* Correspondence address: Department of Materials Science and

Engineering, Warsaw University of Technology, Narbutta 85, 02-524
Warsaw, Poland. Tel.: #[phone]; fax: #[phone].

E-mail address: [e-mail] (D. Krupa).

Biomaterials 22 (2001) 2139}2151

E!ect of calcium-ion implantation on the corrosion resistance

and biocompatibility of titanium

D. Krupa

*, J. Baszkiewicz , J.A. Kozubowski , A. Barcz, J.W. Sobczak, A. Bilin

H ski

,

M. Lewandowska-Szumie"

, B. Rajchel

Department of Materials Science and Engineering, Warsaw University of Technology, Wo!oska 141, 02-507 Warsaw, Poland

Institute of Electron Technology, Al.Lotniko&w 46, 02-668 Warsaw, Poland

Institute of Physical Chemistry of Polish Academy of Science, Kasprzaka 44/52, 01-224 Warsaw, Poland

Department of Transplantology and Central Tissue Bank, Centre of Biostructure Research, The Medical University of Warsaw,

Cha!ubinHskiego 5, 02-004 Warsaw, Poland

Institute of Nuclear Physics, Radzikowskiego 152, 31-342 Cracow, Poland

Received 21 July 2000; accepted 20 November 2000

Abstract

This work presents data on the structure and corrosion resistance of titanium after calcium-ion implantation with a dose of

10

Ca>/cm. The ion energy was 25 keV. Transmission electron microscopy was used to investigate the microstructure of the

implanted layer. The chemical composition of the surface layer was examined by XPS and SIMS. The corrosion resistance was
examined by electrochemical methods in a simulated body #uid (SBF) at a temperature of 373C. Biocompatibility tests in vitro were
performed in a culture of human derived bone cells (HDBC) in direct contact with the materials tested. Both, the viability of the cells
determined by an XTT assay and activity of the cells evaluated by alkaline phosphatase activity measurements in contact with
implanted and non-implanted titanium samples were detected. The morphology of the cells spread on the surface of the materials
examined was also observed. The results con"rmed the biocompatibility of both calcium-ion-implanted and non-implanted titanium
under the conditions of the experiment. As shown by TEM results, the surface layer formed during calcium-ion implantation was
amorphous. The results of electrochemical examinations indicate that calcium-ion implantation increases the corrosion resistance,
but only under stationary conditions; during anodic polarization the calcium-ion-implanted samples undergo pitting corrosion. The
breakdown potential is high (2.7}3 V).

2001 Elsevier Science Ltd. All rights reserved.

Keywords: Titanium; Implantation; Calcium; Corrosion; Calcium phosphate

1. Introduction

Among the metallic materials used in medicine for

implanting parts of the human body, titanium and its
alloys have the best properties. They show a high resist-
ance to corrosion in the environment of physiological

#uids and a good biocompatibility, in particular with

hard human tissues. The high corrosion resistance of
these materials is due to the presence of a TiO "lm on

the metal surface. As has been found, this oxide surface
layer reacts with the nonorganic ions, water and other

species contained in physiological #uids, thereby chang-
ing the properties of the surface [1]. Hanawa et al. [2,3]
have found that, in the simulated body #uid (SBF), a cal-
cium phosphate layer forms spontaneously on the tita-
nium surface and it is this layer which makes titanium
biocompatible. The formation of the apatite, which is
biologically active and whose structure is similar to hu-
man bones is the decisive factor for the integration of the
implant with the bone. The process of spontaneous
formation of calcium phosphate can be accelerated when
the metal surface is "rst subjected to a modifying treat-
ment, which, in the case of biomaterials, has preferably
been ion implantation [4]. The implantation of carbon
and nitrogen ions increases the wear resistance [5,6] and
the corrosion resistance [7}9]. In order to improve the
biocompatibility of titanium and its alloys, Hanawa et al.
[10}13] used calcium ions for the implantation and

0142-9612/01/$ - see front matter

2001 Elsevier Science Ltd. All rights reserved.

PII: S 0 1 4 2 - 9 6 1 2 ( 0 0 ) 0 0 4 0 5 - 1

Table 1
Chemical composition of simulated body #uid (SBF) (m

M

/dm

)

Na

>

K

>

Ca

>

Mg

>

Cl

\

HCO

\

HPO

\

SO

\

142

5

2.5

1.5

147.8

4.2

1

0.5

found that this accelerated the precipitation of calcium
phosphates; this e!ect was more pronounced in pure
titanium than in titanium alloys. They examined the
chemical composition and structure of the surface layers
thus formed and found that, during the calcium-ion im-
plantation, the thickness of the oxide layer increases and
the implanted ions occur in this layer in the form of
calcium oxide CaO and the complex calcium and tita-
nium oxide CaTiO. In in vivo examinations [14], cal-

cium-ion implantation also appeared to be advantageous
for the growth of the bone tissue.

The available literature reports are primarily con-

cerned with improving the bioactive properties of the
titanium surface. There have not however been studies
discussing the e!ect of calcium-ion implantation and the
formation of calcium phosphate precipitates upon the
corrosion resistance of the metallic substrate.

In the present study we attempted to "nd how the

modi"cation of the titanium surface by calcium-ion im-
plantation and the precipitation of calcium phosphates
a!ect the corrosion resistance of titanium.

