* Corresponding author. Tel.: [phone]; fax: [phone].

Biomaterials 20 (1999) 933 — 941

Measurement and evaluation of galvanic corrosion between

titanium/Ti6Al4V implants and dental alloys by electrochemical

techniques and auger spectrometry

Brigitte Grosgogeat , Lucien Reclaru

, Michele Lissac *, Francis Dalard

School of Dentistry, University Claude Bernard of Lyon I, Rue Guillaume Paradin, 69372 Lyon Cedex 08, France

P. X. TECH S.A., Blvd des Eplatures, 2304 La Chaux-de-Fonds, Switzerland

Laboratoire d+Electrochimie et de Physico-chimie des Mate&riaux et des Interfaces, Polytechnic Institut of Grenoble, UMR 5631, INPG-CNRS,

Rue de la Piscine 38402, Saint Martin D+He` res Cedex, France

Received 8 September 1997; accepted 28 November 1998

Abstract

The purpose of this study was to investigate, in different experimental conditions, the galvanic corrosion phenomena which can

exist between a dental suprastructure and a dental implant. The electrochemical behavior of 7 alloy superstructures with titanium and
titanium alloy (Ti6Al4V) implants was investigated by electrochemical means in Fusayama—Meyer de-aerated saliva and
Carter—Brugirard (AFNOR) non de-aerated saliva. Different techniques were used to obtain the value of the galvanic coupling current
and potential for each couple. All showed very low corrosion rates, ranging from 10

\ to 10\ A. Surface analysis confirmed these

results.

1999 Elsevier Science Ltd. All rights reserved

1. Introduction

Titanium and titanium alloy (Ti6Al4V) are now widely

used in odontology because of their excellent character-
istics such as chemical inertia [1, 2] mechanical resis-
tance [3], low density, absence of toxicity [4, 5], and
above all for their biocompatibility [6—10]. Nowadays,
implants and root pins made of these materials are used
by dental surgeons. Assessment of the resistance to cor-
rosion shown by surgical implants is abundantly
documented [11—21]. There remains, however, a problem
concerning the choice of the alloy used for the implant
borne suprastructure. The complexity of the electro-
chemical processes involved in the implant-suprastruc-
ture joint is linked to the phenomena of galvanic
coupling and pitted corrosion [22]. As the assess-
ment corrosion resistance is closely linked to the experi-
ment conditions, we decided to establish a comparative
assessment

of

the

electrochemical

measurements

obtained using different techniques and different prep-
arations to study the galvanic couple in Ti/dental

alloy and Ti6Al4V/dental alloy. A number of aspects
were studied:
(a) the measurement of galvanic currents in galvanic

couple (i) and potential in galvanic couple

(E);

(b) the measurement of galvanic current in the joints;
(c) the assessment of the surface structure.

2. Materials and methods

2.1. Specimen

The investigation of galvanic corrosion by electro-

chemical methods was carried out on seven dental alloys.
The breakdown by class and composition of the alloys
used is given in Table 1. The ratio of anode surface to
cathode surface was equal to 1.

The specimens used were cylinders (diameter 5 mm,

length 15 mm) placed in a polytetrafluoroethylene holder
specifically designed to connect to a revolving electrode
(type EDI 10000 Radiometer) They were polished metal-
lographically with diamond paste (1 mm).

The chemical composition of titanium and titanium

alloy (Ti6Al4V) is also shown in Table 1. The precious

0142-9612/99/$ — see front matter

1999 Elsevier Science Ltd. All rights reserved.

PII: S 0 1 4 2 - 9 6 1 2 ( 9 8 ) 0 0 2 4 8 - 8

Table 1
Classes and composition in percent by weight of the tested alloys in coupling with titanium and Ti6Al4V

