This article appeared in a journal published by Elsevier.

The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/authorsrights
 Author's personal copy

Materials Science and Engineering C 34 (2014) 69–77

Contents lists available at ScienceDirect

Materials Science and Engineering C

journal homepage: www.elsevier.com/locate/msec

Evaluation of taper joints with combined fatigue and crevice corrosion

testing: Comparison to human explanted modular prostheses
L. Reclaru a,⁎, R.A. Brooks b, M. Zuberbühler c, P.-Y. Eschler a, F. Constantin a, G. Tomoaia d
a
PX Group S.A., Dep R&D Corrosion and Biocompatibility Group, Bd. des Eplatures 42, CH-2304 La Chaux-de-Fonds, Switzerland
b
Orthopaedic Research, Addenbrooke's Hospital, University of Cambridge, Box 180 Hills Road, CB2 0QQ Cambridge, United Kingdom
c
Smith & Nephew Orthopaedics AG, Schachenalle 29, 5001 Aarau, Switzerland
d
University of Medicine and Pharmacy Iuliu Hateganu of Cluj-Napoca, Dept. of Orthopaedics and Traumatology, Cluj-Napoca, Romania

a r t i c l e i n f o a b s t r a c t

Article history: The requirement for revision surgery of total joint replacements is increasing and modular joint replacement
Received 6 May 2013 implants have been developed to provide adjustable prosthetic revision systems with improved intra-
Received in revised form 27 August 2013 operative flexibility.
Accepted 9 October 2013
An electrochemical study of the corrosion resistance of the interface between the distal and proximal modules of
Available online 17 October 2013
a modular prosthesis was performed in combination with a cyclic fatigue test. The complexity resides in the
Keywords:
existence of interfaces between the distal part, the proximal part, and the dynamometric screw.
Modular prosthesis A new technique for evaluating the resistance to cyclic dynamic corrosion with crevice stimulation was used and
Fatigue corrosion the method is presented.
Explanted prosthesis In addition, two components of the proximal module of explanted Ti6Al4V and Ti6Al7Nb prostheses were
investigated by optical and electron microscopy.
Our results reveal that: The electrolyte penetrates into the interface between the distal and proximal modules
during cyclic dynamic fatigue tests, the distal module undergoes cracking and corrosion was generated at the
interface between the two models; The comparison of the explanted proximal parts with the similar prostheses
evaluated following cyclic dynamic crevice corrosion testing showed that there were significant similarities
indicating that this method is suitable for evaluating materials used in the fabrication of modular prostheses.
© 2013 Elsevier B.V. All rights reserved.

1. Introduction
Revisions of total joint prostheses such as hip and knee replacements
are becoming more common. Often, revision surgery is more complicated
and less predictable than primary interventions. Therefore, orthopaedic
surgeons have requested a fully flexible and adjustable prosthetic revision
system and more of these modular joint replacement systems with intra-
operative flexibility are being developed.
Ti6Al4V and Ti6Al7Nb alloys show excellent corrosion resistance
[1–6], but in the modular prostheses the combination of the mechanical
process of fatigue and crevice corrosion can generate destruction of the
interface between the distal and proximal modules [7].
Crevice corrosion, wear by friction and fatigue have also been
described at the level of the distal–proximal junction on explanted
modular prostheses of this type made from titanium alloys [7,8].
Considering the multifactorial character of crevice corrosion–fatigue–
fretting, the in-vitro evaluation of a modular prosthesis is difficult
⁎ Corresponding author at: PX Group S.A. Dep R&D, Bd. des Eplatures 42, CH-2304 La
Chaux-de-Fonds, Switzerland. Tel.: +41 32 924 02 90; fax: +41 32 924 02 10.
E-mail address: lucien.reclaru@pxgroup.com (L. Reclaru).
[9]; the complexity of the situation with this implant resides in the
existence of interfaces between the distal module, the proximal
module, and the dynamometric screw in the presence of cyclic
dynamic constraints.
The connection in the majority of such modular implants is designed
as a taper. To achieve optimum biocompatibility, often the modular
parts of these prostheses are made of titanium alloys. Unfortunately,
titanium alloys are notch sensitive and tend to gall. Both of these
properties are detrimental for taper joints. Therefore, control of the
tendency to gall as well as the elimination of notches, either due to
design or manufacturing, is crucial for the long term success of these
modular connections.
In order to go a step further in investigating the safety and
effectiveness of such modular prostheses, a novel technique for
the evaluation of the resistance to cyclic loading under in situ
conditions, the dynamic fatigue test with crevice corrosion stimulation
was established.
The term cyclic (dynamic) fatigue test with stimulation of crevice
corrosion covers various interrelated phenomena: crevice corrosion,
fatigue and tribocorrosion. The term (static) stress corrosion with
stimulation in the crevice covers two distinct phenomena: crevice
corrosion and stress corrosion. In order to clarify these terms, the
definitions are given below: Crevice corrosion is localised corrosion of

0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.msec.2013.10.005
 Author's personal copy

70 L. Reclaru et al. / Materials Science and Engineering C 34 (2014) 69–77

a metal surface at, or immediately adjacent to, an area that is shielded Table 1
from full exposure to the environment because of close proximity Types of modular prostheses tested and experimental conditions.