2. Materials and methods

The material examined was pure commercial titanium

(grade 2). The test specimens in the form of discs 6.2 and
14 mm in diameter and 2 mm thick were polished one
side to the mirror "nish and then their surfaces were
implanted with calcium ions at a dose of 1

;

10

Ca>/cm. The ion energy was 25 keV. The condi-

tions under which the implantation was carried out were
selected so that the calcium concentration was maximum
on the surface. During the implantation, the temperature
of the samples did not exceed 403C. The vacuum in the
target chamber was about 10

\ Pa.

The implantation process was carried out in The Insti-

tute of Nuclear Physics in Cracow.

Characterization of the surface was carried out using

the methods:

E TEM (transmission electron microscope) structural

examinations were made with a Philips EM-300 trans-
mission electron microscope. The test samples were
cut by the electrospark method and then thinned on
the non-implanted surface until a perforation occur-
red.

E SIMS (secondary ion mass spectrometry) pro

"ling was

carried out using a Cameca IMS 6F instrument with
a cesium primary beam at an impact energy of 6 keV.
The species monitored were positively charged CsM
clusters, which enabled determination of the relative
concentration of the sample constituents M with re-
duced matrix e!ects.

E XPS (X-ray photoelectron spectroscopy) investiga-

tions were performed using a VG Scienti"c ESCALAB-

210 spectrometer with a Mg K X-ray source

(1253.6 eV).

The

working

pressure

was

below

5

;10\ Pa. The X-ray power supply was run at 12 kV

and 20 mA. Data analysis of the XPS spectra, involv-
ing non-linear background subtraction least-square
curve "tting by the mixed Gaussian}Lorentzian func-
tions, and quanti"cation, was performed by means of
the VG ECLIPSE program. The binding energy scale
was calibrated by a C 1 s peak at 285.0 eV.

E SEM (scanning electron microscope) was used for sur-

face observations.

The corrosion resistance was examined in a non-

deaerated SBF (simulated body #uid) at a temperature of
373C. The chemical composition of SBF is given in
Table 1. The solution was bu!ered with 50 m

M

trishyd-

roxymetyl aminometane (Trizma) and 45 m

M

hydrochlo-

ric acid (HCl) at a physiological pH of 7.4.

The

following

electrochemical

techniques

were

adopted:

E electrode impedance spectroscopy tests (EIS). The tests

were conducted at a corrosion potential (E) by

applying a sinusoidal potential signal of $50 mV. The
electrode response

was

analysed

in

the

range

0.01 Hz}10 kHz by means of a frequency response ana-
lyser. The program was controlled by the Soft-Kar
IMP-96 program. The results were analysed using the
EQUVCRT program [15];

E the linear polarization procedure (Stern

's method) for

measuring the polarization resistance (R). The

measurements were started at a potential of 20 mV
lower than the corrosion potential E and then the

potential was increased in the anodic direction until
the potential higher by 20 mV than E was achieved.

After each potential increment, the current and poten-
tial were allowed to stabilize for about 5 min. The
polarization resistance R was calculated by the least-

squares method.

E the polarization curves being measured, the samples

were polarized in anodic direction beginning from
a potential of !600 mV up to a potential of 5000 mV.
When the potential reached 5000 mV, it was decreased
at the same rate, down to a potential of 1000 mV so as
to obtain the return curves. The potential variation
rate was 20 mV/min.

Electrochemical examinations of the corrosion resist-

ance were made using an ATLAS 96 electrochemical kit.

2140

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

Fig. 1. Microstructure (a) and electron di!raction pattern (b) of tita-
nium before implantation.

The reference electrode was a saturated calomel elec-
trode. Prior to the measurements, the samples were
exposed to the test conditions for 13 h, and for approxim-
ately 1000}6000 h. The aim of the 13 h exposure was
to allow the corrosion potential E to stabilize.

The long-term exposures were used for examining how
the formation of the phosphate layer and the changes in
the properties of the oxide layers a!ect the corrosion
resistance of titanium. The reference material (non-im-
planted titanium) was also immersed in the SBF and
examined. After polarization, the samples were examined
using an optical microscope and scanning electron
microscope.

2.1. Biocompatibility study in vitro

Titanium samples, both implanted and non-implanted

in the form of cylinders 6.2 mm in diameter and 2 mm
high were placed at the bottom of 96-well culture dishes.
Then human osteoblasts were seeded on the surface of
the samples and on the surface of cell culture modi"ed
polystyrene (bottom of the culture dish) which served as
a control. Cells were seeded at a density of 15 000 cells per
well, and were cultured in a complete culture medium
supplemented with 1.25[OH]D at a concentration of

10n

M

. At least six samples of each type were tested, and

experiments were repeated twice.

2.2. Viability assay

After the week viability of the cells was determined by

means of the XTT assay. This test is used in toxicology
and is based on capacity of mitochondrial dehydroge-
nase enzymes in living cells for converting the XTT
substrate

(2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-

[(phenyloamino) carboxyl]-2H-tetrazolium hydroxide)
into a water-soluble formazan product. The "nal product
of the reaction is measured by the ELISA reader at
450 nm.

