curve of the alloy
showed high polarization resistance and low passive current
density as well as enough resistance to pitting corrosion in
phosphate buffered solutions containing chloride ion.
25)
The
total charge of repassivation reaction of the alloy was lower
than that of crystalline pure zirconium.
6)
These results indi-
cate that the zirconium-base amorphous alloy is sufciently
corrosion resistant in the human body.
On the other hand, the biocompatibility of zirconium-base
amorphous alloy was not examined. The biocompatibility is
dened as the ability to exist in an in vivo environment for
an acceptable period of time with no detrimental effect on
the host.
7)
Therefore, the alloy surface is required neither
to contain nor to release any elements showing cytotoxicity,
carcinogenicity, and so on. The surface characterization of
zirconium-base bulk amorphous alloy is rarely performed as
far as the authors knowledge.
8)
Hence, surface composition
of the alloy must be understood to discuss the corrosion resis-
tance and biocompatibility of the alloy.
The surface of biomaterial contacts directly to body uid
containing various inorganic ions such as chloride and phos-
phate ions and organic compounds such as amino acids and
proteins. The amino acids and proteins are adsorbed by
metals
9, 10)
and sometimes accelerate the dissolution of met-
als.
11, 12)
Therefore, the alloys, whose surface has been char-
acterized for the evaluation of the corrosion resistance and
biocompatibility, are necessary to be exposed to the biologi-
cal environment.
In this study, the surface characterization was performed on
amorphous ZrAlNiCu and ZrAlCu alloys immersed in
Hanks solution, or a cell-culture medium. Effect of inorganic
ions on the surface reaction was evaluated using the Hanks
solution and that of amino acids and proteins was evaluated
using the cell-culture medium.
2. Experimental Methods
2.1 Specimens
Master alloys with the nominal atomic composition of
Zr
65
Al
7.5
Ni
10
Cu
17.5
and Zr
65
Al
7.5
Cu
27.5
were prepared by arc
melting of pure zirconium, aluminum, nickel, and copper
(>99.9 at%) in an argon atmosphere. Weight of the resul-
tant master alloy was > 99.9% of the gross weight of source
metals. The master alloys were re-melted in a quartz nozzle
and rapidly quenched by casting the melt on the rotating cop-
per roll with argon gas pressure. Ribbon alloys with a width
of 25 mm and a thickness of 5055 m were obtained. The
structures of the ribbon alloys were determined to be amor-
phous by X-ray diffraction method with Cu K radiation as
shown in elsewhere.
2, 5)
Free- and roll-sides of the ribbon alloys were polished with
600 grid SiC paper in distilled water and cut into pieces
with a length of 1520 mm. The specimens were ultrasoni-
cally rinsed in distilled water and acetone and dried for a few
days to generate air-formed passive lm and kept in a dessi-
cator. As-polished specimens are named Polished. The
Polished specimens were sterilized in an autoclave under
a saturated vapor pressure at 393 K for 1.8 ks. The specimens
after autoclaving are named Autoclaved. Subsequently, the
Autoclaved specimens were immersed in 5 ml of Hanks
262 S. Hiromoto, K. Asami, A. P. Tsai and T. Hanawa
Table 1 Compositions of Hanks solution and MEM+FBS.
Composition
Concentration/kmol m
3
Hanks solution MEM+FBS
NaCl 1.37 10
1
1.16 10
1
KCl 5.37 10
3
5.36 10
3
CaCl
2
1.26 10
3
1.80 10
3
Na
2
HPO
4
4.23 10
4
8.98 10
4
KH
2
PO
4
4.41 10
4

