Acta Biomaterialia 2 (2006) 693–700

www.actamat-journals.com

Influence of alloying elements on the corrosion stability

of CoCrMo implant alloy in Hank’s solution
a,*
Mirjana Metikoš-Huković , Zora Pilić b, Ranko Babić a, Dario Omanović c

a
Department of Electrochemistry, Faculty of Chemical Engineering and Technology, University of Zagreb, Savska 16,
P.O. Box 177, 10000 Zagreb, Croatia
b
Faculty of Education, University of Mostar, Matice Hrvatske bb, Mostar, Bosnia and Herzegovina
c
Center for Marine and Environmental Research, Rudjer Bošković Institute, P.O. Box 180, 10002 Zagreb, Croatia

Received 15 February 2006; received in revised form 26 May 2006; accepted 6 June 2006

Abstract

The behavior of a CoCrMo alloy and its components was studied in simulated physiological solution (Hank’s solution) using cyclic
voltammetry (CV) and electrochemical impedance spectroscopy (EIS). The passivation of all samples occurred spontaneously at the open
circuit potential. The composition of the oxide films as a function of the solution components and the applied potential is discussed. The
electrochemical properties of the spontaneously passivated electrodes at the open circuit potential were studied by EIS. The polarization
resistance (Rp) and the electrode capacitance (Cdl) were determined. The higher Rp of the alloy than of the chromium pointed to the
stabilizing effect of the other alloy components. The concentration of the metallic ions in a simulated physiological solution, measured
by inductively coupled plasma-mass spectrometry, was in accordance with the values of both the Rp determined from EIS data and cur-
rent densities measured with CV.
 2006 Published by Elsevier Ltd. on behalf of Acta Materialia Inc.

Keywords: Cobalt chromium molybdenum alloy; Implant; Physiological solution; Passivity; Metal ion release

1. Introduction pins and plates [3]. Their biocompatibility is closely linked

Metallic materials play an essential role in assisting with
the repair or replacement of bone tissue that has become dis-
eased or damaged. Metals are more suitable for load-bearing
applications compared with ceramics or polymeric materials
because they combine high mechanical strength and fracture
toughness [1]. However, the main limitation of these metallic
materials is the release of the toxic metallic ions that can lead
to various adverse tissue reactions and/or hypersensitivity
reactions [2]. The currently approved and commonly used
metallic biomaterials include Co-based alloys, titanium
and stainless steels. Each material has its own advantage.
One of the main applications of CoCrMo alloys is in dental
skeletal structures and orthopedic implants such as screws,
*
Corresponding author. Tel.: +385 1 4597 140; fax: +385 1 4597 139.
E-mail address: mmetik@fkit.hr (M. Metikoš-Huković).
to their high resistance towards corrosion which is attributed
to the spontaneous formation of a passive oxide film. The
integrity of the oxide film has been strongly correlated with
the chemical and mechanical stability of implants. In spite
of the chemical stability of the alloys, the release of metallic
ions into the body tissue is well established [4–6]. It occurs by
various mechanisms, including: (a) uniform passive dissolu-
tion resulting from the slow diffusion of metal ions through
the passive film; and (b) the local breakdown of the passivity
as a consequence of the various types of localized corrosion,
or various mechanical events such a fretting corrosion. Since
the oxide layer on the CoCrMo alloys determines the prop-
erties of the implant surface, its characteristics are very
important. Yet, electrochemically oriented studies on the
passivity of CoCrMo alloys are scarce [3,7–11].
Herein we aimed to study the corrosion stability of the
CoCrMo alloy and the properties of the surface oxide layer