between the metal and the surface of another material [11]; Fatigue Series Alloy Surface Tolerance Code Mechanical
corrosion is a process in which a metal fractures prematurely under conditions
conditions of simultaneous corrosion and repeated cyclic loading at 1st series Ti6Al7Nb Without specific Uniform #1 Dynamic
lower stress levels or fewer cycles than would be required in the treatment #2 constraint
absence of the corrosion environment [11]; Tribocorrosion groups #3
#4 Static constraint
together all the mechanical and chemical interactions which cause the
2nd series Ti6Al4V 2nd type of Adapted to every #5 Dynamic
degradation of solids in relative displacement with or without contact anodizing level of the taper #6 constraint
lubricant [12] and Stress corrosion cracking results from the combined (“Ti-anodizing”) #7
action of corrosion and a static compressive or tensile stress, the latter #8 Static constraint
being applied or residual [12].
The aim of the study presented here was to compare the results of
this in vitro technique with an evaluation of retrieved implants to
determine whether the method reproduces failures observed in vivo. The mechanical parameters were developed by Plus Orthopaedics
Under dynamic mechanical load, all interfaces of the taper joint AG using the following characteristics: a minimum of 0.25 kN and a
may be exposed to fretting and bearing in mind the surrounding maximum 3.3 kN loads, an amplitude of 1.1 mm at a frequency of
body fluid and the crevice condition while implanted, this fretting 10 Hz and a control function of cyclic load (sinusoidal).
will be amplified by corrosion and become tribo-corrosion. The first Two types of mechanical tests were conducted, one test under
modular prosthetic system tested with this new technique is a dynamic loading for 5 million fatigue cycles and a second test under a
modular revision hip stem made from either Ti6Al4V or Ti6Al7Nb static force of 981 N during the equivalent time corresponding to
alloys. 5 million dynamic fatigue cycles. Five million dynamic fatigue cycles
correspond to approximately 5 years of walking for a person with a
bodyweight of 100 kg.
2. Materials and methods
The parameters of sample placement followed the requirements of
the ASTM F1440 standard.
2.1. Preparation of samples

The test samples were complete modular prostheses prepared

2.3. Corrosion behaviour evaluation
by Plus Orthopedics AG, Fig. 1. The test was conducted on two
series of these modular prostheses. The main difference between
The potentiostatic electrochemical technique (controlled potential
them is the alloy composition: The first series was made of
coulometry) adapted in accordance with the ASTM F746–87 standard
Ti6Al7Nb alloy and the second of Ti6Al4V alloy. Table 1 presents
[10] was used to evaluate the corrosion behaviour of the prostheses.
the differences between the two types of modular prostheses and
This consists of causing an excitation of a given potential for a very
the experimental conditions.
short period of time and then putting on a constant potential for a
For each series of prostheses, four prostheses were evaluated: 3
given time. The steps of electrochemical measurement for one
specimens were used in cyclic fatigue dynamic tests with stimulation
electrochemical cycle are shown in Table 2. For ten electrochemical
of crevice corrosion and one prosthesis was used under static stress
cycles, 40 electrochemical measurements were performed at four
stimulation of crevice corrosion, summarised in Table 1.
levels: 600 mV, 650 mV, 700 mV and 750 mV. The electrochemical
cycles were performed at the same time as the mechanical tests.
2.2. Fatigue behaviour evaluation Ten electrochemical cycles correspond to a million mechanical
fatigue cycles. The test was conducted for a total of 5 million dynamic
For the fatigue tests, an LFV 10–63 kN series machine (Walter & fatigue cycles.
Bai AG, Switzerland) was used and adapted for fatigue testing and
research of biomedical implants (Hip Implant Prostheses according
to ISO 7206-4 and 6).
A mechanical–electrochemical cell was designed especially for
these tests (Plus Orthopaedics, AG), Fig. 2. It is cylindrical with a
central glass tube to guide it onto the sample holder. The distal
module of each prosthesis was embedded in a resin. (elastic modulus
3.8 MPa).The embedding of the tapered shape was up to 1 cm from
the boundary between the distal module and the proximal module.

Fig. 1. Samples tested (modular prostheses): a. distal part and b. proximal part with the
dynamometric screw. Fig. 2. Mechanical–electrochemical cell used for corrosion–fatigue tests.
 Author's personal copy

L. Reclaru et al. / Materials Science and Engineering C 34 (2014) 69–77 71

Table 2
Steps of electrochemical measurement for one electrochemical cycle. 30

No. Step Potential [V] Time [min]