Mg
2
SO
4
7.39 10
4
8.11 10
4
NaHCO
3
4.17 10
3
2.382.62 10
2
Amino acid 5.5 10
3
Protein/kg m
3
< 10
solution
13)
(Hanks) or Eagles minimum essential medium
14)
containing 10 vol% fetal bovine serum (MEM + FBS). The
immersion was performed in the air containing 5% CO
2
and a saturated water vapor at 310 K for 2.4 Ms (28d) in
an incubator. The specimens after immersion in Hanks and
MEM+FBS are named Hanks and MEM+FBS, respec-
tively. The composition of Hanks solution and MEM+FBS
are summarized in Table 1.
2.2 SEM and EDS
The free-side surfaces of Polished and MEM+FBS of
the ZrAlNiCu alloy were observed by a scanning electron
microscope, SEM, (HITACHI, S-3100HT) with accelerating
voltage of 25 kV.
Bulk composition of Polished of the ZrAlNiCu alloy
was measured using an energy dispersive X-ray spectrometer,
EDS, (OXFORD MODEL 7021, LINK ISIS) with accelerat-
ing voltage of 20 kV. Three different regions on the free-side
surface of Polished with 20 m 20 m were measured.
The bulk composition was calculated from the intensity of an
EDS peak of each alloying element.
2.3 XPS
X-ray photoelectron spectroscopy, XPS, was performed
with an electron spectrometer (SSI-SSX100). X-ray source
was the monochromatized Al K line (1486.61 eV). Take-
off angle of photoelectron was 35

to specimen surface. All

binding energies in this paper are relative to the Fermi level.
The spectrometer was calibrated against Au 4f
7/2
(binding
energy, 84.07 eV) and Au 4f
5/2
(87.74 eV) of pure gold and
Cu 2p
3/2
(932.53 eV), Cu 2p
1/2
(952.35 eV), and Cu Auger
L
3
M
4,5
M
4,5
line (kinetic energy, 918.65 eV) of pure copper.
The energy values were based on published data.
15)
The free-
side surface of the specimen was characterized. The compo-
sition and thickness of the surface oxide lm and composi-
tion of the substrate just under the surface oxide lm were
simultaneously calculated according to the method described
elsewhere.
16, 17)
Empirical
1821)
and theoretically calculated
22)
data of relative photoionization cross-sections was used for
quantication. The relative photoionization cross-sections,

ij
/
O 1s
, are summarized in Table 2, where relative photoion-
ization cross-section of a level j electron of an element i to
that of O 1s electrons.
Table 2 Values of
ij
/
O 1s
used in the quantication of the surface com-
position.
Photoionization cross-section
Level
Zr 3d Zr 3d Al 2p Al 2p Cu 2p
3/2
Cu 2p
3/2
(Metal) (Oxide) (Metal) (Oxide) (Metal) (Oxide)

ij
/
O1s
2.40 2.40 0.187 0.187 5.71 5.71
Reference 19 19 15 15 19 19
Level
Ni 2p
3/2
Ni 2p
3/2
P 2p Na 1s Cl 1s S 2p
(Metal) (Oxide)

ij
/
O1s
7.47 1.53 2.59 2.91 1.017 0.515
Reference 17 17 17 19 18 16
Level N 1s C 1s