1742-7061/$ - see front matter  2006 Published by Elsevier Ltd. on behalf of Acta Materialia Inc.
doi:10.1016/j.actbio.2006.06.002
694 M. Metikoš-Huković et al. / Acta Biomaterialia 2 (2006) 693–700
in a simulated physiological solution (Hank’s solution). As
controls, the individual components, i.e. cobalt, chromium
and molybdenum were also studied. The corrosion stability
was estimated by determining both the polarization resis-
tance and the concentration of the dissolved metallic ions
in the solution. The study was performed using cyclic vol-
tammetry (CV) and electrochemical impedance spectros-
copy (EIS). During the last three decades, EIS has
proved to be advantageous for the characterization of the
various oxide films on the metal surfaces. This technique
requires minimal invasive procedures; i.e. neither the oxi-
dation nor the reduction was forced to take place in the
open circuit mode. From the EIS data we can obtain infor-
mation about the passive film dissolution (the rate and type
of the control) and the passive film characteristics (film
thickness, passive film dielectric constant, diffusion coeffi-
cient of the diffusing species, etc.).
2. Experimental
The study was performed with spectroscopically pure
cobalt, chromium and molybdenum and CoCrMo alloy
(Good Fellow). The alloy was composed of 30 wt.% Cr,
6 wt.% Mo, and the remainder Co. The samples of Cr
and alloy were cylindrical. The lateral surfaces of the cylin-
ders were coated with polyester, exposing only the plane of
the cylinder bases to the solution. The surface areas of the
Cr and alloy samples were 0.385 cm2 and 0.196 cm2,
respectively. The rectangular shaped samples of Co and
Mo were embedded in the Teflon holders; their surface area
was 1 cm2. The electrodes were polished with SiC paper
(600 grit) followed with diamond paste (30, 15, 6 and
1.5 lm), degreased in ethanol in an ultrasonic bath and
rinsed with distilled water. Prior to each measurement,
electrodes were polarized at 1.50 V vs. a saturated calo-
mel electrode (SCE) to remove oxides that were air-formed
at the surface.
Measurements were performed in Hank’s solution, pH
6.8. Its composition was 8.00 g/L NaCl, 0.40 g/L KCl,
0.35 g/L NaHCO3, 0.25 g/L NaH2PO4 · H2O, 0.06 g/L
Na2HPO4 · 2 H2O, 0.19 g/L CaCl2 · 2H2O, 0.19 g/L
MgCl2, 0.06 g/L MgSO4 · 7H2O and 1.00 g/L glucose.
All chemicals were of p.a. purity.
The experiments were performed in a two-compartmen-
tal glass cell at 298 K. The counter electrode was a large Pt-
plate, and the reference electrode was a SCE. All potential
values are reported vs. SCE. Measurements were carried
out using an Solatron SI 1287 Electrochemical Interface
and a Solatron SI 1255 HF Frequency Response Analyser.
The concentrations of the metallic ions released into solu-
tion were determined in ppb (lg/L) by high resolution induc-
tively coupled plasma-mass spectrometry (HR ICP-MS,
Element 2, Thermo Finnigan, Germany). The instrument
settings were as follows. Resolutions (m/Dm): LR = 300,
MR = 4000, HR = 10,000; Torch: Fasel typemm i.d.;
Nebulizer: Micro Mist, AR40-1-F02, 0.2 mL/min; Spray
Chamber: Jacketed Twister, 20 mL, quartz; Cones: Sample
and skimmer – nickel, 1.1 and 0.8 mm; Sample uptake:
ASX 510, CETAC autosampler; Internal standard: In
(1 lg/L). The isotopic mass numbers used were Co 59, Cr
52, Mo 95 or 98.
All specimens with the known surface area (those of Co,
Mo and CoCrMo alloy were about 4 cm2, while Cr samples
were about 10 cm2) were surface-finished with waterproof
emery paper of 400 and 1000 grit and then ultrasonically
cleaned and placed in the high density polyethylene (HDPE)
bottles containing 0.5 L Hank’s solution both at pH = 6.8
and pH = 2. The HDPE bottles used for the immersion tests
were carefully cleaned with 10 vol.% concentrated HNO3
solution and thoroughly rinsed with ultra pure deionised
water to remove impurities. All bottles were placed at room
temperature 23 ± 2 C. Every day the bottles were agitated
in order to homogenize the solution. The concentrations of
released metal ions in solutions were determined after 30
days of specimens’ immersion. The calibration curves were
created with six points obtained by increasing concentra-
tions of standard. A test solution without a metal specimen
was used to estimate the blank. From each separate solution,
the aliquot of 5 mL was transferred to the sample bottle and
acidified with concentrated HNO3 (50 lL). Prior to ICP-MS
analysis the indium standard was added to the sample. The
quantity of metal released (lg/cm2) was calculated using
the following formula: (0.5 L)/[(metal concentration) 
(blank)]/(surface area of the specimen).
3. Results and discussion
3.1. Cyclic voltammetry
Both CoCrMo alloy and its individual components (Co,
Cr and Mo) were investigated in Hank’s solution (pH 6.8)
using cyclic voltammetry. The Co cyclic voltammogram
recorded from 1.5 to 0.0 V is shown in Fig. 1. During
the anodic sweep, the oxidation of the metal surface starts
at around 0.9 V. The current increases up to 0.7 V, then
slightly decreases up to 0.2 V showing a plateau after
which a slight increase is observed. At E > 0.0 V the cobalt
electrode undergoes pitting corrosion. Within the cathodic
0.2 Co
A
0.0
j / mA cm-2
-0.2
-0.4
C
-0.6
-1.2 -0.9 -0.6 -0.3 0.0
E /V
Fig. 1. Cyclic voltammogram of Co in Hank’s solution, m = 20 mV s1.