2.3. Alkaline phosphatase (ALP) activity measurement

ALP was chosen to label the condition of osteoblasts

because the changes in ALP activity correspond to the
stimulation of osteoblasts in culture. Moreover, it is one
of the most important promoter of osteogenesis in vivo,
and thus * an important factor in predicting the in vivo
situation on the basis of the observations performed in
the cell culture. ALP was measured by means of an
Alkaline Phosphatase SIGMA Diagnostic Kit after
8 days of exposure of cells to titanium samples.

2.4. Morphological observations

Morphological observations of the cells in the culture

wells were performed systematically during the observa-

tion period (8 days) by means of an inverted micro-
scope. After an 8 days observation period the cells were

"xed, dehydrated, and

subjected to morphological in-

vestigations using optical and scanning electron micro-
scopes.

3. Results

3.1. TEM results

The starting material contained dislocations and

high density of subgrains with blurred boundary that
occurred within the original grains with size above
1

m. This is characteristic microstructure of the

non-recrystallized material subjected to plastic deforma-
tion (Fig. 1). Fig. 2 shows a di!raction image of a layer
formed as a result of calcium-ion implantation. The blur-
red rings indicate that the surface layer has become
amorphous. Judging from the diameter of the rings we

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

2141

Fig. 3. Scanning electron micrographs of the titanium surface (a) and calcium-ion-implanted titanium (b) after long time exposure 3000 h.

Fig. 2. Electron di!raction pattern of calcium-ion-implanted titanium.

can estimate the spacing between the closest neighbours
at 2.33 A

s .

3.2. SEM results

An image of the samples exposed to SBF is shown in

Fig. 3. After the long-term exposures, on both implanted
and non-implanted titanium surfaces we can see precipi-
tates, which do not form a solid layer but have an insular
character.

3.3. SIMS results

Fig. 4a shows the theoretical depth distribution of

calcium, obtained using the TRIM code [16] and correc-
ted for the sputtering e!ect, which takes place during
implantation. The concentration pro"les of calcium, tita-
nium and oxygen observed in the surface layer formed by
calcium-ion implantation are shown in Fig. 4b. When

comparing the real and theoretical calcium concentration
pro"les we can see that in reality the calcium penetration
depth is greater than that calculated theoretically. The
maximum concentration occurs at a depth of about 4 nm.
Fig. 4b shows that the concentration of oxygen in the
implanted layer is increased to a depth of 20 nm but it is
lower than the concentration of calcium. Hence we can
infer that, partially, the implanted calcium occurs in a me-
tallic form. Fig. 4c shows the concentration pro"les of Ti,
O, Ca and P in the near-su rface layer of the samples
implanted with Ca ions and exposed in an SBF for 3000 h.
Fig. 4d shows the pro"les of these elements in non-im-
planted samples. When comparing the pro"les of Fig. 4b
and c, we can see that long-term exposures increase the
concentrations of calcium and oxygen. The near-surface
layer in addition contains phosphorus. The variation of
the compositions and concentrations of all these elements
can be attribu ted to the formation of calciu m phosphates
on the sample surface. The pro"les shown in Fig. 4d
indicate that phosphates also form on the surface of non-
implanted titanium, but here the concentrations of cal-
cium and phosphorus in the near-surface layer are lower
than those in the calcium-ion-implanted samples.

3.4. XPS results

The XPS method was used for an analysis of the

surface layers formed by calcium-ion implantation and,
for comparison, the surface layers on non-implanted tita-
nium. The samples were analysed before immersion and
after immersion. We also examined the composition of
the surface layers formed on the samples "rst exposed in
an SBF for 13 h and then polarized anodically. The XPS
results are given in Tables 2 and 3. Table 2 shows the
variation of the chemical composition of the layers and
Table 3 gives the values of the binding energy, (BE),
determined for the main components (Ti2p, O1s,

Ca2p, P2p) of the surface layers, and their percent-

age shares. The XPS spectra obtained were deconvoluted
using the data given in Refs. [2,17].

2142

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

Fig. 4. SIMS depth pro"les of elements: (a) theoretical calcium concentration depth pro"le in titanium. Depth pro"les of elements in calcium-ion-
implanted titanium; (b) before exposure; (c) after exposure for 3000 h; (d) depth pro"les of the elements in titanium after exposure for 3000 h.

Surface contaminants, such as carbon, silicon, alumi-

nium and nitrogen were identi"ed during the XPS analy-
sis in both exposed and non-exposed samples. The

atomic concentrations of Si, Al and N were below 2%.
The dominant contaminant was carbon, whose concen-
tration ranged from 24 to 80.3 at% (Table 2).

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

2143

Table 3
BE values of the main XPS peaks and relative chemical composition of the samples of titanium and calcium-ion-implanted titanium

Specimen

Time
(h)

Ti

O

Ca

P

(%)

Ti

>

(%)

Ti

>

(%)

Ti

>

(%)

Ti

(%)

O

\

(%)

OH

\

(%)

HO

(%)

PO

\

(%)

HPO

\

(%)

HPO\

(%)

Ti

0

458.5

456.4

454.1

530.0

532.0

534.0

82

10.5

7.5

68.8

26.7

4.5

3000

459.1

457.0

454.2

530.6

532.1

533.9

348.1

133.9

87.8

6.7

5.5

45.6

43.2

11.2

100

100

13

459.5

456.6

531.1

532.5

534.1

348.4

134.2

94.5

5.5

57.6

33.6

8.8

100

100

Ti}Ca

0

459.1

457.0

531.1

532.9

534.9

347.8

65.2

34.8

22.7

69.3

8

100

3000

459.2

457.4

455.7

531.0

532.5

534.1

348.2

133.9

81.9

11.7

6.3

60.6

30.6

8.8

100

100

13

458.7

530.2

531.6

534.1

347.1

133.2

100

66.7

23.9

9.4

100

100

Specimens after polarization.