ij
/
O1s
0.644 0.341
Reference 19 19
3. Results and Discussion
3.1 SEM and EDS
SEM images of Polished and MEM+ FBS of the Zr
AlNiCu alloy are shown in Fig. 1. On MEM + FBS,
scratches similar to those on Polished were observed. In
addition, no pit and corrosion product were observed on
MEM + FBS, indicating that the signicant corrosion did
not occur on the alloy during immersion in MEM+FBS.
The composition of the ZrAlNiCu alloy determined by
EDS is shown in Table 3. Concentration of aluminum was
smaller than the nominal concentration. Aluminum seemed to
be slightly underestimated because of its lownominal concen-
tration and relatively low sensitivity in EDS due to its small
atomic weight. The concentrations of other alloying elements
were almost same as the nominal concentration. Thus, the
bulk composition was assigned as the nominal composition.
3.2 Composition of substrate just under surface oxide
lm
XPS peaks of Zr 3d and Ni 2p were decomposed into peaks
originating from metallic and oxidic states according to bind-
ing energy data.
23, 24)
The spectrum at Al 2p electron region
was overlapped with that of Cu 3p electron region, then the
spectrum was decomposed into four peaks originating from
metallic and oxidic states of Al 2p electrons and from Cu
3p
3/2
and Cu 3p
1/2
electrons, according to the binding en-
ergy data.
23, 24)
The spectrum at Cu 2p electron region was
decomposed into two peaks. One of the peaks is assigned to
an overlapped peak of metallic state, Cu
0
, and cuprous state,
Cu
+
, and the other to cupric state, Cu
2+
.
23, 24)
Because the dif-
ference in the binding energy between Cu
0
and Cu
+
states is
small, its meaningful decomposition was impossible. In ad-
dition, intensity of Auger Cu L
3
M
4,5
M
4,5
peaks, which could
assist the decomposition of Cu
0
and Cu
+
, was almost unde-
tectable. Thus, the spectrum Cu
0
+ Cu
+
was conventionally
treated as metallic state peak. This will cause overestimation
of the amount of Cu in metallic state. The photoelectrons
originated from metallic state are assumed to be excited from
the alloy substrate just under the surface oxide lm of the al-
loy.
Compositions of the alloy substrate of the ZrAlNiCu
and ZrAlCu alloys determined by XPS are shown in Fig. 2.
Surface Characterization of Amorphous ZrAl(Ni, Cu) Alloys Immersed in Cell-Culture Medium 263
(a) Polished (b) MEM+BS
Fig. 1 SEM images of Polished and MEM+FBS of amorphous ZrAlNiCu alloy. (a) Polished; (b) MEM+FBS.
Table 3 Bulk composition of Zr
65
Al
7.5
Ni
10
Cu
17.5
amorphous alloy ana-
lyzed by EDS (at%).
Concentration (at%)
Zr Al Ni Cu
68.9 0.3

3.7 0.2 9.7 0.1 17.7 0.4

: Standard deviation.
Bulk compositions of the alloys are also shown as a reference.
The composition of MEM+FBS of the ZrAlNiCu alloy
was not determined because photoelectrons in metallic state
were not detected probably due to the thick layer of adsorbed
amino acids and proteins. Concentrations of zirconium, [Zr],
of Polished decreased to zero after an autoclave treatment,
indicating that zirconium is preferentially oxidized and prob-
ably incorporated into the surface oxide. Concentrations of
aluminum, [Al], of Polished, Autoclaved, and Hanks
were higher than those in the bulk composition, indicating the
enrichment of aluminum in the substrate. Concentrations of
copper, [Cu], of Autoclaved and MEM+FBS were higher
than those in the bulk composition. Enrichment of copper in
the substrate of amorphous ZrAlNiCu alloy also occurs in
a deaerated phosphate buffered saline solution.
5)
In addition,
the enrichment of copper at interface between lm and sub-
strate is also observed on bulk amorphous Zr
55
Al
10
Cu
30
Ni
5
alloy anodized at 0.84 V(SCE) for 20 h in air-saturated 0.1-M
Na
2
SO
4
with pH 8.
8)
Thus, the enrichment of copper in the
substrate seems to occur when the alloy is oxidized. Concen-
trations of nickel, [Ni], of Polished and Autoclaved were
similar to that in the bulk composition and decreased to zero
after immersion in Hanks.
The [Al] and [Cu] increased and [Zr] decreased after au-
toclaving. Thus, the enrichment of aluminum and copper in
the substrate was enhanced with the preferential oxidation of
zirconium. After immersion in Hanks solution, the [Al] in-
creased while the [Cu] and [Ni] decreased. After immersion
in MEM + FBS, the [Al] decreased to zero while [Cu] in-
0
10
20
30
40
50
60
70
80
90
100
Zr Al Cu
Element
A
t
o
m
i
c

C
o
n
c
e
n
t
r
a
t
i
o
n
,

C
(b) Zr-Al-Cu
Substrate
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
0
10
20
30
40
50
60
70
80
90
100
Zr Al Cu Ni
Element
A
t
o
m
i
c