 M. Metikoš-Huković et al. / Acta Biomaterialia 2 (2006) 693–700
portion of the voltammogram, a well expressed cathodic
peak is observed at 0.95 V followed by the current
increase due to the hydrogen evolution. It has been
reported that the Co anodic polarization curve shows the
primary region of passivation in neutral and slightly alka-
line solutions within the examined potential range (up to
0.0 V) [12]. The oxide layer is presumably composed of a
hydrated cobalt(II) oxide, Co(OH)2 and/or anhydrous
CoO. The higher-valence oxide begins to form at E >
0.114 V vs. SCE. The reversible potential corresponding
to the reaction
þ 
Co + 2H2 O ! Co(OH)2 + 2H + 2e ð1Þ
is E/V = 0.095  0.0591 pH vs. a standard hydrogen elec-
trode [13]. The cathodic reduction of Co(OH)2 or CoO
leads to the metallic cobalt only, without any detectable
amount of Co2+ ions [12]. It is well known that the electro-
chemical behavior of cobalt is very sensitive to the solution
composition. Burke et al. [14] investigated the cobalt under
potential cycling conditions in neutral and alkaline solu-
tions and observed that the activity of the cobalt oxide film
decreases in the presence of NaH2PO4. They attributed
such behavior to the formation of insoluble cobalt phos-
phate, or at least to the incorporation of the phosphate
ions in the oxide layer. On the other hand, Davies and Bur-
stein [15] and Gervasi et al. [16] showed that the bicarbon-
ate ions (also present in Hank’s solution) are aggressive
towards Co(II) oxide layer in slightly alkaline solutions.
According to them, the passive layer in the Co(II) potential
region is composed of both CoCO3 and Co(OH)2. Under
the influence of HCO 3 ions, the dissolution of CoCO3
arises from the formation of CoðCO3 Þ2 2 in the upper range
of the Co(II) potential region. The aggressive effect of
bicarbonate ions is inhibited by the presence of borate ions
[15]. According to Burke et al. [14], it seems that the phos-
phate ions have an inhibiting effect on the corrosion pro-
cess of the cobalt in the Co(II) potential region; thus the
reactivation, which is the characteristic of the pure bicar-
bonate solution [15,16], is not observed with Hank’s
solution.
The electrochemical behavior of the chromium in
Hank’s solution was characterized by the cyclic voltammo-
gram presented in Fig. 2. There are three distinct regions in
the potential range examined: the hydrogen evolution
region, the passive region and the transpassive region
followed by the oxygen evolution. The switch from the
cathodic to the anodic current took place without an
activation–passivation transition as in acid solutions [17].
It is well known that even a brief contact of the chromium
with damp air is sufficient to create a thin oxide layer on its
surface [18]. A relatively large current generated during
hydrogen evolution is caused by the electron tunneling
through the thin film depleted semiconducting oxide
[19,20]. Within the passive region, the increase in the cur-
rent with the increasing potential can be interpreted in
terms of the point defect model [21]. Thus, the increasing
anodic polarization favors the production of the metal
695
0.2
Cr
0.1 A
j / mA cm-2
0.0
C
-0.1
-0.2
-1.5 -1.0 -0.5 0.0 0.5 1.0
E /V
Fig. 2. Cyclic voltammogram of Cr in Hank’s solution, m = 20 mV s1.
vacancies near the film–solution interface [22], and an
increase in passive current. The transpassive region of chro-
mium begins at about 0.5 V as indicated by a faster increase
of the current related to the solid-state oxidation of Cr(III)–
to Cr(VI)–oxygen containing species. Depending on the pH
these species are either Cr2 O2 2
7 (pH < 6) or CrO4 (pH > 6)
[23]. The cathodic current within the negatively sweep is
attributed to the combination of the capacitive charging
and the solid-state electroreduction of Cr(VI)– to Cr(III)–
oxygen containing species (the current peak C).
The passivity of the chromium is due to the formation of
Cr(III)–oxygen species at the electrode surface that pro-
ceeds via solid state reactions [24]. Based on the ex situ
spectroscopy studies, the passive film on chromium has
been identified as Cr2O3 [25], or CrOOH [26]. According
to the X-ray photoelectron spectroscopy (XPS) studies of
Moffat and Latanision [27] the passive film formed in sul-
phuric acid is composed of the hydrated trivalent chro-
mium oxide, presumably CrOOH. A bilayer structure of
the passive film on chromium has been proposed in both
acidic [28] and slightly alkaline solutions [29]. A hydrated
Cr(III) oxide, i.e. Cr(OH)3, was suggested to be formed
first. Underneath, based upon the potential and time of
the passivity, an anhydrous Cr2O3 or CrOOH film is
formed. Surface enhanced Raman spectroscopy (SERS)
studies of Oblonsky and Devine [30], describe the passive
film on chromium in borate buffer as mainly Cr(OH)3,
but with evidence of Cr(OH)2 as well. Studies of chromium
passivation in acidic chloride solutions performed by
Bjornquist and Olefjord [17] using electron spectroscopy
for chemical analysis (ESCA) have shown that chloride
ions are present inside the film, i.e. they replace OH ions
in Cr(III)–oxygen containing species. The kinetics of the
film formation indicates that the film grows according to
a high field model. The thickness of the passive film formed