Current (uA)
20
1. Excitation 0.800 1
2. Potentiostatic level 0.600 36
3. Excitation 0.800 1 10
4. Potentiostatic level 0.650 36
5. Excitation 0.800 1
6. Potentiostatic level 0.700 36 0
7. Excitation 0.800 1
8. Potentiostatic level 0.750 36
9. Potentiostatic level −0.500 20 -10
0 10 20 30 40
Time (min)
The reasoning behind the choice of measuring technique is based on
Fig. 3. Potentiostatic curves recorded for the equaliser of 0.600 V vs. SCE during the first
the following considerations: million dynamic fatigue cycles for prosthesis #7.
– The materials chosen in the manufacture of components do not
really pose any problems for the general form of corrosion
has a total area of 115.9 cm2. The electrolyte used in the test was a
resistance. On the other hand, seepage of electrolyte into the spaces
solution of HCl 0.07 N (300 ml) prepared from Titrisol® 1.0 N (Merck).
of the conical parts of the interface between the distal and proximal
For extraction tests the release solution volume (ml)/total sample
modules may generate localised corrosion in the crevice shape,
surface (cm2) ratio was equal to 3. The choice of electrolyte extraction
along with a tribocorrosion process.
was based on thermodynamic considerations [13] (solubilities), on
– The readings of the applied electrical potential remain in the range
standards (EN-71-3) about bioavailability [14] and on constraints of
where there is no process of decomposition of the electrolyte into
the analysis technique (simple matrix causes no pertubations).
hydrogen and oxygen.
The reactors were first washed with concentrated nitric acid for
– A cathode potential equaliser at − 500 mV vs. the saturated
24 h, then thoroughly rinsed with deionized water (18 MΩ·cm) to
calomel electrode (SCE) was added with the aim of analysing
eliminate the acid completely and finally they were dried. The
the depassivation–repassivation capacity of the materials for
extraction tests were conducted at 37 ± 2 °C. After 168 h, the
the areas tested.
prosthesis was removed from the extraction solution, rinsed with
– Using this technique, the total quantity of electrical charge consumed
dionized water and then dried. 50 ml of the extraction solution was
during the experiment can be easily measured. For analytical
then analysed by inductively coupled optical emission spectroscopy
determinations, the total electrical charge passed in the experiment
(ICP-OES) and mass spectroscopy (MS). A blank solution was measured
is easily related to the concentration of electroactive species in the cell.
as a reference. Two surface treatments of prostheses were made on the
The potentiostat used was the model PAR 273A EG&G (Princeton Ti6Al7Nb alloy. Treatment 1 (Corundum blasted) and Treatment 2
Applied Research). The counter electrode is made of platinum and the (Corundum blasted + mechanical + chemical cleaning) were used for
reference electrode is a saturated calomel electrode. The electrochemical the cation extraction tests. For each treatment, three prostheses were
cell was provided with a Luggin capillary for the reference electrode used.
(Fig. 2). The test environment was a physiological solution of NaCl at a
concentration of 9 g/l [10] in ultra-pure water (electrical resistivity 2.5. Explanted samples
18 MΩ∙cm) added to a level such that it covers 1 cm of the height of the
proximal module. Two components of modular prostheses were explanted after
To eliminate any transfer of electrical charge during the 7 years from one patient (female, DOB 1925). The proximal modules
electrochemical measurements, the sample holder was protected of Ti6Al7Nb (PL-06, Plus 662, Gr BS), (#9) and Ti6Al4V (PL-06, Plus
with an insulation of PTFE. Another protection was made in the 662, Gr. BL), (#10) alloys were analysed by optical and scanning
plane zone of the sample holder-coating of the distal model with electron microscopy. These were compared with the prostheses
a silicone rubber protective coating for electronic components used in the cyclic fatigue dynamic tests with stimulation of crevice
(3140 Dow Corning). Thus the metal parts of the sample holder corrosion for any evidence of deterioration and to correlate these
and the reaction resin were covered and there was no transfer of observations with the results obtained from the laboratory tests.
charge between the counter electrode and the reference electrode
when readings were being taken. Table 3
After the tests, the sample was cut up manually using a saw in order Quantity of electrical charge (mC) consumed for the first million dynamic fatigue cycles
to investigate inside the crevice at the interface of the distal and (ten electrochemical cycles) during the testing of prostheses #2 (Ti6Al7Nb) and #7
(Ti6Al4V).
proximal modules.
Cycle Level 0.600 V Level Level Level Total (mC)
2.4. Cation extraction 0.650 V 0.700 V 0.750 V

#2 #7 #2 #7 #2 #7 #2 #7 #2 #7
The release of cations from implants is an important subject. The
1 108.2 37.3 69.7 21.9 54.1 17.8 50.3 16.5 282.3 93.5
toxic effects of cations released from prosthetic implants have been 2 30.9 12.3 26.3 7.4 25.9 7.3 25.3 8.1 108.4 35.1
associated with clinical implant failure, osteolysis, cutaneous allergic 3 23.0 10.1 18.4 5.3 18.9 5.1 20.3 5.7 80.6 26.2
reactions and accumulation at remote sites. 4 19.3 8.9 15.4 4.1 16.0 3.8 16.7 4.0 67.4 20.8
5 15.7 8.4 13.2 3.5 13.3 3.0 14.8 3.3 57.0 18.2
Metals from orthopaedic implants are released into surrounding
6 10.8 8.2 10.5 3.1 10.2 2.7 10.5 2.9 42.0 16.9
tissue by various mechanisms: corrosion, wear and associated 7 10.7 7.8 8.3 2.9 7.9 2.4 8.6 2.6 35.5 15.7
electrochemical processes such as stress corrosion, corrosion fatigue 8 8.7 7.9 6.8 2.8 7.3 2.3 7.8 2.5 30.6 15.5
and fretting corrosion. 9 8.1 7.9 6.4 2.7 6.4 2.2 6.8 2.4 27.7 15.2
To minimize the volume to surface ratio, glass Pyrex reactors were 10 7.8 8.1 5.7 2.7 5.9 2.2 6.1 0.9 25.5 13.9
Total (mC) 243.2 116.9 180.7 56.4 165.9 48.8 167.2 48.9 757.0 271.0
developed and adapted to the prostheses shapes. The femoral element
 Author's personal copy

72 L. Reclaru et al. / Materials Science and Engineering C 34 (2014) 69–77

100

Electrical Charge (mC) Consumed

75

50

25

0
0 40 80 120 160 200
200 electrochemical measurements for 5 million fatigue cycles

#1 #2 #3 #4

Fig. 4. Behaviour of localised corrosion during the 5 million mechanical fatigue cycles for Ti6Al7Nb series-#1, #2, #3, and #4. Prosthesis #4 has been evaluated for corrosion resistance
without cyclic dynamic forces but under a load of 981 N.