C
o
n
c
e
n
t
r
a
t
i
o
n
,

C
Bulk
Polished
Autoclaved
Hanks
MEM+FBS
(a) Zr-Al-Ni-Cu
Substrate
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
Fig. 2 Compositions of the alloy substrate just under the surface oxide lm
on the amorphous (a) ZrAlNiCu and (b) ZrAlCu alloys. Undetected:
Photoelectrons in metallic state of the element were not detected.
creased up to 100 at% on the ZrAlCu alloy, while substrate
composition of MEM+FBS of the ZrAlNiCu alloy was
not obtained.
264 S. Hiromoto, K. Asami, A. P. Tsai and T. Hanawa
Fig. 3 Thickness, t , of the surface oxide lm on the amorphous (a)
ZrAlNiCu and (b) ZrAlCu alloys. Uncalculated: t were not cal-
culated due to the lack of signals from elements in metallic state.
3.3 Thickness and composition of surface oxide lm
Thickness, t , of the surface oxide lm of the ZrAlNi
Cu and ZrAlCu alloys is shown in Fig. 3, wherein the t on
MEM+FBS of the ZrAlNiCu alloy could not be deter-
mined because no signal originating from metallic state were
detectable. The t was large in this order: MEM + FBS,
Hanks, and Autoclaved. Therefore, oxidation of the al-
loys proceeded in a saturated water vapor and aqueous solu-
tions.
In MEM + FBS, the amino acids and proteins form an
adsorption layer on the surface oxide lm
9, 10)
and the ad-
sorption layer was covered with so-called contaminant car-
bon layer.
16, 17)
However, the quantication method for three
layers on the substrate was not developed. Then, to apply the
conventional quantication method
16, 17)
in which the contam-
inant carbon and surface oxide layers are assumed to exist on
the substrate, the following two models were assumed. One
is that amino acids and proteins are uniformly contained in
the surface oxide lm and the contaminant carbon apparently
does not exist, where the amount of adsorbed amino acids and
proteins is estimated. The other is that all amount of carbon is
originated from the contaminant carbon, although some part
of the carbon should be coming from adsorbed amino acids
and proteins in MEM+ FBS. In any case, it should be men-
tioned that both models are too simplied, however, the lat-
ter model is reasonable to quantify the surface oxide lm be-
cause the amino acids and proteins in the surface oxide lm
on MEM+FBS is negligible if they were incorporated.
Based on the former model, the compositions of the sur-
face oxide lm of MEM + FBS of the ZrAlNiCu and
ZrAlCu alloys were calculated as listed in Table 4. Nickel
was not incorporated into the lm of the ZrAlNiCu alloy.
Concentrations of zirconium, aluminum and copper of both
alloys were similar each other. Difference in these concen-
trations between both alloys was reduced because concentra-
tions of carbon and nitrogen were too large. Concentrations
of carbon and nitrogen of both alloys were also similar each
other. Therefore, both alloys adsorb almost the same amount
of amino acids and proteins. In any case, this model could not
be used for the discussion on the composition of the surface
Table 4 Composition of the surface oxide lm on MEM+FBS assuming
that adsorbed amino acids and proteins dispersed in the lm.
Alloy
Concentration (at%)
Zr Al Cu Ni P Na Cl S O C N
ZrAlNiCu 2.9 1.2 0.0 0.0 0.3 0.7 0.3 0.0 26.0 58.5 10.1
ZrAlCu 3.9 1.0 0.1 0.4 0.0 0.3 0.7 23.4 58.5 11.8
Polished
Autoclaved
Hanks
MEM+FBS
0
5
10
15
20
25
Zr Al Cu Ni P Na Cl S
Element
A
t
o
m
i
c

C
o
n
c
e
n
t
r
a
t
i
o
n
,

C
0
10
20
30
40
50
60
70
80
90
100
O
A
t
o
m
i
c

C
o
n
c
e
n
t
r
a
t
i
o
n
,

C
(a) Zr-Al-Ni-Cu
Surface oxide
U
n
d
e
t
e
c
t
e
d
N
o

e
x
i
s
t
e
n
c
e
N
o

e
x
i
s
t
e
n
c
e
N
o

e
x
i
s
t
e
n
c
e
N
o

e
x
i
s
t
e
n
c
e
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
Zr Al Cu P Na Cl S O
0
5
10
15
20
25
Element
A
t
o
m
i
c