Class

Code

Composition in weight%

Conventional alloys
High gold

1

68.5 Au

12.5 Ag

2.3 Pt

4.2 Pd

10.3 Cu

2.2 Zn

Low gold

2

61.5 Au

28.0 Ag

3.0 Pd

6.3 Cu

1.2 In

Silver-based

3

2.0 Au

58.5 Ag

0.1 Pt

27.4 Pd

10.4 Cu

1.6 Zn

PFM

alloys

High gold

A

81.0 Au

3.0 Ag

12.0 Pt

2.1 Pd

1.8 In

0.1 Ir

Low gold

B

45.0 Au

5.0 Ag

38.9 Pd

8.5—6.0 Sn

6.0 Cu

In 1.3 Ga

0.1 Ru

Pd-based

C

2.0 Au

79.0 Pd

9.6 Cu

9.0 Ga

1.5 Zn

0.3 Ru

Co—Cr

D

65.0 Co

20.0 Cr

6.0 Mo

6.0 W

2.0 Si

0.8 Nb

0.02 C

Titanium

Ti

Base

0.05 N

0.08 C

0.0125 H

0.25 Fe

0.25 O

Ti6Al4V

Ti6Al4V

Base

)

0.07 N

)

0.08 C

)

0.015 H

)

0.3 Fe

)

0.2 O

5.6—6.75 Al

3.5—4.5 V

This code is consistent with previous studies involving the same alloys.

PFM-alloys: Porcelain fused to metal technique.

dental alloys used were supplied by Qualident SA
(Geneva, Switzerland),

the

non-precious

alloy

by

Flamarc (Chaplan, France), the titanium by Strauman
(Saint-Imier, Switzerland), and the titanium alloy
(Ti6A14V) by Cezus (Ugine, France).

2.2. Electrolytes

Two test solutions were used:

E de-aerated artificial saliva (Fusyama and Meyer)

(23, 24) at temperature of 37°C and pH 5 with the
following composition: NaCl 0.4; KCl 0.4; CaCl)

2HO 0.795; NaHPO 0.69 and urea 1.0 g/l. The test

milieu was de-aerated with N for 24 h.

E non-de-aerated artificial saliva (Carter-Brugirard AF-

NOR/NF (French Association of Normalization) 591-
141) at a temperature of 37°C and pH 6.7 (by addition
of lactic acid) with the following composition: NaCl
0.7, KCl 1.2, NaHPO 0.26, NaHCO 1.5, KSCN

0.33 and urea 1.3 g/l. The oxygen content in the non
de-aerated medium ranges between 5.2 and 6.4 mg/l,
whereas in a de-aerated saliva, this content drops to
about 0.2 mg/l, at 37°C, as measured with the Viscolor
Oxygene Kit SA 10 (Macherey-Nagel-Du¨ren Germany).

The electrical conductibility of the salivas was measured
and found to be situated in the region of 330—340

;

10

\ )\ cm\.

2.3. Electrochemical measurements

In our investigation we used a rotating electrode tech-

nique, with a rotation speed of 500 rpm. A saturated
calomel electrode (SCE) was used as a reference. In both
experiments the potentiodynamic control of the working
electrode was provided by an EG & G PAR model 273A
potentiostat. The overall system was controlled using
a PC-compatible microcomputer. The variation of the
galvanic current with time for the studied galvanic couple

was controlled by an EG & G PAR model 273A poten-
tiostat, which was modified according to the manu-
facturer’s instructions to be used as a zero resistance
ammeter. This modification allowed measurements of the
current variation down to 100 pA. This is the same elec-
trical setup used in a previous study measuring the
galvanic current of titanium-dental alloy couple [22].
Similar assemblies have been used by other authors for
measuring galvanic cells [25—27].
We investigated a series of parameters such as:

(a) In the Fusayama—Meyer artificial saliva:

E The open circuit potential after 24 h of immer-

sion in the de-aerated electrolyte of each elec-
trode.

E Tafel slopes and calculation of the corrosion

current densities, etc., in the region $150 mV
SCE from the corrosion potential (Ecorr). For
all cases, the scan rate was 0.1 mV/s.

E Measurement of the variation of the galvanic

current and the common potential in the coup-
ling for 15 h. The quantity of current generated
by coupling was integrated. In galvanic tests the
ratio of anode surface to the cathode surface was
equal to one.

E Application of the mixed potential theory.

The surface of the polished titanium was passivated at

#

800 mV SCE for 30 min. The titanium was kept per-

manently immersed in the electrolyte for all the measure-
ments taken.