on chromium (0.3–2.5 nm) is a linear function of the ano-
dic potential [25,27,31].
The molybdenum electrochemical behavior in Hank’s
solution was characterized by the cyclic voltammogram
presented in Fig. 3. The well expressed hydrogen evolution
region up to 1.2 V is followed by the passive region
696 M. Metikoš-Huković et al. / Acta Biomaterialia 2 (2006) 693–700
1.0 Mo
0.5
j / mA cm -2
0.0
-0.5 C
-1.0
-1.2 -0.9 -0.6 -0.3 0.0 0.3
E/V
Fig. 3. Cyclic voltammogram of Mo in Hank’s solution, m = 20 mV s1.
which is characterized by a slightly expressed hump with the
current maximum at a potential of 0.7 V. The increase in
current above 0.0 V corresponds to the transpassive region
of molybdenum. The change from the cathodic to the
anodic current took place without an activation–
passivation transition, like chromium. By reversing the
anodic sweep, the voltammogram shows a small cathodic
current peak at 0.5 V which probably corresponds to the
reduction of higher valence molybdenum-oxygenated spe-
cies formed in the transpassive region. In the literature,
there are different opinions on the nature of the passivity
of molybdenum. Investigating molybdenum in sulphuric
acid solution, Brox et al. [32] came to the conclusion that
Mo is passivated by the chemisorption of oxygen. A thick
film composed of a complex multivalent Mo–oxygen com-
pound is formed on the surface in the transpassive region,
when the current starts to increase rapidly. However, the
majority of authors [33–37] have found that the passive
state of Mo in acidic solutions is caused by the formation
of the Mo–oxygen compounds of the variable oxidation
state, depending upon the applied potential. The ESCA
data of Urgen et al. [34], and the XPS data of Lu and Clay-
ton [35] obtained by the investigation of Mo in acidic chlo-
ride solutions indicated that inside the passive film Mo is
present as Mo(III), Mo(IV), Mo(V) and Mo(VI); the frac-
tions of different valence states were noticed at different
anodic potentials. According to Urgen et al. [34], the passive
film contains mostly Mo(III), whereas Mo(VI) is the major
species in the transpassive films. On the other hand, Lu and
Clayton [35] observed Mo(III), probably Mo2O3 in the
early stage of passive film formation. Later though, the
main components of the passive film were MoO2 and
MoO(OH)2. The transpassive product of Mo was a mixture
of MoO2, MoO(OH)2, Mo3O5 and MoO3, while HMoO 4 in the outer part of the layer, whereas Cr-oxide was present
was observed as a dissolution product. In a slightly acidic
solution Wang et al. [33], using SERS, observed that
MoO2 was the main component of the passive film on
molybdenum. The reason why Mo(III) species were not
observed might be the higher pH of the solution; Mo(III)
oxides are stable in the acid solutions only [38]. On the bases
of cyclic voltammetry and the rotating ring-disc results,
3.0
Co30Cr6Mo
2.0 n=1
j / mA cm-2
1.0
n=5
0.0
-1.0
-2.0
-1.5 -1.0 -0.5 0.0 0.5 1.0
E/V
Fig. 4. Cyclic voltammogram of CoCrMo alloy in Hank’s solution,
m = 20 mV s1.
Hull [37] has proposed that the passive films on Mo in acidic
solutions are composed of MoO2 and MoO3.
The cyclic voltammogram of CoCrMo alloy recorded in
Hank’s solution is shown in Fig. 4. Two anodic current
peaks are observed. The first broad current peak with a
maximum at 0.53 V was assigned to the electroforma-
tion of Co(II) oxide/hydroxide at the pre-existing Cr2O3
layer. The anodic current peak at 0.55 V was related to
the secondary passivity, i.e. the solid state oxidation of
Cr(III) present in the passive layer to Cr(VI). At 0.8 V
simultaneous electroformation of soluble CrO2 4 species
and oxygen evolution took place [39]. Within the cathodic
portion of the voltammogram no current peak was
observed even when a more sensitive scale had been applied.
During successive cycles the anodic charge decreased very
rapidly, indicating that the oxide layer formed in the previ-
ous cycle was not reduced completely within the cathodic
portion of the voltammogram and that during the repetitive
cycling a considerable decrease in the active electrode area
available for the oxidation process occurs.
Electrochemically related studies of the CoCrMo alloy
and the composition of passive films are scarce [3,7–11].
Milosev and Strehblow [7], and Kocijan et al. [8] studied
the CoCrMo orthopedic alloy in simulated physiological
solutions (pH 7.5) by XPS. They found that the electro-
chemical oxidation resulted in the formation of a complex
layer whose composition and thickness was dependent
upon the applied potential. At low potentials the passive
film thickness was about 3.1 nm, consisting mostly of
Cr2O3 and Cr(OH)3. At higher potentials both Co and
Mo oxides entered the passive layer; Co as CoO, and Mo
as MoO3. Consequently the film thickness significantly
increased. The amount of Co- and Mo-oxides was higher
mainly in the inner part, closer to the oxide/metal interface.
Hodgson et al. [9] investigated also the composition of the
passive film of CoCrMo alloy in simulated biological solu-
tion (pH 7.4) by XPS. They reported that the passive films
formed at potentials below 0.5 V vs. SCE were highly
enriched (90%) with Cr(III) oxide and Cr(III) hydroxide.
Cobalt was the main constituent of the alloy, but its con-
 M. Metikoš-Huković et al. / Acta Biomaterialia 2 (2006) 693–700 697