2.6. Surface characterization
An optical microscope and a scanning electron microscope Jeol JSM
6300 were used to evaluate the surface of the prostheses after the
mechanical fatigue cycles and of the explanted implants.
3. Results and discussions
3.1. Electrochemical characterization
For all the electrochemical tests the potentiostatic plots were
recorded in steps of +50 mV between the open circuit potential up to
+0.750 V/SCE. Fig. 3 shows the potentiostatic curves recorded for the
equaliser of 0.600 V/SCE during the first million mechanical fatigue
cycles for a prosthesis made on Ti6Al4V alloy (#7). It should be noted
that the quantities of current recorded during the scan were of the
order of micro-amperes.
From the potentiostatic curves recorded, the quantities of electrical
charge were calculated by integration in microcoulombs consumed in
the fatigue process for each potential equaliser.
By way of example, Table 3 shows the quantities of electrical charge
consumed in the first million dynamic fatigue cycles for prostheses #2
(Ti6Al7Nb alloy) and #7 (Ti6Al4V alloy).
It should be noted that the total electrical quantity of charge
consumed for the first million dynamic fatigue cycles by sample #2
(Ti6Al7Nb) is 2.5 times bigger than that consumed by sample #7
(Ti6Al4V). Prosthesis #2, which is made of Ti6Al7Nb alloy shows a
worse corrosion behaviour than prosthesis #7 made of Ti6Al4V alloy.
The entirety of results (200 electrochemical measurements
corresponding to 50 electrochemical cycles) for the 1st series of
prostheses measured in corrosion is shown in Fig. 4. Sample #4 was
evaluated for corrosion resistance without cyclic dynamic forces but
under a static load of 981 N, during the equivalent time corresponding
to 5 million dynamic fatigue cycles.
The results for the 2nd series are shown in Fig. 5. All the specimens
made of Ti6Al4V alloy and tested under dynamic fatigue cycles present
better corrosion crevice behaviour than those made of Ti6Al7Nb alloy.
Sample #8 was evaluated under the same conditions as sample #4.
It should also be noted that the curve shapes vary greatly
between the two series. The harmonic shape of the curves of
prostheses from the 2nd series is probably due to the surface
treatment employed and the degree of degradation [15,16]. Given
that the last equaliser from the electrochemical measurement cycle
is at − 0.500 V vs. SCE (Table 2), that is a depassivation cathode
equaliser; the repassivation capacity of the prosthesis coating of
this series shows a less rapid aptitude for repassivation. However, in

50
Electrical Charge (mC) Consumed

40

30

20

10

0
0 40 80 120 160 200
200 electrochemical measurements for 5 million fatigue cycles
#5 #6 #7 #8

Fig. 5. Behaviour of localised corrosion during the 5 million mechanical fatigue cycles for Ti6Al4V series-#5, #6, #7, and #8. Prosthesis #8 has been evaluated for corrosion resistance
without cyclic dynamic forces but under a load of 981 N.
 Author's personal copy

L. Reclaru et al. / Materials Science and Engineering C 34 (2014) 69–77 73

4000
the level of the cones and, as a result, initiation of crevice corrosion
Cumulated Charge (mC)

3000
2000
1000
0
#1 #2 #3 #4 #5 #6 #7 #8
Code samples
Fig. 6. Electrical charges accumulated over 5 million fatigue test cycles. Prostheses #4 and
#8 were evaluated under static stress.
the first equaliser of 0.600V the quantities of electrical charge are higher
in relation to the last equaliser at 0.750V vs. SCE (Table 3). Also, it can be
observed that the potentiostatic curves always start in the cathode
region (negative current readings) (Fig. 4). Electrochemically prepared
oxide layers may be in the crystalline or amorphous state depending on
the anode potential and the electrolyte solution used and they are
adsorbed during repassivation, so that the ions of the medium are
incorporated in the newly formed oxide layer [17,18].
Also for the same series made of Ti6Al4V alloy, the curve shapes do
not have any correlation to the cyclic fatigue movement applied in the
test. It can be noted that under static fatigue test, the curve of prosthesis
#8 has the same form as in the cyclic dynamic fatigue tests (Fig. 5).
Another important observation that can be made is that in the
dynamic fatigue tests of the first series, after around 3 million cycles,
crystallised sodium chloride was found at the level of the press screw
of the two modules. This indicates that there is an effect of pumping
the electrolyte into the space of the assembly of the two modules at
becomes possible. For the prostheses of the second series, this process
may be noted more rapidly, from around 2.5 million cycles. Other
researchers studied the corrosion–fatigue behaviour of Ti and its alloys
(Ti6Al4V, Ti6Al7Nb) and observed also small amounts of salts from
the media remaining on the specimen surface [19,20] or corrosion
products [21] after different cycle numbers [22,23].
During the static mechanical electrochemical measurements (#4
and #8), the electrolyte does not reach the level of the press cones, so
this electrolyte pumping phenomenon is not noted.
Fig. 6 summarises the results obtained in the electrochemical-
fatigue tests. Analysing the sum of charges accumulated during the
entire test period it can be noted that the 1st series prostheses
show the worst corrosion behaviour. With regard to the composition
of each alloy, and thus the corrosion resistance of each constituent
metal, Ti, Al, V, Mo and Zr to a lesser extent displayed a strong
resistance to both pitting and crevice corrosion. Conversely, Nb
appeared to be much more vulnerable as a metal [24].
3.2. Microscopic observation of cyclic fatigue tested prostheses
When interpreting the microscopic observations, it is essential to
take into account the fact that it is not possible to consider the various
phenomena involved separately, which are: crevice corrosion, fatigue
(cyclic dynamic test), stress (static test) and tribocorrosion (cyclic
dynamic test). These processes are intimately interrelated by complex
mechanisms; the observation is limited to recording composite effects.
This is why the phenomenon studied in this report is designated by the
term fatigue corrosion.
Fig. 7 presents photographs of the 3 prostheses following dynamic
fatigue cyclic testing with crevice corrosion stimulation. Fig. 7a and b
show images of the distal module of prosthesis #1 made of Ti6Al7Nb
alloy revealing a crack on its surface after the dynamic fatigue cycles.