(b) In AFNOR artificial saliva:

E The open circuit potential after 24 h and 6 days

of immersion in the non-de-aerated electrolyte
of each electrode.

E Polarization curves (anodic and cathodic) were

scanned after 8 days of immersion between

!

500 and 1000 mV vs. SCE. For all cases, the

scan rate was 0.5 mV/s.

E Construction of Evans diagrams from graphs.

934

B. Grosgogeat et al. / Biomaterials 20 (1999) 933— 941

Table 2
Open-circuit potential vs. time of alloys studied

Code alloys

Fusayama saliva

AFNOR saliva

After 24 h

After 24 h

After 6 days

Potential

Ranking

Potential

Ranking

Potential

Ranking

(mV vs. SCE)

(mV vs. SCE)

(mV vs. SCE)

1

#

113

4

#

25

3

#

91

3

2

#

76

5

#

12

5

#

36

5

3

#

62

6

!

52

6

!

26

7

A

#

171

2

#

96

1

#

119

2

B

#

142

3

#

60

2

#

134

1

C

#

215

1

#

20

4

#

81

4

D

!

133

9

!

100

8

!

84

9

TI

#

11

7

!

70

7

#

21

6

Ti6Al4V

#

15

8

!

168

9

!

81

8

Passivation of the titanium was obtained by immersion
in the artificial saliva for 6 days.

Galvanic corrosion is a very complex phenomenon.

Six basic factors are involved in galvanic corrosion: (1)
potentials; (2) polarization; (3) electrode areas; (4) resis-
tance and galvanic current; (5) the electrolyte medium; (6)
aeration, diffusion and agitation of the electrolyte.

Two dissimilar metals, placed separately in the same

electrolyte, exhibit two different corrosion potentials. As
long as there is no interaction between the two metals,
the imbalance of the potentials remains unchanged.
However, as soon as an electrical contact is achieved
between these two metals, immersed in the same electro-
lyte, the system is affected and an equilibrium is reached
between the two potentials. The kinetics of this equilib-
rium may be controlled by such factors as reactions at
the surface of the electrode, resistance of the circuit,
among other. The common potential which is reached
more or less rapidly is affected by the polarization of the
electrodes. Three types of polarization and control are
known: anodic control (when the anode is much more
polarized than the cathode), cathodic control and mixed
control.

Consequently, the galvanic corrosion can be studied

using a variety of tests ranging from full-scale component
exposures to electrochemical tests.

By choosing these selected electrochemical measure-

ments we could obtain more information to explain some
basic facts concerning galvanism, and to derive data from
them which is directly related to a clinical environment.

2.4. Scanning electron microscope analysis

Some of the corroded and non-corroded surfaces were

observed by scanning electron microscopy. A JEOL JSM
6400 Instrument, was used.

2.5. Auger spectrometry analysis

Three series of titanium samples with different surface

states were analysed:
E Non-passivated surface after polishing with metallo-

graphic emery paper and finishing with 1

m diamond

dust.

E Passivated surface after 15 h in non-de-aerated AF-

NOR artificial saliva at #500 mV/SCE

E Passivated surface after 5 min in sulfuric acid (concen-

tration: 150 g/l) at 5000 mV.

The surfaces were examined using a RIBER-CMA

spectrometer in the following experimental conditions:
(a) electronic analysis: pressure in the working environ-

ment,(10

\ torr; surface area of the specimen ana-

lysed, 0.05 mm

; angle between the sample and the

analysis canon, 60°; primary energy, 2000 eV;

(b) ionic scraping: ions Xe

>; erosion speed, 1 nm/min;

angle between the sample and the scraping canon,
30°.

3. Results

3.1. Open circuit potential

The measurements obtained for open-circuit potential

in both salivas are presented in Table 2. In general, the
open-circuit potential measured in the de-aerated saliva
(Fusyama—Meyer) is higher than in the non-de-aerated
saliva (AFNOR). The dispersion of the corrosion poten-
tial

measurements

obtained

is

in

line

with

the

Traube—Wagner theory of mixed potential [28]. The
electrical potential at the metal—electrolyte surface is
closely dependent on the conditions of the experiment,
namely the nature of the electrolyte, the concentration,

B. Grosgogeat et al. / Biomaterials 20 (1999) 933— 941

935

Fig. 1. Evolution of potential in open circuit plotted against time for
alloys in AFNOR-type artificial saliva.