centration in the film was around 5%, except at 0.5 V where 90

it reached a maximum of about 20%. The concentration of 105 Cr 80
Mo in the form of Mo(IV) oxide in the passive film was
small, approximately 1%. In the passive region, the major 70

-Phase angle / degrees

species that dissolved from the alloy was Co. In short-term 104 60
experiments, the adsorption/incorporation of the specific
50

IZI / Ω cm2
ions (phosphates, hydrogen carbonates) on the film surface
was not observed. Hanawa et al. [10] investigated CoCrMo 103 40
alloy in Hank’s solution (pH 7.4) using XPS. They found
30
that Co dissolved from the film; the resulting film was com-
posed of chromium oxide with a small amount of molybde- 102 20
num oxide. They also found that calcium phosphate was 10
generated on/or in the film. Li et al. [11] studied the corro-
sion films on CoCrMo alloy surface in artificial saliva and 101 0
10-3 10-2 10-1 100 101 102 103 104 105 106
sodium chloride solution by Raman spectroscopy. From
Frequency / Hz
their results it appears that Cr and Co in the alloy play
important roles in passivation. Mo species exist only in Fig. 6. Impedance spectra of Cr in Hank’s solution measured at the open
the air-formed film, and it readily dissolves upon exposure circuit potential.
to the solution.
5
10 90
3.2. Impedance spectroscopy
Mo 80

A set of impedance spectra recorded with the electrodes 4 70

10
of CoCrMo alloy and individual components (Co, Cr and

-Phase angle / degrees

60
Mo) in Hank’s solution at open circuit potentials, pre-
IZI / Ω cm2

sented as Bode plots, is shown in Figs. 5–8. The spectra 3

50
show that in the higher frequency region, logjZj tends to 10
40
become constant, with the phase angle values falling rap-
idly towards 0 with increasing frequency. This is a 30
2
response typical of the resistive behavior and corresponds 10
20
to the solution resistance, RX. Inside the medium and low
frequency regions, the systems behave differently. In the 10

medium frequency range, a linear relationship between 10

1
0
logjZj and log f is observed for all samples, but with differ- 10-3 10-2 10-1 100 101 102 103 104 105 106
Frequency / Hz
ent slopes (always less than 1) and phase angle maxima
(less than 90), indicating that the passive films were Fig. 7. Impedance spectra of Mo in Hank’s solution measured at the open
not fully capacitive. Inside the low frequency range an circuit potential.

4 6
10 80 10 90
Co30Cr6Mo 80
Co 70
5
10
60 70
-Phase angle / degrees

3
-Phase angle / degrees

10
50 4
60
10
IZI / Ω cm2

IZI / Ω

40 50
3
30 10 40
2
10
20 30
2
10 20
10

1
10 0 10
1
10
0
-3 -2 -1 0 1 2 3 4 5 6
10-3 10-2 10-1 100 101 102 103 104 105 106 10 10 10 10 10 10 10 10 10 10
Frequency / Hz Frequency / Hz

Fig. 5. Impedance spectra of Co in Hank’s solution measured at the open Fig. 8. Impedance spectra of CoCrMo alloy in Hank’s solution measured
circuit potential. at the open circuit potential. Surface area is 0.196 cm2.
698 M. Metikoš-Huković et al. / Acta Biomaterialia 2 (2006) 693–700

inductive behavior is observed for cobalt (Fig. 5). With
chromium (Fig. 6) though, the slope of the logjZj vs. log f
and the phase angle against log f curves indicate War-
burg-like behavior. Warburg-like impedance was also
observed with chromium in a borate buffer, pH 9.3 [40].
The resistive behavior of the Mo (Fig. 7) and alloy
(Fig. 8) electrodes becomes more pronounced, especially
for Mo, in the low frequency range, but the region where
logjZj does not depend on log f (i.e.. the dc limit) is not
reached until 102 Hz.
Since any interface undergoing an electrochemical reac-
tion is typically analogous to an electronic circuit consist-
ing of a specific combination of resistors and capacitors,
the electrochemical systems under the study can be
described in terms of their electrical equivalent circuits
(EECs). Thus, the first task was to identify the best-fit
EEC, based on Boucamp’s software [41], in order to estab-
lish the values of circuit elements. The intention was then
to correlate their values with cyclic voltammetry data and
solubility data experimentally determined. The usual guide-
lines for the selection of the best-fit EEC were followed: (i)
a minimum number of circuit elements were to be
employed; (ii) the v2 error was suitably low (v2 6 104),
and the errors associated with each element were up to
5%. The EECs that satisfied the above criteria were used
to fit the experimental data and are presented in Fig. 9.
Since the measured capacitive response is not generally
ideal due to the certain heterogeneity of the electrode sur-
face [42,43], a constant phase element (CPE) was intro-
duced for fitting the spectra, instead of an ideal
capacitance element. Its impedance may be defined by:
Z(CPE) = [Q(jx)n]1, where Q is the constant, x is the
angular frequency, and n is the CPE power. The factor n
is an adjustable parameter that usually lies between 0.5
a As the n1 and n2 exponents are very high, the constant
RΩ CPE 1
R1 L The impedance data of the alloy (Fig. 8) were fitted
R2
and 1 [44]. When n = 1, the CPE describes an ideal capac-
itor with Q equal to the capacitance (C). For 0.5 < n < 1,
the CPE describes a distribution of dielectric relaxation
times in frequency space, and when n = 0.5 the CPE repre-
sents a Warburg impedance with diffusional character.
The Co impedance data (Fig. 5) were fitted with the
EEC presented in Fig. 9a, and the resultant EIS parameters
are given in Table 1. The value of R1 representing the
charge transfer resistance, Rct, is quite low (2.7 X cm2)
and indicates the low stability of cobalt in Hank’s solution.
The constant phase element Q1 represents the double layer
capacitance, Cdl, as shown by the high value of the n1 expo-
nent. The R2L element is usually attributed to the relaxa-
tion of the corrosion products on the electrode surface.
The cell resistance, RX, equals 10 ± 1 X. The same value
was observed for all of the specimens and was not inserted
in Table 1.
The impedance data of chromium (Fig. 6) were fitted
with the EEC presented in Fig. 9b, and the resultant EIS
parameters are given in Table 1. The polarization resis-
tance, Rp, equals the sum of Rct (R1) and the diffusion resis-
tance, RD [45]. It is quite large (Rp = 350 X cm2) indicating
the high stability of the chromium in Hank’s solution.
Since the n1 exponent is very high, the constant phase ele-
ment Q1 is essentially the Cdl.
The impedance data of molybdenum (Fig. 7) were pos-
sible to model by the electrical equivalent circuit composed
of the cell resistance, RX, in series with two RC parallel
combination, each composed of a resistance, R, and a con-
stant phase element, CPE (Fig. 9c). The physical meaning
given to the circuit is association with the passive film/elec-
trolyte interface (R1CPE1) and the passive film itself
(R2CPE2). The polarization resistance of the molybdenum
(Rp = 41.9 kX cm2) equals the sum of the R1 (Rct) and the
passive film resistance, R2. It is much larger than that of
cobalt, but smaller than that of chromium as expected.
phase elements CPE1 and CPE2 represent the Cdl and the
passive film capacitance, respectively.
with the EEC used to fit the molybdenum data (Fig. 9c),
and the resultant EIS parameters are given in Table 1.