C
o
n
c
e
n
t
r
a
t
i
o
n
,

C
0
10
20
30
40
50
60
70
80
90
100
A
t
o
m
i
c

C
o
n
c
e
n
t
r
a
t
i
o
n
,

C
(b) Zr-Al-Cu
Surface oxide
N
o

e
x
i
s
t
e
n
c
e
N
o

e
x
i
s
t
e
n
c
e
N
o

e
x
i
s
t
e
n
c
e
N
o

e
x
i
s
t
e
n
c
e
U
n
d
e
t
e
c
t
e
d
U
n
d
e
t
e
c
t
e
d
Fig. 4 Compositions of the surface oxide lm on the amorphous (a)
ZrAlNiCu and (b) ZrAlCu alloys. Undetected: Photoelectrons in
metallic state of the element were not detected. No existence: The element
did not exist in the environment.
oxide lm.
Based on the latter model with the contaminant carbon cov-
ering surface oxide lm, compositions of the surface oxide
lm of all specimens were calculated as shown in Fig. 4.
In addition, relative concentration of aluminum to zirco-
nium, [Al]/[Zr] was calculated as shown in Fig. 5. Phospho-
rus, sodium, and chlorine contained in Hanks solution and
MEM+ FBS were detected in Hanks and MEM+ FBS
of the ZrAlNiCu alloy whereas sulfur was not detected.
Phosphorous, sodium, chlorine, and sulfur were detected in
Hanks and MEM+FBS of ZrAlCu alloys, except that
sodium in MEM+ FBS and chlorine in Hanks were not
detected.
The binding energies of P 2p electrons were 132.6
134.5 eV, indicating that phosphorus existed as phosphate
ions, H
x
PO
(3x)
4
, and phosphate salt, such as zirconium
phosphate.
24)
Concentration of phosphorous of Hanks was
larger than that of MEM + FBS of both alloys, indicat-
ing that the adsorbed amino acids and proteins probably in-
Surface Characterization of Amorphous ZrAl(Ni, Cu) Alloys Immersed in Cell-Culture Medium 265
Fig. 5 Relative concentration of aluminum to zirconium, [Al]/[Zr], on the
amorphous (a) ZrAlNiCu and (b) ZrAlCu alloys.
hibits the incorporation of H
x
PO
(3x)
4
into the surface oxide
lm. The binding energies of Na 1s electrons were 1072.0
1072.1 eV, originating from Na
+
, and those of Cl 2p electrons
were 197.8199.2 eV, originating from Cl

as a chloride.
24)
Sodium and chloride ions were possibly located over the sur-
face oxide lm because sodiumchloride is not incorporated in
the surface oxide lm on ZrTi alloy.
25)
The binding energy
of S 2p electrons was 169.1 eV in Hanks and 164.1 eV in
MEM+FBS of the ZrAlCu alloy, and those were origi-
nating from SO
2
4
and sulde, respectively.
24)
Therefore, the
SO
2
4
is incorporated from Hanks solution into the surface
oxide lm. On the other hand, the sulde is originated from
the adsorbed amino acids and proteins.
The ratio [Al]/[Zr] in the bulk composition was 0.12 that
was smaller than those in the Polished of both alloys
(Fig. 5), indicating that aluminum was enriched in the sur-
face oxide lm. After autoclaving, during which the surface
is oxidized without releasing any alloying elements, the [Zr]
in the surface oxide lm of the ZrAlNiCu alloy increased,
whereas that of the ZrAlCu alloy decreased (Fig. 4). On
the other hand, the ratio [Al]/[Zr] in the surface oxide lm of
both alloys decreased (Fig. 5) and [Zr] in the alloy substrate
decreased (Fig. 2), indicating that zirconium is preferentially
oxidized against aluminum and incorporated into the surface
oxide lm. Concentrations of copper and nickel, [Cu] and
[Ni], in the surface oxide lm increased after autoclaving as
shown in Fig. 4, and [Cu] especially increased as much as
[Al].
After immersion in Hanks solution, the [Zr] and [Al] in the
surface oxide lm decreased (Fig. 4). The ratio [Al]/[Zr] of
the ZrAlNiCu alloy did not change (Fig. 5). On the other
hand, that of the ZrAlCu alloy decreased (Fig. 5), indicat-
ing that aluminum preferentially dissolves against zirconium.
The [Ni] decreased to zero and the [Cu] also decreased after
immersion in Hanks solution (Fig. 4). As mentioned previ-
ously, the [Cu] and [Ni] in the alloy substrate also decreased
after immersion in Hanks solution (Fig. 2). Therefore, cop-
per and nickel dissolve from the alloy in aqueous solutions
containing inorganic ions.
After immersion in MEM + FBS, the [Zr] in the surface
oxide lm decreased (Fig. 4), leading to the increase in the
Fig. 6 Relative concentration of OH