Fig. 2. Evolution of potential in open circuit plotted against time for
Titanium and Ti6Al4V in AFNOR-type artificial saliva.

Fig. 3. Evans diagrams constructed from polarization curves for
Ti/OC and Ti6Al4V/OC.

Fig. 4. Evans diagrams constructed from polarization curves for
Ti/OD and Ti6Al4V/OD.

the temperature, the pH, the surface state of the metal,
etc. As a result, the electrochemical reactions at the
metal—electrolyte interface vary with time. In Table 2 we
also show the ranks obtained by the 9 alloys ranging
from rank 1 (the highest positive figure measured) to rank
9 (the most negative measurement, indicating the worst
behaviour).

Figures 1 and 2 show the variations of the open-circuit

potential measurements over 6 days for the 9 alloys
studied.

3.2. Potentiodynamic polarization curves

The polarization curves were obtained in pseudo-

stable conditions in order to allow us to construct the
Evans diagrams. This explains the 6-day immersion in
the AFNOR saliva and the slow scanning speed. Fig-
ures 3 and 4 show the Evans diagrams for the Ti/alloy
CC, Ti6A14V/alloy CC and Ti/CoCr, Ti6A14V/CoCr.

3.3. Galvanic current predicted after the mixed
potential theory

This theory is based on two hypotheses: (a) any electro-

chemical reaction can be divided into two or more oxida-
tion or reduction reactions, and (b) there can be no net
accumulation of electrical charges during an electro-
chemical reaction. When two different corroding alloys
are coupled electrically in the same electrolyte, both
alloys are polarized so that each corrodes at a new rate
[29, 30]. In practice, the application of the mixed poten-
tial theory allows us to trace the Tafel lines for each alloy,
and to sum the measurements of the anode and cathode
lines of the couple under investigation [30]. In this way,
two new Tafel lines are obtained and their intersection
provides the values of E and i (Table 3).
3.4. Galvanic currents

Direct measurement of galvanic couples are shown in

Figs. 5 and 6. The galvanic currents remain weak (to the

936

B. Grosgogeat et al. / Biomaterials 20 (1999) 933— 941

Table 3
Comparative analysis of results obtained using two different measurement techniques and two different artificial salivas for galvanic couples (Ti/dental
alloys and Ti6A14V/dental alloys)

Couples

E (mV vs. SCE)

i (nA/cm)

Potential

Evans

Potential

Evans

mixed theory

diagrams

mixed theory

diagrams

Fusayama saliva

AFNOR saliva

Fusayama saliva

AFNOR saliva

¹

i/Alloys

1

#

78

#

37

222

600

2

#

80

#

12

303

350

3

#

79

!

60

193

250

A

#

67

#

15

83

350

B

#

82

#

40

241

660

C

#

86

#

25

99

530

D

!

170

!

10

198

430

¹

i6Al4V/Alloys

1

#

90

#

43

1196

540

2

#

94

#

12

2022

390

3

#

101

!

58

922

80

A

#

70

#

7

81

350

B

#

93

#

45

1334

1860

C

#

79

#

34

113

460

D

!

165

!

97

279

10

Fig. 5. Measurement of galvanic current for Ti/dental alloy couples.

Fig. 6. Measurement of galvanic current for Ti6Al4V/dental alloy
couples.

order of nA). For the positive values of the galvanic
current, the dental alloys are cathodic in relation to their
coupled Ti or Ti6A14V, so there is little or no corrosion
of the alloy. On the other hand, when the current is
negative, the dental alloys are anodic, and more cor-
rosion is observed. Quantitative assessment of the gal-
vanic couple is shown by digital integration of i galv and
time function in Table 4.