Table 2
Relative concentrations of elements in surfaces of titanium and cal-
cium-ion-implanted titanium and the ratios of concentrations of cal-
cium to those of phosphorus, [Ca]/[P]

Specimen

Immersion
time (h)

Relative concentration (at%)

[Ca]/[P]

Ti

Ca

P

O

C

Ti

0

17.7

*

*

41.0

34.3

3000

12.4

2.9

5.0

53.5

26.2

0.57

13

14.2

2.9

4.4

52.3

26.2

0.65

Ti}Ca

0

1.4

1.5

*

17.8

80.3

3000

12.0

2.1

3.8

52.6

29.4

0.55

13

13.1

5.9

3.9

53.1

24.0

1.51

Specimens after polarization.

3.4.1. XPS analyses prior to immersion
3.4.1.1. Non-implanted titanium. From the deconvolu-
tion of the Ti2p and O1s spectra (Table 3) obtained for

non-implanted titanium we can see that the dominant
component of the surface layer is titanium oxide
TiO. The other components found in the layer were

TiO and metallic titanium. The occurrence of the sig-

nal due to metallic titanium suggests that the TiO layer

was thin.

3.4.1.2. Calcium-ion-implanted titanium.

The XPS re-

sults obtained for calcium-ion-implanted titanium bear
an error due to the high contamination of the surface
with carbon. During the implantation, the sample surfa-
ces undergo oxidation, which is evidenced by the alter-
ation of the surface colour. After the implantation, the
samples were light yellow. The deconvolution of the
Ti2p spectrum indicates that titanium oxides TiO

and TiO or non-stoichiometric TiO oxide are present

in the surface layer. The lack of the Ti

signal suggests

that, during the implantation, the oxide layer has been
thickened. The value of the binding energy determined
for Ca2p (347.8eV) provides evidence that the im-

planted calcium occurs in the form of CaO or Ca(OH).

The surface layer formed during the calcium-ion im-
plantation appears to be composed of TiO, TiO and

CaO or Ca(OH). The absence of the Ca signal can be

attributed to the speci"city of the XPS method * the
composition of the surface layer is only examined to
a depth of about 3 nm.

3.4.2. XPS analyses after immersion

3.4.2.1. Non-implanted titanium. In addition to Ti and O,
other ions such as phosphorus and calcium were present
on the Ti surface after its immersion in an SBF. A decon-
volution of the Ti2p spectrum has showed that the

dominant component was Ti

>, whereas a deconvolution

of the Ca2p and P2p spectra revealed the presence of

calcium in the Ca

> form and phosphorus in the form of

phosphates. The [Ca]/[P] ratio was 0.57, a value close to
the [Ca]/[P] ratio in Ca(HPO) )HO.
3.4.2.2. Calcium-ion-implanted titanium.

The surface

layer contained Ti, Ca, P and O. A deconvolution of the
Ti2p spectrum showed the presence of titanium at

the three oxidation degrees: Ti

>, Ti> and Ti>. The

content of Ti

> increased and the content of Ti> de-

creased, compared to that found in the non-implanted
samples. These changes in the shares of the individual Ti
fractions indicate that during the long-term exposures,
the oxide layer was rebuilt. The binding energy of cal-
cium was 348.2 eV, which indicates that calcium present
in the surface layers occurs as a two-valence ion, whereas

2144

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

Fig. 5. Anodic polarization curves for titanium after various time of exposition.

Table 4
Results of electrochemical experiments

Immersion
time (h)

Ti

Ti#Ca

E (mV) R (M cm) EA (mV) R (M cm)

13

!

112

6.3

!

295

17.7

1000

!

6

10.8

!

38

21

1660

#

128

17.2

2000

!

20

50.5

3200

#

65

23.1

!

82

57.5

the binding energy of phosphorus, equal to 133.9 eV,
indicates to the presence of phosphorus in the form of the
HPO

\

[2]. The [Ca]/[P] ratio was 0.55, which is sim-

ilar to the [Ca]/[P] ratio in Ca(HPO) ) HO.
3.4.3. XPS analyses after anodic polarization
3.4.3.1. Non-implanted

titanium. The

surface

layers

produced on the samples exposed for 13 h and polarized
anodically contained titanium and oxygen and in
addition calcium and phosphorus. After the polarization,
the concentrations of calcium and phosphorus in the
surface layers were comparable with their concentrations
determined after a 3000 h exposure. The Ti2p

spectrum was deconvoluted into two components Ti

>

and Ti

>. The dominant component of this spectrum

was Ti

>. The Ca2p peak at 348.4eV and the

P2p peak at 134.2eV can be attributed to the forma-

tion of hydrated Ca(HPO). The [Ca]/[P] ratio was

0.65.