to O
2
, [OH

]/[O
2
] on the amor-
phous (a) ZrAlNiCu and (b) ZrAlCu alloys.
ratio [Al]/[Zr] as shown in Fig. 5. Therefore, the prefer-
ential dissolution of zirconium from the surface oxide lm
can be promoted by amino acids and proteins. The [Cu] and
[Ni] in the surface oxide lm decreased after immersion in
MEM+ FBS (Fig. 4), indicating that copper and nickel also
dissolve in the solution containing amino acids and proteins.
On the other hand, copper dissolves in the solution containing
amino acids and proteins less than in the solution consisting
of inorganic ions because copper was enriched in the alloy
substrate (Fig. 2). Stability of metal ions in solution contain-
ing organic compounds depends on the stability of complex of
the metal ions and organic compounds. Since the magnitude
of chelate formation constant of alloying elements with ethy-
lene diamine tetra acetic acid, EDTA, is larger in the order of
Zr
4+
> Cu
2+
> Ni
2+
> Al
3+
,
26)
stability of complex formed
between alloying elements and amino acids and proteins in
MEM+ FBS is possibly larger in the same order. As the re-
sult, zirconium dissolved preferentially to copper and nickel
in MEM+FBS.
3.4 Oxygen in surface oxide
Concentrations of oxygen species in MEM+ FBS were
larger than those of Polished, Autoclaved, and Hanks
of both alloys as shown in Fig. 4, because some part of oxy-
gen species of MEM+FBS came from the adsorbed amino
acids and proteins.
The XPS spectrum of O 1s was decomposed into three
spectra originating from O
2
, hydroxide or hydroxyl group,
OH

, and hydrate and/or adsorbed water according to the

binding energy.
27)
Relative concentrations of OH

to O
2
,
[OH

]/[O
2
], are shown in Fig. 6. The ratio [OH

]/[O
2
]
decreased after autoclaving, indicating that dehydration of
the surface oxide proceeded at 393 K in a saturated water va-
por. On the other hand, the ratio [OH

]/[O
2
] increased after
immersion in Hanks solution and in MEM + FBS, respec-
tively. Therefore, the hydration reaction of the surface oxide
seems to be preferential to the dehydration reaction in aque-
ous solutions. The [OH

] was much larger than [O

2
] on
MEM+FBS, indicating that the hydration reaction may be
preferential to the dehydration reaction with the existence of
amino acids and proteins. In this case, some part of [OH