b Table 1
RΩ CPE 1 Impedance parameters of Co, Cr, Mo and CoCrMo alloy in Hank’s
solution
R1 W EIS parameters Co Cr Mo Alloy
Ecorr (V) 0.54 0.31 0.25 0.35
106CPE1 (X1 cm2 sn) 29.3 36.0 16.6 32.7
c n1 0.80 0.84 0.94 0.90
RΩ CPE 1
R1 (kX cm2) 2.7 34.4 33.6 104
106CPE2 (X1 cm2 sn) – – 25.3 16.3
R1 CPE 2 n2 – – 0.95 0.99
R2 (kX cm2) – – 8.3 499
R2 WR (kX cm2) – 314 – –
103L (Hcm2) 1.5 – – –
R2 (kX cm2) 0.65 – – –
Fig. 9. Equivalent electrical circuits used to fit the impedance spectra.
 M. Metikoš-Huković et al. / Acta Biomaterialia 2 (2006) 693–700 699

The polarization resistance of the alloy (Rp = 603 kX cm2) oxide film of Cr. These results are in relatively good agree-
is the largest in comparison with the alloy components ment with those obtained by Okazaki and Gotoh [4]: the
indicating very high stability of alloy in Hank’s solution. quantities of Co and Mo are a little higher, while the
Both constant phase elements, CPE1 and CPE2, are essen- amount of Cr is practically the same. The observed rela-
tially the ideal capacitances, as shown by the high values of tively small differences in solubility could be ascribed to
the n1 and n2 exponents. They represent the Cdl and the the different media used in the immersion tests, the different
passive film capacitance, respectively. The double layer temperatures, as well as to the treatment of metal speci-
capacitances for all of the samples are typical for the Cdl mens before immersion.
of the passive oxide layers [27,46,47].
4. Conclusion
3.3. Cobalt, chromium and molybdenum concentration in
Hank’s solution
The excellent corrosion resistance of the CoCrMo alloy
in Hank’s solution is due to the highly protective passive
The results obtained in static immersion test are listed in
(oxide) film that is the key factor for its compatibility.
Tables 2 and 3. Generally, the dissolution of the metal ions
At the open circuit potential all pure metals and alloy
is dependent on the metal/alloy itself and the pH of the
underwent passivation. During the anodic polarization
solution. Comparing the quantity of metal ions released
Co is inclined to undergo localized corrosion at low poten-
from the individual pure metal specimens at pH = 6.8, a
tials, while Cr and Mo undergo transpassive transition at
very high metal content of cobalt and molybdenum can
more noble potentials.
be noted compared to chromium (Table 2). Almost five
The results show that CoCrMo alloy exhibits ‘‘chromium-
to six orders of magnitude higher concentrations of Co
like’’ passivity. Although the corrosion behavior of the pas-
are obtained. This suggests that the oxide film formed on
sive alloy is determined by the presence of Cr, its corrosion
Co is much more soluble than that of Cr, as follows from
resistance is higher than that of Cr due to the beneficial
the solubility products of their hydrated species. These
effect of Mo on the alloy passivity.
results are in a good agreement with the corresponding
The results of quantitative analysis of the dissolved
polarization resistances obtained by impedance spectros-
metal ions in Hank’s solution showed that the quantities
copy. By lowering the pH, the dissolution of both Co
of Co and Mo released from pure metals are much higher
and Cr is increased by about 80%. The oxides of Mo are
(by a factor 103) than those from CoCrMo alloy, while
more stable in acidic than in neutral or slightly alkaline
the quantity of Cr released from the alloy is slightly higher
media, which is in accordance with the literature data
than those from the pure Cr. By decreasing the pH of solu-
[48]. Comparing these results with those obtained from
tion, the quantities of all metal ions released from the alloy
immersion tests of the CoCrMo alloy, it can be observed
increase. The same trend is observed for pure metals,
that the quantities of released Co and Mo ions are extre-
except Mo, whose stability increases by decreasing the
mely reduced, while the amount of dissolved Cr is slightly
pH of the solution.
increased. Moreover, the higher solubility of all specimens
is obtained at pH = 2. This indicates that overall process of
metal dissolution in Hank’s solution is controlled by the
References