#1 #2 #3

a c e

b d f
560x

Fig. 7(f) Ti Al Nb O Na Si Cl K
1 88.20 6.95 4.84 - - - - -
2 82.84 6.11 - - 3.28 0.71 3.63 3.43
3 72.60 5.69 7.18 14.54 - - - -
Fig. 7. Ti6Al7Nb alloy prostheses after dynamic fatigue cyclic testing with crevice corrosion stimulation: Prosthesis #1 showing a crack on the distal module (a and b); prosthesis #2 shows
coloration and salt deposition at the interface of the distal and proximal modules (c) and on the proximal module (d); Prosthesis #3 showing salt deposition on the distal module with EDX
analysis (e and f).
 Author's personal copy

74 L. Reclaru et al. / Materials Science and Engineering C 34 (2014) 69–77

Prosthesis #2 shows interference coloration of the alloy due to Table 4

oxidation or a deposit at the interface of the two modules, Fig. 7(c) The quantities of cations released (microgrammes/litre) in a medium of 0.07 N HCl. The
values of cations released are the average of three prostheses tested.
and the presence of a deposit at the level of the crevice (Fig. 7(d)).
Scanning electron microscopy shows microscopic cracks located on Element Method Blanc Treatment 1 Treatment 2 Corundum
the distal module at the level of the crevice. Griza [25] observed Corundum blasted blasted + mechanical +
chemical cleaning
three regions of crack initiation on an uncemented hip stem, two
being in the transverse plane of the neck, in the intersection with As ICP-OES volatile b1 1.8 ± 1.6 1.1 ± 0.5
hydrides
the hole, whereas the other was in the high part of the neck, which
Ba ICP-MS 0.24 0.7 ± 0.2 0.4 ± 0.3
suffers from higher positive tension due to external load. EDX Be ICP-MS b0.1 b0.1 b0.1
analysis in this region revealed only Al, Ti and Nb. Br ICP-MS 3.54 5.6 ± 2.9 5.3 ± 1.8
Prosthesis #3 shows holes on the distal module at the level of the Cd ICP-MS b0.02 0.04 ± 0.08 0.07 ± 0.2
crevice as well as coloration at the same level on the proximal module, Co ICP-MS 0.02 0.17 ± 0.1 0.78 ± 0.03
Hg ICP-MS 0.81 0.12 ± 0.04 0.3 ± 0.5
Fig. 7(e). From the EDX (spectrum 1 and 3) it appears that the inside of
Li ICP-MS 0.11 0.14 ± 0.07 0.16±
the hole is depassivated (lack of oxygen in spectrum 1); the edges of the Mo ICP-MS b0.02 b0.02 b0.02
hole contain elements Na, Cl, Si and K. Also, on the distal module, at the Nb ICP-MS b0.02 63 ± 18 116 ± 23
level of the distal/proximal module interface, where deposits were Pb ICP-MS 1.9 0.9 ± 0.2 0.70 ± 0.5
Sb ICP-MS 0.06 0.05 ± 0.04 0.09 ± 0.02
observed; the EDX analysis in particular shows the presence of elements
Sn ICP-MS 0.13 0.10 ± 0.04 0.14 ± 0.02
Na, Si, Cl and K, Fig. 7(f). Zr ICP-MS b0.02 0.17 ± 0.04 0.03 ± 0.02
Fig. 8(a.) reveals the same penetration phenomenon of the Al ICP-OES 18 191 ± 19 157 ± 63
electrolyte at the interface of the distal/proximal modules for a Ca ICP-OES 0.0 0.01 ± 0.02 0.07 ± 0.1
Ti6Al4V prosthesis as was observed for Ti6Al7Nb prostheses. Visual Cr (total) ICP-OES b0.5 2.0 ± 0.4 0.90 ± 0.0
Cu ICP-OES b2 3.4 ± 4.7 b2
inspection established the presence of wear and corrosion in this
Fe ICP-OES b2 35 ± 12 45 ± 16
area, revealed by a rough aspect of the surface of the proximal module Ni ICP-OES b2 b2 b2
(Fig. 8(a)) as well as, to a lesser degree, on the distal module. This P ICP-OES b10 b10 b10
phenomenon is probably due to tribocorrosion and crevice corrosion. S ICP-OES b0.02 b0.02 b0.02
Ti ICP-OES 0.7 1453 ± 179 1563 ± 110
Inspection of the distal module did not reveal the presence of cracks.
V ICP-OES b0.2 b0.2 b0.2
Prosthesis #6 also shows an increase in the roughness of the Zn ICP-OES b2 6.4 ± 2.3 9 ± 12
surface at the level of the crevice (Fig. 8(c.)). The EDX analyses in