3.4.3.2. Calcium-ion-implanted titanium.

The surface

layers produced on the samples exposed for 13 h and
subjected to anodic polarization contained Ti, O, Ca and
P. The Ti2p spectrum shows the presence of titanium

but only at the 4# oxidation degree. The values of the
binding energies of Ca2p (347.1eV) and P2

(133.2 eV) indicate that hydroxyapatite has formed. The
[Ca]/[P] ratio was 1.52, a value close to that known for
hydroxyapatite (1.67).

Judging from the results obtained for the samples

polarized anodically, during the polarization of titanium

in an SBF, we deal not only with an oxidation process
but also with the Ca and P ions building themselves into
the surface layer.

3.5. Corrosion results

3.5.1. DC current measurements

The results of electrochemical examinations of non-

implanted and calcium-ion- implanted titanium are given
in Table 4 and in Figs. 5}8.

Table 4 gives the values of the corrosion potential

E and polarization resistance R as a function of the

exposure time in an SBF.

3.5.1.1. Non-implanted titanium.

Results obtained for

non-implanted titanium indicate that the longer the ex-
posure time, the higher the corrosion resistance of tita-
nium. This is evidenced by the increase of the corrosion

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

2145

Fig. 6. Anodic polarization curves for calcium-ion-implanted titanium after various time of exposition.

Fig. 7. Scanning electron micrographs of the surface of calcium-ion-implanted titanium after a long time exposure (2000 h) and after polarization:
(a) surface after polarization; (b) surface after polarization and cleaning (etched in concentrated HNO).

potential E and the polarization resistance R with

increasing exposure time. The advantageous e!ect of the
prolonged exposure time upon the shape of the anodic
polarization curves can be seen in Fig. 5. The polariza-
tion curves plotted for samples exposed for 1000}6000 h,
within the potential range from !200 to 600 mV, show
a decrease of the anodic current density compared with
that measured in the samples exposed for 13 h. The
longer the exposure-time, the greater the reduction of the
current density. Within the potential range from 1000 to
2500 mV, the anodic current density also decreases with
respect to that in the 13 h exposed samples, but this
decrease is insigni"cant * the di!erence is a few micro-
amperes. Above a potential of 2.5 V, the current density

gradually increases in all the samples, its value being
higher in longer exposed samples.

3.5.1.2. Calcium-ion-implanted titanium.

The values of

the polarization resistance R determined for calcium-

ion-implanted titanium suggest that the implantation
improves the corrosion resistance of titanium. After com-
parable exposure times, the polarization resistance
R was two}threefold higher in calcium-ion-implanted

samples than in non-implanted samples. The calcium-ion
implantation, on the other hand, decreases the corrosion
potential E. As the exposure time increases, the cor-

rosion potential increases, but the E values still remain

lower than those measured in non-implanted titanium

2146

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

Fig. 8. Anodic polarization curve for calcium-ion-implanted titanium obtained with reverse scan.

(Table 4). Based on the polarization curves taken after
various exposure times it is impossible to determine
unequivocally the in#uence of calcium-ion implantation
on the corrosion resistance of titanium. By comparing
the anodic polarization curves obtained for non-im-
planted samples (Fig. 5), implanted samples (Fig. 6) after
exposure for 13 h, we "nd that in the calcium-ion-im-
planted samples the anodic current densities are lower
than in non-implanted samples. The decrease of the cur-
rent density is observed within the entire range of the
potentials examined. However at a potential of about
2800 mV, the current density in the implanted samples
increases in a step manner, and above this potential it

#uctuates.

From Fig. 6 we can see that the anodic

polarization curves change adversely at prolonged expo-
sure times. In the samples exposed for 1000}3000 h,
the anodic current density only decreases at potentials
between!400 and#300 mV. Above 300 mV, it gradual-
ly increases up to the values obtained for non-implanted
titanium. At potentials higher than 2700 mV, we observe
violent #uctuations of the anodic current from 1 to
above 1000

A. Moreover, during the anodic polariza-

tion, we could see a white precipitate in addition to
the calcium phosphates formed during the long-term
exposures. An image of the sample surfaces after the
anodic polarization is shown in Fig. 7a. After the precipi-
tates were removed, corrosion pits could be seen in
a microscopic image (Fig. 7b). The observed #uctuations
of the anodic current can be attributed to the formation
of these pits. For the samples exposed for 3200 h, the

breakdown potential should be taken to be equal to
2700 mV.

Fig. 8 shows an example of the anodic polarization

curve taken with a return polarization for a sample
exposed for 3200 h. During the return polarization, the
values of the anodic current density were lower, even
though corrosion pits are present on the sample surface.

3.5.2. AC current measurements

The results of the impedance examinations are given in

Table 5 and in Fig. 9. The Bode diagrams shown in Fig. 9
can be interpreted by using a simple equivalent circuit
with a single time constant, composed of a resistor and
a constant phase element (CPE). The equivalent circuits
used for interpreting the impedance spectra are shown in
Fig. 10. However, the values of the resistance of the
barrier layer (R) calculated from the model shown in

Fig. 10a were widely spread, and thus we used a simpli"-
ed equivalent model, shown in Fig. 10b. The results of the
calculations based on this simpli"ed model are given in
Table 5.

The impedance spectra of Fig. 9 indicate that the oxide

layer formed on titanium, irrespective of whether im-
planted or non-implanted, has a capacitive character.
Within a wide frequency range, the phase angle is close to

!