]
266 S. Hiromoto, K. Asami, A. P. Tsai and T. Hanawa
originated from the adsorbed amino acids and proteins.
3.5 Surface composition of the zirconium-base amor-
phous alloy in human body
Dissolution of copper and nickel from the ZrAlNiCu
and ZrAlCu alloys in Hanks solution and MEM + FBS
is indicated in this study. However, copper was enriched in
the substrate after immersion in MEM+ FBS though nickel
and copper in the surface oxide lm dissolved. Therefore,
the dissolution of nickel and copper possibly completes after
releasing all amount in the lm. Hence, the continuous dis-
solution of copper and nickel through the implantation in the
human body may not occur.
The phosphate salt such as zirconium phosphate was
formed on the ZrAlNiCu and ZrAlCu alloys, indicat-
ing that the surface of the zirconium-base amorphous alloy is
inert in body uid.
25)
4. Conclusions
In the ZrAlNiCu and ZrAlCu alloys polished in wa-
ter, aluminum is enriched in the surface oxide lm and sub-
strate just under the lm. After autoclaving, zirconium is
preferentially oxidized and aluminum and copper are thus en-
riched in the substrate. By immersion in Hanks solution,
copper and nickel dissolve from the alloy, leading to enhance
the enrichment of aluminum in the substrate. In MEM+FBS,
amino acids and proteins cause the preferential dissolution of
zirconium and aluminum in the substrate disappears, causing
the signicant enrichment of copper in the substrate. Phos-
phorous in the solutions is incorporated in the surface oxide
lm and forms phosphate salts, such as zirconium phosphate.
On the other hand, the incorporation of phosphorous is inhib-
ited by amino acids and proteins.
REFERENCES
1) A. Inoue: Acta Mater. 48 (2000) 279306.
2) S. Hiromoto, A.-P. Tsai, M. Sumita and T. Hanawa: Corros. Sci. 42
(2000) 16511660.
3) S. Hiromoto, A.-P. Tsai, M. Sumita and T. Hanawa: Corros. Sci. 42
(2000) 21672185.
4) S. Hiromoto, A.-P. Tsai, M. Sumita and T. Hanawa: Corros. Sci. 42
(2000) 21932200.
5) S. Hiromoto, K. Asami, A.-P. Tsai, M. Sumita and T. Hanawa: J. Elec-
trochem. Soc. (in press).
6) S. Hiromoto and T. Hanawa: Electrochimica Acta (in press).
7) D. Hill: Design Engineering of Biomaterials for Medical Devices,
(Wiley, Chichester, 1998) pp. 157.
8) A. Gebert, K. Buchhols, A. Leonhard, K. Mummert, J. Eckert and L.
Schultz: Mater. Sci. Eng. A 267 (1999) 294300.
9) A. P. V. A. do Serro, A. C. Fernandes, B. de J. V. Saramago and W.
Norde: J. Biomed. Mater. Res. 46 (1999) 376381.
10) B. A. Ivarsson, P.-O. Hegg, K. I. Lundstrom and U. L. F. Jonsson: Col-
loids and Surfaces 13 (1985) 169192.
11) C. G. F. Clark and D. F. Williams: J. Biomed. Mater. Res. 16 (1982)
125134.
12) Y. Mu, T. Kobayashi, M. Sumita, A. Yamamoto and T. Hanawa: J.
Biomed. Mater. Res. 49 (2000) 238243.
13) J. H. Hanks and R. E. Wallace: Proc. Soc. Exp. Biol. Med. 71 (1949)
196.
14) H. Eagle: Science 130 (1959) 432.
15) K. Asami: J. Electron. Spectrosc. 9 (1976) 469478.
16) K. Asami, K. Hashimoto and S. Shimodaira: Corros. Sci. 17 (1977)
713723.
17) K. Asami and K. Hashimoto: Corros. Sci. 24 (1984) 8397.
18) H. Yoshioka, H. Habazaki, A. Kawashima, K. Asami and K. Hashimoto:
Corros. Sci. 32 (1991) 313325.
19) K. Teramoto, K. Asami and K. Hashimoto: Corros. Sci. 27 (1978) 57
61.
20) K. Hashimoto, M. Kasaya, K. Asami and T. Masumoto: Corros. Eng.
26 (1977) 445452.
21) K. Hashimoto, K. Asami and K. Teramoto: Corros. Sci. 19 (1979) 314.
22) J. H. Scoeld: J. Electron Spec. 8 (1976) 129137.
23) C. D. Wagner: Practical Surface Analysis, 2nd ed., Vol. 1, Eds. D.
Briggs and M. P. Seah (Wiley, Chichseter, 1990) App. 5, pp. 595.
24) C. D. Wagner, W. M. Riggs, L. E. Davies, J. F. Moulder and G. E.
Muilenberg: Handbook of X-ray Photoelectron Spectroscopy, (Perkin-
Elmer, Eden Prairie, MN, 1979) pp. 54.
25) T. Hanawa, O. Okuno and H. Hamanaka: J. Japan Inst. Metals 56 (1992)
11681173.
26) K. Ueno: Cheletmetric Titration, 10th ed., (Nanko-do, Tokyo, 1964)
pp. 3132.
27) K. Asami and K. Hashimoto: Corros. Sci. 17 (1977) 559570.