#5 #6 #7

a c
e

b d f

Fig. 8(d) Ti Al V O Si Cl Zr
1 92.64 2.47 3.94 - 0.95 - -
2 70.76 1.96 3.68 20.11 2.77 - 0.72
3 84.98 2.19 3.96 - 8.67 0.20 -

Fig. 8(f) Ti Al V Na Si O
1 77.09 6.25 3.49 0.65 12.53 -
2 63.33 4.98 2.90 1.59 26.89 0.31
3 63.10 4.41 3.15 0.72 28.62 -

Fig. 8. Ti6Al4V alloy prostheses after dynamic fatigue cyclic testing with crevice corrosion stimulation: Prosthesis #5 showing coloration and salt deposition on the distal (a) and proximal
modules (b); Prosthesis #6 showing a rough area on the distal modules (c and d) and also on prosthesis #7 on the distal module (e) with EDX analysis below (d and f).
 Author's personal copy

L. Reclaru et al. / Materials Science and Engineering C 34 (2014) 69–77 75

the rough area of the distal module reveal the presence of elements
Zr, Cl and Si (Fig. 8(d)).
A cross-section of the proximal module of prosthesis #7 is presented
in Fig. 8(e and f). Just as for prostheses #5 and #6, this cross-section
reveals salt deposits as well as wear and/or corrosion in the crevice
area; on the other hand, the distal module is much less marked by this
phenomenon than the distal module of prosthesis #6.
The EDX analyses are measured on the distal module in the crevice
area. The presence of Na, Si and O can be observed (Fig. 8(f)).
Examination of the distal and proximal modules of prosthesis #8
(Ti6Al4V alloy), which had been subjected only to a static test revealed
that the electrolyte did not penetrate into the distal/proximal module
interface; there is no evidence of corrosion phenomenon. This was
also seen for prosthesis #4 made of Ti6Al7Nb alloy.
3.3. Cation extraction
The data are presented in Table 4. The values presented in the table
are an average of concentrations of the chemical elements released from
three prosthesis analysed for each treatment. There were no significant
differences in the release of elements in a dilute solution of hydrochloric
acid 0.07 N between the two surface treatments of prostheses made
from Ti6Al7Nb alloy. The elements at the detection limit (Be, Mo, Ni,
P, S and V) and those that have the released value identically to the
blank solution (Br, Hg, Pb and Sn) are not considered important.
Concerning the alloying elements, the presence of titanium and
aluminium is comparable for both series of implants and conform
with those found in the literature [26–29]. Spriano [26] also reported
an increase of metal ion concentration after long exposure of Ti6Al7Nb
alloy to an SBF solution.
On the other hand, the concentration of niobium cations released
from the surface of the Treatment 1 prostheses is smaller than for
surface Treatment 2 (63 to 116 μg·l−1, respectively).
The most important impurity is iron which was between 35 and
45 μg·l− 1 and almost identical for the two types of prostheses. The
specimens of surface Treatment 2 release less chromium than of
surface Treatment 1, 0.9 and 2 μg·l−1 respectively. The prostheses with
surface Treatment 1 released 3.4 μg·l−1 cooper. The prostheses with
surface treatment 1 released 1.4–2.6 μg·l−1 arsenic. The prostheses
with mechanical and chemical cleaning released 0.78 μg·l−1 cobalt

Ti6Al7Nb prostheses Ti6Al4V prostheses

d
a

c
f

Fig.9(c) C O Na Al Ti Nb
1 4.7 41.6 0.4 1 48.6 3.8
2 2.8 37.2 - 3.2 53.1 3.7
3 4.5 43.1 0.4 1 46.6 4.4

Fig. 9(f) C O Na Al Si Ti V
1 2.76 62.68 0.72 1.74 13.61 17.69 0.80
2 27.43 45.26 0.32 1.42 8.02 16.89 0.66