[1] Staiger MP, Pietak AM, Huadmai J, Dias G. Magnesium and its
Table 2
alloys as orthopedic biomaterials. Biomaterials 2005;27:1728–34.
The quantity of metal ions dissolved during 30 days of immersion of pure
[2] Wapner KL. Implications of metallic corrosion in total knee
metal specimens in Hank’s solutions at a specified pH
arthroplasty. Clin Orthop Relat Res 1991;271:12–20.
Metal specimen Quantity of released metal (lg cm2) [3] Contu F, Elsner B, Bohni H. Corrosion behaviour of CoCrMo
pH = 2.0 pH = 6.8 implant during fretting in bovine serum. Corros Sci 2005;47:1863–75.
[4] Okazaki Y, Gotoh E. Comparison of metal release from various
Co 2808.8 ± 39.0 608.0 ± 5.5 metallic biomaterials in vitro. Biomaterials 2005;26:11–21.
Cr 0.030 ± 0.003 0.004 ± 0.001 [5] Hanawa T. Metal ion release from metal implants. Mater Sci Eng C
Mo 18.30 ± 0.30 64.30 ± 0.80 2004;24:745–52.
[6] Lin H, Bumgardner JD. Changes in surface oxide composition of Co–
Cr–Mo implant alloy by macrophage cells and their released reactive
Table 3 chemical species. Biomaterials 2004;25:1233–8.
The quantity of metal ions dissolved during 30 days of immersion of [7] Milosev I, Strehblow HH. The composition of the surface passive film
CoCrMo alloy specimen in Hank’s solutions at a specified pH formed on CoCrMo alloy in simulated physiological solution.
Electrochim Acta 2003;48:2767–74.
Metal specimen Quantity of released metal (lg cm2) [8] Kocijan A, Milosev I, Pihlar B. Cobalt-based alloys for orthopaedic
pH = 2.0 pH = 6.8 applications studied by electrochemical and XPS analysis. J Mater
Sci: Mater Med 2004;15:643–50.
Co 1.343 ± 0.012 0.734 ± 0.013
[9] Hodgson AWE, Kurz S, Virtanen S, Fervel V, Olsson CA, Mischler
Cr 0.085 ± 0.001 0.014 ± 0.001
S. Passive and transpassive behaviour of CoCrMo in simulated
Mo 0.098 ± 0.003 0.040 ± 0.003
biological solutions. Electrochim Acta 2004;49:2167–78.
700 M. Metikoš-Huković et al. / Acta Biomaterialia 2 (2006) 693–700
[10] Hanawa T, Hiromoto S, Asami K. Characterization of the surface
oxide film of a Co–Cr–Mo alloy after being located in
quasi-biological environments using XPS. Appl Surf Sci 2001;183:
68–75.
[11] Li Y, Wang K, He P, Huang BX, Kovacs P. Surface-enhanced
Raman spectroelectrochemical studies of corrosion films on implant
Co–Cr–Mo alloy in biosimulating solutions. J Raman Spectrosc
1999;30:97–103.
[12] Sato N, Ohtsuka T. Anodic oxidation of cobalt in neutral and basic
solution. J Electrochem Soc 1978;125:1735–40.
[13] Pourbaix M. Atlas of electrochemical equilibrium. Oxford: Perg-
amon Press; 1966, p. 324.
[14] Burke LD, Lyons ME, Murphy OJ. Formation of hydrous oxide
films on cobalt under potential cycling conditions. J Electroanal
Chem 1982;132:247–61.
[15] Davies DH, Burstein GT. The electrochemical behaviour of cobalt in
bicarbonate and borate electrolytes. Corros Sci 1980;20:973–87.
[16] Gervasi CA, Biaggio SR, Vilche JR, Arvia AJ. The electrochemical
behaviour of cobalt in carbonate–bicarbonate solutions. Corros Sci
1989;29:427–37.
[17] Bjornkvist L, Olefjord I. The electrochemistry of chromium in acidic
chloride solutions: anodic dissolution and passivation. Corros Sci
1991;32:231–42.
[18] Metikos-Hukovic M, Ceraj-Ceric M. P-type and n-type behavior of
chromium oxide as a function of the applied potential. J Electrochem
Soc 1987;134:2193–7.
[19] Gerischer H. The impact of semiconductors on the concepts of
electrochemistry. Electrochim Acta 1990;35:1677–99.
[20] Schmickler W. Direct elastic electron tunnelling through oxide films. J
Electroanal Chem 1977;84:203–7.
[21] Macdonald DD. The point defect model for the passive state. J
Electrochem Soc 1992;139:3434–49.
[22] Bojinov M, Fabricius G, Laitinen T, Mikela K, Saario T. Coupling
between ionic defect structure and electronic conduction in passive
films on iron, chromium and iron–chromium alloys. Electrochim
Acta 2000;45:2029–48.