3.5. Auger spectrometry

The study of surface oxidation on titanium was carried

out in two stages:
(a) The first stage consisted of measuring the erosion

speed of the layer of titanium oxide formed by
anodisation at 5 V in sulfuric acid. The thickness of

the oxidized layer was calculated at between 1.15 and
1.25 nm. In these experimental conditions, the speed
at which the oxidized layer is formed is approxi-
mately 2.3 nm/V [31]. The profiles of the concentra-
tions concerning titanium, oxygen and carbon in the
anodized sample were plotted against erosion time.
A swift decrease in the carbon content was noted. The
presence of carbon is generally attributed to the sur-
face pollution caused by the use of organic solvents to
clean the surfaces of the samples.

The profiles of oxygen and titanium concentra-

tions have an inverted relationship: whereas the
curve of oxygen peaks diminishes with the depth of
the erosion, the curve of titanium peaks increases
until it flattens out when the heart of the metal is
reached.

B. Grosgogeat et al. / Biomaterials 20 (1999) 933— 941

937

Table 4
Galvanic behaviour and cumulative charges in Fusayama saliva of
couples studied

Code alloy

Couples Ti/Alloy

Couples Ti6Al4V/Alloy

(

lC/cm)

(

lC/cm h)

(

lC/cm)

(

lC/cm/h)

1

59.9

4.0

17.20

1.2

2

67.6

4.5

436.7

29.1

3

42.6

2.8

124.8

8.3

A

59.6

3.9

335.3

22.4

B

57.8

3.8

225.9

15.1

C

59.7

4.0

296.3

19.8

D

!

116.7

!

11.1

!

5541

!

46.4

Fig. 7. Profile of the composition of the titanium oxide layer before the
corrosion test.

Fig. 8. Profile of the composition of the titanium oxide layer after the
corrosion test in AFNOR-type artificial saliva.

The length of time necessary for the erosion of the

oxidized layer was obtained by the Tangential Plot-
ting method, and was found to be 830$30 s. The
resulting erosion speed thus obtained was 1 nm/min.

(b) The second stage consisted in the analysis of the

composition of the titanium oxide layers before and
after the corrosion tests. The analysis of the initial
surface of the titanium revealed the presence of oxy-
gen, carbon, and titanium. The presence of titanium
can be attributed to surface pollution. The carbon
content decreases swiftly after the first manometers
(Fig. 7).

The heart of the metal was reached after 190$10 s of

erosion, which corresponds to a thickness of approxi-
mately 3 nm. The analysis of the composition of the
oxidized layer after the corrosion test revealed the pres-
ence of a phosphorus. The presence of phosphorus is due
to the composition of the artificial saliva which contains
phosphate ions. The concentration of carbon decreases
slowly (Fig. 8). There must be adsorption of urea or of
carbonate ions on the surface of the sample. The titanium
content is stable after 520 s, corresponding to an oxidized

layer approximately 8 nm thick. The oxygen content was
small both on the surface and at the heart of the sample,
and the highest reading obtained was after 250 s of ero-
sion time. This form of evolution is probably linked to
the adsorption of urea or of phosphate in the oxidized
layer.

4. Discussion

Open-circuit potential measurement over time leads to

the conclusion that an immersion time of two days is
necessary for a pseudo-stationary state to be reached. In
order to study the electrochemical behaviour of titanium
in conditions close to those of clinical reality, it is abso-
lutely necessary to have oxidized layers characteristic of
a pseudo-stationary state. Hoar [11] and Fraker [31]
showed that the passivation of the titanium oxide layer is
closely dependent on the length of immersion, and that it
represents an ageing process. We integrated this using
two passivation techniques: anodic polarization and im-
mersion.

The comparison of measurements obtained in different

experimental conditions show that
E The electric potential of the silver-based alloy C3 is

negative in AFNOR (French Association of Normal-
ization) — type artificial saliva, whereas it is positive in
Fusayama-type artificial saliva. This can be explained
by the formation of silver chloride on the surface of the
alloy insoluble in Fusayama artificial saliva. In this
case the silver chloride passivates the surface of the
alloy. On the other hand, in AFNOR artificial saliva
the silver chloride probably forms the soluble silver
complex due to the presence of thiocyanate ions and
no passivation occurs.