903. This is in agreement with the results reported in

Refs. [18,19]. The values of the exponent

CPE, which

#uctuate between 0.93 and 0.97, indicate that the layer

formed on the titanium surface is compact and has the
character of a barrier layer. A long-term exposure in an

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

2147

Table 5
Impedance parameters of surface layers on titanium and calcium-ion-implanted titanium (equivalent circuit Fig. 10b)

Immersion time (h)

Ti

Ti#Ca

R ( cm)

Y!.# (F/cm)

R ( cm)

Y!.# (F/cm)

13

20.5

6.4

0.96

14.2

10.9

0.93

1000

20.4

4.1

0.94

10.7

8.3

0.924

1650

19.5

5.7

0.97

2000

12.0

6.9

0.93

3200

14.2

9.4

0.94

14.7

6.7

0.93

4200

20.4

9.0

0.93

Fig. 9. Bode plots for titanium (a) and for calcium-ion-implanted titanium (b). (1) Exposure for 13 h; (2) exposure for 3200 h.

Fig. 10. Equivalent circuits. (a) Equivalent circuit for one time con-
stant; (b) simpli"ed version of equivalent circuit (a). Explanation of
symbols: R solution resistance; R resistance of barrier layer; CPE

* constant-phase element.

SBF has no signi"cant e!ect on the shape of the impe-
dance spectra.

From the results given in Table 5 we can infer that,

with non-implanted titanium, an increase of the exposure
time slightly increases the capacitance of the barrier layer
(Y!.#), which can be explained in terms of the surface

layer structure being changed [20]. In the calcium-ion-
implanted titanium, we observe a reverse e!ect, namely,

the capacitance of the barrier layer decreases with in-
creasing exposure time. This can perhaps be due to the
calcium ions passing from the oxide layer to the SBF
[21,22].

3.6. Biocompatibility examinations

The viability of the cells was found to be una!ected by

the substrate. The results are shown in Fig. 11, where we
can see that the viability of the osteoblasts cultured in
direct contact with non-implanted titanium and cal-
cium-ion-implanted titanium is the same and is at a level
comparable to the control (viability of the cells cultured
on the bottom of the culture dish).

The ALP activity in the culture in contact with tita-

nium surface appeared to be at the same level as that in
the control culture (Fig. 12).

With calcium-ion-implanted titanium, the cells ex-

pressed excellent spreading (Fig. 13) as can be seen from
SEM results.

2148

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

Fig. 11. Results of the XTT assay.

Fig. 12. Results of the ALP activity.

Fig. 13. Scanning electron micrograph of the cells cultured on the
surface of calcium-ion-implanted titanium after four-day exposure.

4. Discussion

The surfaces of the samples, implanted as well as non-

implanted, subjected to long-term exposures appeared to
be covered with calcium phosphates (Fig. 3). According
to SIMS results, the concentrations of calcium, phos-

phorus and oxygen are higher on calcium-implanted
samples than on non-implanted samples exposed for the
same time (Fig. 4c and d); the calcium concentration is
higher than that of phosphorus which suggests that the
Ca/P ratio exceeds unity. The XPS examinations
(Table 2) however give the Ca/P ratio equal to about 0.5,
which is close to the Ca/P ratio in Ca(HPO) ) HO.

From a deconvolution of the P2p spectrum (Table 3),

on the other hand, we infer that the binding energy of
phosphorus is 133.9 eV, a value closest to that of
CaHPO ) 2HO [2]. It is however di$cult to identify

the type of the phosphate based on the value of the
measured binding energy since di!erent sources give dif-
fering values. That the CaHPO )2HO phosphate is

formed can be supposed from the examinations of the
e!ect of pH upon the stability of various phosphate ions
[23]. Within the pH range from 7.2 to 12.3, the most
stable ion is HPO

\

. Hydroxyapatites form at a pH

above 8 [24].

The discrepancy between the SIMS and XPS results is

probably due to the fact that the calcium phosphate does
not form a homogeneous compact layer. The results
obtained from various areas of the surface may di!er
signi"cantly [25].

Examinations of the corrosion resistance of calcium-

ion-implanted titanium have not given an unequivocal
answer whether the calcium implantation improves the
corrosion resistance of titanium. The increased polariza-
tion resistance R in the implanted samples with respect

to that in non-implanted samples suggests that the e!ect
of the implantation is advantageous.

When analysing the polarization curves, we can see

that, within the potential range examined, the advantage-
ous e!ect of calcium-ion implantation only occurs in the
samples exposed for 13 h. The increase of the corrosion
resistance under stationary conditions and the decreased
anodic current density observed in 13 h exposed samples
can be explained in terms of the oxide layer being
thickened and the structure of the surface layer being
altered (amorphization of the surface during the im-
plantation). That the oxide layer has been thickened we
can infer from the XPS examinations (absence of a signal
from the substrate * Table 3).

As the exposure time increases, the density of the

anodic current (within the range from 600 to 2800 mV)
increases to reach the values obtained for non-implanted
titanium. This can be attributed to the alteration of
the chemical composition of the oxide layer during the
exposure, an e!ect suggested by the results of the impe-
dance examinations (Table 5). Hanawa et al. [21,22]
report that, during long-term exposures in the solutions
with a pH of 7, the incubation period is followed
by the migration of the calcium ions from the implanted
layer to the solution. In simpli"cation, the changes
that occur in the surface layer can be written as:
CaTiOPTiO [21].