[23] Keisall GH, House CI, Guyanga FP. Chemical and electrochemical
equilibria and kinetics in aqueous Cr(III)/Cr(II) chloride solutions. J
Electroanal Chem 1988;244:179–202.
[24] Dobbelaar AL, De Wit JHW. Impedance measurements and analysis
of the corrosion of chromium. II The passivation. J Electrochem Soc
1990;137:2038–46.
[25] Seo M, Saito R, Sato N. Ellipsometry and Auger analysis of
chromium surface passivated in acidic and neutral aqueous solutions.
J Electrochem Soc 1980;127:1909–12.
[26] Sugimoto K, Matsuda S. Passive and transpassive films on Fe–Cr
alloys in acid and neutral solutions. Mater Sci Eng 1980;42:181–9.
[27] Moffat TP, Latanision RM. An electrochemical and X-ray photo-
electron spectroscopy study of the passive state of chromium. J
Electrochem Soc 1992;139:1869–79.
[28] Maurice V, Yang WP, Marcus P. XPS and STM investigation of the
passive film formed on Cr(1 1 0) single-crystal surfaces. J Electrochem
Soc 1994;141:3016–27.
[29] Kim JS, Cho EA, Kwon HS. Photo-electrochemical analysis of
passive film formed on Cr in pH 8.5 buffer solution. Electrochim Acta
2001;47:415–21.
[30] Oblonski LJ, Devine TM. A surface enhanced Raman spectroscopic
study of the passive films formed in borate buffer on iron, nickel,
chromium and stainless steel. Corros Sci 1995;37:17–41.
[31] Haupt S, Strehblow HH. The formation of the passive layer on Cr in
0.5 M H2SO4 . A combined electrochemical and surface analytical
study. J Electroanal Chem 1987;228:365–92.
[32] Brox B, Hua WY, Olefjord I. Electrochemical surface analyses of
Mo(1 0 0) single crystal polarized in 0.5 M H2SO4. J Electrochem Soc
1988;135:2184.
[33] Wang K, Li YS, He P. In situ identification of surface species on
molybdenum in different media. Electrochim Acta 1998;43:2459–67.
[34] Urgen M, Stolz U, Kirchheim R. ESCA measurements of films on
molybdenum formed in the passive and transpassive region. Corros
Sci 1990;30:377–91.
[35] Lu YC, Clayton CR. An XPS study of the passive and transpassive
behavior of molybdenum in deaerated 0.1 M HCl. Corros Sci
1989;29:927–37.
[36] Bojinov M, Betova I, Raicheff R. Transpassivity of molybdenum in
H2SO4 solution. J Electroanal Chem 1995;381:123–31.
[37] Hull MN. On the anodic dissolution of molybdenum in acidic and
alkaline electrolytes. J Electroanal Chem 1972;38:143–57.
[38] Pourbaix M. Atlas of electrochemical equilibrium. Oxford: Perg-
amon Press; 1966, p. 277.
[39] Metikos-Hukovic M, Omanovic S, Babic R, Milosev I. An electro-
chemical study of the passive film formed on Ni–Cr (80/20) coating.
Ber Bunsenges Phys Chem 1994;98:1243–9.
[40] Marijan D, Gojic M. Electrochemical study of the chromium
electrode behaviour in borate buffer solution. J Appl Electrochem
2002;32:1341–6.
[41] Boukamp BA. A package for impedance/admittance data analysis.
Solid State Ionics 1986;18–19:136–40.
[42] Macdonald JR. Impedance spectroscopy, emphasizing solid materials
and systems. New York, NY: Wiley; 1987, p. 39.
[43] Martini EMA, Muller IL. Characterization of the film formed on iron
in borate solution by electrochemical impedance spectroscopy.
Corros Sci 2000;42:443–54.
[44] Rammelt U, Reinhard G. On the applicability of a constant phase
element (CPE) to the estimation of roughness of solid metal
electrodes. Electrochim Acta 1990;35:1045–9.
[45] Scully JR. Polarization resistance method for determination of
instantaneous corrosion rates. Corrosion 2000;56:199–218.
[46] Haupt S, Strehblow HH. Combined electrochemical and surface
analytical investigations of the formation of passive layers. Corros Sci
1989;29:163–82.
[47] Castro EB, Vilche JR. Investigation of passive layers on iron and
iron-chromium alloys by electrochemical impedance spectroscopy.
Electrochim Acta 1993;38:1567–72.
[48] Badawy WA, Al-Kharafi FM. Corrosion and passivation behaviors
of molybdenum in aqueous solutions of different pH. Electrochim
Acta 1998;44:693–702.