Fig. 9. Analyses of the proximal module of Ti6Al7Nb alloy explanted prosthesis #9 (a, c and EDX analysis) and evaluated following cyclic dynamic testing with stimulation of crevice
corrosion, #2 (b). The same analyses for prostheses made on Ti6Al4V alloy proximal module: explanted, #10 (d, f and EDX analysis) and evaluated after cyclic dynamic tests with
stimulation of crevice corrosion, #7 (e).
 Author's personal copy
76 L. Reclaru et al. / Materials Science and Engineering C 34 (2014) 69–77
compared to 0.17 μg·l−1 for the non-cleaned implants. Some of these
cations released into the solution probably come from the glassy reactor
(Pb and Sn) and from the HCl impurities (according to Merck
information). Some of them are considered accidental impurities
(As, Cd and Cu). The presence or the absence of these accidental
impurities did not relate to the type of surface treatment.
3.4. Explanted prostheses
The objective of this part of the study was the observation and
analysis of two components of explanted modular prostheses (proximal
module) and comparison with the prostheses tested under dynamic
fatigue cycles. The distal parts could not be compared because there
were insufficient explanted parts.
Fig. 9 shows a comparison of the explanted proximal module made
from Ti6Al7Nb with the proximal module made of the same alloy after
cyclic dynamic testing with stimulation of crevice corrosion. Significant
similarities are apparent particularly with regard to the phenomena of
electrolyte penetration, salt deposition and corrosion. Fig. 9 shows
that there has been penetration of biological fluid between the
interfaces of the proximal and distal modules.
The arrows in Fig. 9(b) indicate the location where the EDX analyses
were taken for the Ti6Al7Nb prosthesis. Observation by SEM of the
upper part of the proximal/distal module shows the presence of a
deposit from the biological fluids which has penetrated the interface,
certified by the presence of carbon and sodium, Fig. 9(c). The point in
Fig. 9(c) shows the presence of localised mechanical wear which is in
accordance with Fig. 9(a). The pumping effects of the electrolyte
were observed after 3 millions dynamic fatigue cycles. This type of
effect was also observed on the explanted prostheses.
A comparison of the explanted Ti6Al4V prosthesis and the one
tested for 5 millions dynamic fatigue cycles, Fig. 9(d and e) show
that the location of the visible patches on the explanted proximal
module Fig. 9(d) correspond approximately to the electrolyte
deposits found during the cyclic dynamic tests with stimulation of
crevice corrosion (Fig. 9(e)). The SEM images do not show any
apparent localised corrosion in these patches, Fig. 9(f). The EDX
analyses show the presence of carbon and sodium and indicate that
the biological fluid penetrated and diffused at the proximal/distal
module interfaces. An analogous pumping electrolyte phenomenon
at the interface of proximal and distal modules was observed in the
evaluation of the modular prostheses after dynamic fatigue cycling
with crevice stimulation.
4. Conclusion
The results obtained by electrochemical analysis and visual
observation reveal better corrosion behaviour for prostheses made
of anodised type 2 Ti6Al4V alloy compared to prostheses made of
Ti6Al7Nb alloy. The latter implants showed cracking at the level of
the distal module and holes near the crevice corrosion zone. The
most important impurities released from these prostheses were Fe,
Ba, Zn, Co, Cr, and Zr. However, some of the cations released in the
solution were considered to be accidental impurities (As, Cd and Cu).
Electrolyte had penetrated into the interface between the distal
and proximal modules of the prostheses made from Ti6Al7Nb alloy
during dynamic cyclic tests with stimulation of crevice corrosion,
which was not the case during the static stress tests. The interference
coloration at the level of the distal/proximal module interface on
samples of the 1st series (Ti6Al7Nb) is witness to reactions of the
electrolyte with the substrate; without there being any evidence of
corrosion. The prostheses made of Ti6Al4V alloy did not show any
coloration, cracks or holes however, at the level of the crevice
corrosion zone, an increase in surface roughness was observed
after the dynamic fatigue corrosion test (particularly on prosthesis
#6). Observation of the samples submitted solely to static stress
tests did not show corrosion.
Microscopic evaluation of the explanted prostheses certifies the
electrolyte pumping phenomena, salt deposition and corrosion
products at the interface between the proximal and distal modules as
also observed following cyclic fatigue loading. This analysis indicates
that the use of dynamic fatigue cycle testing with stimulation of
crevice corrosion for the evaluation of biomaterials used for modular
prostheses is justified.
Acknowledgements
Dr. Roger Brooks acknowledges funding support from the National
Institute for Health Research.
References
[1] I.C. Lavos-Valereto, S. Wolynec, I. Ramires, A.C. Guastaldi, Electrochemical
impedance spectroscopy characterization of passive film formed on implant
Ti-6Al-7Nb alloy in Hank's solution, J. Mater. Sci. Mater. Med. 15 (2004) 55–59.
[2] A. Baron, W. Simka, W. Chrzanowski, EIS tests of electrochemical behaviour of