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

2149

As a result of the calcium migration into the solution,

the oxide layer becomes thinner and, after a su$ciently
long time, it acquires the electrochemical properties close
to those of non-implanted titanium.

Polarization of calcium implanted samples above a po-

tential of 2.8 V a!ects adversely their corrosion resist-
ance. The polarization curves (Fig. 6) show #uctuations
of the anodic current above a potential of 2.8}3 V de-
pending on the exposure time. The current #uctuations
observed in long-term exposed samples can be attributed
to the corrosion pits that form on the sample surfaces
(Fig. 7). Polarization above 2.8 V probably enhances the
migration of calcium ions into the solution. A SIMS
analysis of the polarized and non-polarized samples has
shown that, after the polarization, the calcium concentra-
tion in the surface layer decreases [26].

The mechanism of the initiation of the corrosion pits

on the surface of calcium-ion-implanted titanium is not
easy to explain. We can suppose that the reason is the
alteration of the chemical composition of the oxide layer
(due to the migration of calcium ions to the solution,
which results in the passive layer being locally
weakened). The course of the polarization curves ob-
tained in the present study is non-typical of the existence
of pits. A typical course of the potential}current curve
above the breakdown potential should show a continu-
ous increase of the anodic current due to the develop-
ment of existing pits and the formation of new pits. The
curves obtained in our experiments show #uctuations of
the anodic current, which suggest that the development
of pits are stopped after the initial period of their violent
growth. This could be due to the dissolution of the
implanted layer, followed by repassivation of the tita-
nium surface, or to the pit cavities being blocked by
insoluble corrosion products (titanium oxides and tita-
nium and calcium phosphates), Fig. 7a.

The return polarization curve, too, is not typical of

pitting corrosion (Fig. 8). The repassivation potential
cannot be determined since, during the return polariza-
tion, the values of the anodic current density are lower
than those observed when the potential is being in-
creased. The return polarization curve also shows anodic
current #uctuations within the 4}3 V potential range,
which can re#ect the formation of new corrosion pits.

The pitting corrosion of calcium-ion-implanted tita-

nium was also observed by Spector et al. [27]. They
found that titanium implanted with calcium ions (at
a dose of 2.5

;10Ca>/cm, with a beam energy of

120 keV) underwent pitting corrosion under stationary
conditions during an exposure in a 0.9% NaCl solution
bu!ered with phosphates.

The occurrence of pitting corrosion was also observed

on the surface of the Ti6Al14V titanium alloy implanted
with silicon ions [28]. The breakdown potential fell with-
in the range from 2.5 to 3.5 V depending on the silicon
dose.

The results of our examinations of the corrosion resist-

ance di!er from those reported by Hanawa et al. [13],
even though the conditions of the calcium-ion implanta-
tion were similar (a dose of 1

;10, energy of 18 keV).

The discrepancies may results from the di!erent condi-
tions under which the corrosion resistance was measured.
Hanawa et al. polarized the samples within the potential
range from 0 to 2 V and found that the anodic current
was decreased compared with that in non-implanted
samples. The increased corrosion resistance of calcium-
ion-implanted titanium reported by them suggests that
they examined samples after short-term exposures.

5. Conclusions

The results obtained permit us to conclude that
Calcium-ion implantation with a dose of 1

;10

Ca

>/cm results in an amorphization of the surface layer,

Calcium-ion implantation with a dose of 1

;10

Ca

>/cm increases the corrosion resistance under sta-

tionary conditions,

Titanium subjected to calcium-ion implantation with

a dose of 1

;10 Ca>/cm undergoes pitting corrosion

during anodic polarization. The break-down potential in
an SBF at a temperature of 373C is 2.7}3 V depending on
the exposure time. At longer exposure times, the suscepti-
bility of calcium-ion-implanted titanium to pitting cor-
rosion under anodic polarization conditions increases.

Calcium phosphate formed on the titanium surface dur-

ing the exposure does not a!ect the corrosion resistance.

The results obtained in the present experiments con-

"rm the biocompatibility of titanium and calcium-ion-

implanted titanium tested in contact with osteoblasts
in vitro.

Calcium-ion implantation increases the corrosion res-

istance of titanium under stationary conditions but, dur-
ing its polarization, calcium-ion-implanted titanium
undergoes pitting corrosion at high potentials. This e!ect
should be taken into account when considering the ap-
plication of this material to medical purposes.

Acknowledgements

The authors acknowledge the support of the State

Committee for Scienti"c Research through the Grant
No. 7T08C 024 13.

References

[1] Sundgren J-E, BodoK P, LundstroKm I. Auger electron spectro-

scopic studies of the interface between human tissue and implants
of titanium and stainless steel. J Colloid Interface Sci 1986;
110:9}20.

2150

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

[2] Hanawa T, Ota M. Calcium phosphate naturally formed on

titanium in electrolyte solution. Biomaterials 1991;12:767}74.

[3] Hanawa T, Ota M. Characterization of surface "lm formed

on titanium in electrolyte using XPS. Appl Surf Sci 1992;55:
269}76.

[4] Sioshansi P. Medical application of ion beam processes. Nucl

Instr and Meth 1987;B19/20:204}8.