Ti6Al4V and Ti6Al7Nb alloys, J. Achiev. Mater. Manuf. Eng. 21 (1) (2007) 23–26.
[3] Z. Paszenda, W. Walke, S. Jadacka, Electrochemical investigations of Ti6Al4V and
Ti6Al7Nb alloys used on implants in bone surgery, J. Achiev. Mater. Manuf. Eng.
38 (1) (2010) 24–32.
[4] S. Luiz de Assis, S. Wolynec, I. Costa, Corrosion characterization of titanium alloys by
electrochemical techniques, Electrochim. Acta 51 (8–9) (2006) 1815–1819.
[5] M. Aziz-Kerrzo, K.G. Conroy, A.M. Fenelon, S.T. Farrell, C.B. Breslin, Electrochemical
studies on the stability and corrosion resistance of titanium-based implant
materials, Biomaterials 22 (12) (2006) 1531–1539.
[6] S. Tamilselvi, V. Raman, N. Rajendran, Corrosion behaviour of Ti–6Al–7Nb and
Ti–6Al–4V ELI alloys in the simulated body fluid solution by electrochemical
impedance spectroscopy, Electrochim. Acta 52 (3) (2006) 839–846.
[7] D.C. Rodrigues, R.M. Urban, J.J. Jacobs, J.L. Gilbert, In vivo severe corrosion and
hydrogen embrittlement of retrieved modular body titanium alloy hip implants,
J. Biomed. Mater. Res. B Appl. Biomater. 88 (1) (2009) 206–219.
[8] T.M. Grupp, T. Weik, W. Bloemer, H.-P. Knaebel, Modular titanium alloy neck
adapter failures in hip replacement — failure mode analysis and influence of implant
material, BMC Musculoskelet. Disord. 11 (3) (2010) 1–12.
[9] J.L. Gilbert, J.J. Jacobs, The mechanical and electrochemical processes associated with
taper fretting crevice corrosion: a review, in: D.E. Marlowe, J.E. Parr, B.M. Mayor
(Eds.), Modularity of Orthopaedic Implants, STP 1301, ASTM, USA, 1997.
[10] ASTM F746–87, Standard test method for pitting or crevice corrosion of metallic
surgical implants, (reapproved 1994) Medical Devices and Service, vol. 13.011998.
187–192 (Philadelphia).
[11] ASTM G15–08, Standard terminology relating to corrosion and corrosion
testing, Metals Test Methods and Analytical Procedures, vol. 03.022008. 57–61
(Philadelphia).
[12] D. Landolt, Corrosion et chimie de surfaces des métaux, Presses Polytechniques et
Universitaires Romandes, Lausanne, 1993. 552.
[13] M. Pourbaix, Atlas of electrochemical equilibria in aqueous solutions, National
Association of Corrosion Engineers, Houston, Texas, 1974. 644.
[14] EN 71–3, Safety of toys, part 3: migration of certain elements, Directive 88/378/EEC
(C 188, 2005-08-02)1994.
[15] Z. Cai, H. Nakajima, M. Woldu, A. Berglund, M. Bergmann, T. Okabe, In vitro
corrosion resistance of titanium made using different fabrication methods,
Biomaterials 20 (2) (1999) 183–190.
[16] L. Reclaru, R. Lerf, P.-Y. Eschler, A. Blatter, J.M. Meyer, Evaluation of corrosion on
plasma sprayed and anodized titanium implants, both with and without cement,
Biomaterials 24 (18) (2003) 3027–3038.
[17] C. Fleck, D. Eifler, Corrosion, fatigue and corrosion fatigue behaviour of metal
implant materials, especially titanium alloys, Int. J. Fatigue 32 (6) (2010) 929–935.
[18] C. Leinenbach, D. Eifler, Influence of oxidation treatment on fatigue and
fatigue-induced damage of commercially pure titanium, Acta Biomater. 5 (7)
(2009) 2810–2819.
[19] C. Leinenbach, C. Fleck, D. Eifler, The cyclic deformation behaviour and fatigue
induced damage of the implant alloy TiAl6Nb7 in simulated physiological media,
Int. J. Fatigue 26 (8) (2004) 857–864.
[20] M.K. Dimah, F. Devesa Albeza, V. Amigó Borrás, A. Igual Muñoz, Study of the
biotribocorrosion of titanium biomedical alloys in simulated body fluids by
electrochemical techniques, Wear 294–295 (2012) 409–418.
[21] V. Chandrasekaran, W.L. Sauer, A.M. Taylor, D.W. Hoeppner, Evaluation of the
fretting corrosion behaviour of the proximal pad taper of a modular hip design,
Wear 231 (1) (1999) 54–64.
[22] C. Leinenbach, D. Eifler, Fatigue and cyclic deformation behaviour of surface-modified
titanium alloys in simulated physiological media, Biomaterials 27 (9) (2006)
1200–1208.
[23] J.R. Goldberg, J.L. Gilbert, The electrochemical and mechanical behaviour of
passivated and TiN/AlN-coated CoCrMo and Ti6Al4V alloys, Biomaterials 25 (5)
(2004) 851–864.
 Author's personal copy
L. Reclaru et al. / Materials Science and Engineering C 34 (2014) 69–77
[24] M.A. Khan, R.L. Williams, D.F. Williams, In-vitro corrosion and wear of titanium
alloys in the biological environment, Biomaterials 17 (22) (1996) 2117–2126.
[25] S. Griza, M. Reis, Y. Reboh, A. Reguly, T.R. Strohaecker, Failure analysis of
uncementated total hip stem due to microstructure and neck stress riser, Eng. Fail.
Anal. 15 (7) (2008) 981–988.
[26] S. Spriano, M. Bosetti, M. Bronzoni, E. Vern, G. Mainac, V. Bergo, M. Cannas,
Surface properties and cell response of low metal ion release Ti–6Al–7Nb
alloy after multi-step chemical and thermal treatments, Biomaterials 26 (11)
(2005) 1219–1229.
77
[27] D. Zaffe, C. Bertoldi, U. Consolo, Element release from titanium devices used in oral
and maxillofacial surgery, Biomaterials 24 (6) (2003) 1093–1099.
[28] D. Martini, M. Fini, M. Franchi, V. DePasquale, B. Bacchelli, M. Gamberini, A. Tinti, G.
Giavaresi, V. Ottani, M. Raspanti, S. Guizzardi, A. Ruggeri, Detachment of titanium
and fluorohy-droxyapatite particles in unloaded endosseous implants, Biomaterials
24 (7) (2003) 1309–1316.
[29] J. Wang, B. Wicklund, R. Gustilo, D. Tsukayama, Titanium, chromium and cobalt ions
modulate the release of bone-associated cytokines by human monocytes/macrophages
in vitro, Biomaterials 17 (23) (1996) 2233–2240.