[5] Buchanan RA, Rigney Jr ED, Williams JM. Ion implantation of

surgical Ti6Al4 V for improved resistance to wear-accelerated
corrosion. J Biomed Mater Res 1987;21:355}66.

[6] Williams JM, Buchanan RA, Rigney Jr ED. Improvement in wear

performance of surgical Ti}6Al}4V alloy by ion implantation of
carbon or nitrogen. Proceedings of the ASM Conference on
Applications of Ion Plating and Ion Implantation to Materials
Atlanta, GA, June 3}5 1985. p. 141}50.

[7] Krupa D, Jezierska E, Baszkiewicz J, KaminHski M, WierzchonH T,

Barcz A. E!ect of nitrogen ion implantation on the structure and
corrosion resistance of OT-4-0 titanium alloy. Surf Coat Technol
1996;79:240}5.

[8] Krupa, D, Baszkiewicz J, Jezierska, et al. E!ect of nitrogen ion

implantation on the corrosion resistance of OT-4-0 titanium alloy
in 0.9% NaCl environment. Surf Coat Technol 1999;111:86}91.

[9] Krupa D, Jezierska E, Baszkiewicz J, et al. E!ect of carbon ion

implantation on the structure and corrosion resistance of OT-4-0
titanium alloy. Surf Coat Technol 1999;114:250}9.

[10] Hanawa T, Asaoka K, Ukai H, Murakami K. Calcium phosphate

precipitation on titanium in electrolyte and its acceleration by
calcium-ion-implantation. Proceedings of the Symposium on
Compatibility of Biomedical Implants San Francisco 1994.
p. 126}37.

[11] Hanawa T, Ukai H, Murakami K. X-ray photoelectron spectro-

scopy of calcium-ion-implanted titanium. J Electron Spectrosc
Relat Phenom 1993;63:347}54.

[12] Hanawa T, Ukai H, Murakami K, Asaoka K. Structure of sur-

face-modi"ed layers of calcium-ion-implanted Ti}6Al}4V and
Ti}56Ni. Mater Trans JIM 1995;36:438}44.

[13] Hanawa T, Murakami K, Kihara S. Calcium phosphate precipi-

tation on calcium-ion-implanted titanium in electrolyte. Charac-
terization and Performance of Calcium Phosphate Coatings for
Implants ASTM STP 1196 Philadelphia, PA April 1994.
p. 170}84.

[14] Hanawa T, Kamiura Y, Kohgo T, et al. Early bone formation

around calcium-ion-implanted titanium inserted into rat tribia.
J Biomed Mater Res 1997;36:131}6.

[15] Boukamp BA. Equivalent circuit * EQUIVCRT Manual, Tech-

nical University of Twente, 1989.

[16] Ziegler JG, Biersack JP. The stopping and range of ions in solids.

Software Trim-92. Pergamon Press, New York, 1985.

[17] NIST X-ray photoelectron spectroscopy database version 2.0.

Standard Reference Data Program Gaithersburg, MD 20899.

[18] KroHlikowski A, Dowgird A. Oxide "lms on titanium * formation

and properties corrosion resistant alloys * 3rd International
Symposium Cracow 20}22 V, 1996. p. 61}9.

[19] Pan J, Thierry D, Leygraf C. Electrochemical impedance spectro-

scopy study of the passive oxide "lm on titanium for implant
application. Electrochim Acta 1996;41:1143}53.

[20] E!ah EAB, Bianco PD, Ducheyne P. Crystal structure of the

oxide layer on titanium and its changes arising from immersion.
J Biomed Mater Res 1995;29:73}80.

[21] Hanawa T, Asami K, Asaoka K. Microdissolution of calcium ions

from calcium-ion-implanted titanium. Corr Sci 1996;38:1579}94.

[22] Hanawa T, Asami K, Asaoka K. AES studies on the dissolution of

surface oxide from calcium-ion-implanted titanium in nitric acid
and bu!er solutions. Corr Sci 1996;38:2061}7.

[23] De Aza PN, Gutian F, Merlos A, Lora-Tamayo E, De Aza S.

Bioceramics * simulated body #uid interfaces: pH and its in#u-
ence of hydroxyapatite formation. J Mater Sci: Mat Med
1996;7:399}402.

[24] BakoH Z, Kotsis I. Composition of precipitated calcium phosphate

ceramics. J Ceram Int 1992;18:373-8.

[25] Serro AP, Fernandez AC, Saramago B, Lima J, Barbosa MA.

Apatite deposition on titanium surfaces * the role of albumin
adsorption. Biomaterials 1997;18:963}8.

[26] Krupa D, Baszkiewicz J. Unpublished data.
[27] Spector M, Vilars PA, Hsu, et al. Strength of attachment of bone

to hydroxyapatite coated and calcium ion-implanted titanium in
a canine model. 20th Annual Meeting of the Society for Bio-
materials, Boston, 1994. p. 330.

[28] Baszkiewicz J, KaminHski K, Kozubowski J, et al. Implantation of

silicon ions into a surface layer of the Ti6Al4V titanium alloy and
its e!ect upon the corrosion resistance and structure of this layer.
J Mat Sci 2000;35:767}75.

D. Krupa et al. / Biomaterials 22 (2001) 2139}2151

2151