E In both artificial salivas the open-circuit measure-

ments of titanium are very similar, in spite of the fact
that the passivation techniques are different. A further
observation can be made: 24 h immersion of titanium

938

B. Grosgogeat et al. / Biomaterials 20 (1999) 933— 941

Fig. 9. Anodic control of Ti6Al4V/O2 couple in Fusayama-type artifi-
cial saliva.

Fig. 10. Mixed control of Ti/OB couple in Fusayama-type saliva.

is insufficient for a pseudo-stationary state to be
achieved.

E The open-circuit measurement of potential for

Ti6Al4V in AFNOR-type artificial saliva remains
negative after 6 days of immersion, whereas anodic
passivation brings Ti6Al4V into the cathodic zone,

#

15 mV vs. SCE. In AFNOR-type artificial saliva, the

6 days of immersion are probably insufficient for
a positive potential to be obtained.

E The open-circuit measurements of potential in

Fusayama-type artificial saliva are higher than those
obtained in AFNOR-type artificial saliva.

E The classification of the alloys used in Table 1 shows

that, except for the Pd-based alloy

CC, the alloys

occupy approximately the same positions. Thus the
two artificial salivas used produce the same non-
equivocal data for the interpretation of the results.
Consequently, it can be stated that the electrochemical
behaviour of the alloys does not vary even if they are
tested in different artificial salivas.

Similarly, although the artificial salivas are different, the
densities of galvanic corrosion currents are low for all the
couples studied (Table 3), whether they are obtained by
Evans diagrams based on graphs or by using the mixed
potential theory. They are in the region of a few hundred
nanoamperes.

The results obtained by the theory of mixed potentials,

by direct measurement of galvanic couple (Figs. 5 and 6),
or by the quantities of cumulated charges used during the
whole duration of the test (Table 4) show that the
measurements obtained for the Ti6Al4V/alloy couples
are systematically higher. We also calculated a mean
value per hour for the charge cumulated for each couple
studied over a period of 15 h (Table 4). The values ob-
tained are in agreement with the overall results.

As we saw above, the Ti or Ti6Al4V couples with

dental alloys introduce a new series of values for the
mixed potential. The galvanic battery created in this
way functions under anodic, cathodic (Fig. 9) or mixed
(Fig. 10) control, according to the potential.

When the control of the cells was anodic, anodic polar-

ization was predominant. The intensity of the galvanic
corrosion process depends on the polarization of the
anode, and particularly on the variations of the state of
the surface of the anode. When the control was of a mixed
nature, the intensity of the galvanic corrosion process
was due to the common contribution of the polarization
phenomena.

The surfaces of the alloys, before and after the cor-

rosion tests were observed by scanning electron micro-
scopy. On the whole, the examination of the surfaces of
the dental alloys revealed no damage due to the cor-
rosion process (Fig. 11a and b) but it did produce evi-
dence of oxidation of the Ti6Al4V-based implants.

After Souto and Burstein [32] the Ti6Al4V alloy

shows transient microscopic breakdown of the passive

state induced by the presence of chloride ions. In anodic
polarization of the Ti6Al4 V breakdown event are ob-
served as small currents. These breakdown events involve
localized depassivation of the passive surface followed by
repassivation. After polarization tests Mozhi et al. [33]
by the transmission electron microscopy technic (TEM)
examined the microstructures in these alloys. The various
precipitate phases observed, play an important role in the
passivation of Ti—Al alloys. In the Al—3%Ti—3%V alloy
ternary intermetallics of different morphologies are seen
along with AlTi intermetallic particles. The conti-

nuous anodic dissolution of these precipitates is believed
to contribute to the smooth increase in the passive cur-
rent with the change of morphologies of [the] surface. In
our case the precious alloys remain in their immunity
range and will therefore acts as cathodes, while Ti6Al4V
will be under anodic control. Consequently, the invest-
igation of Ti6Al4V after galvanic coupling tests con-
firmed the breakdown event observed by Souto and
Burstein.

The theoretical variation of the oxidized layer after the

corrosion test was calculated using Faraday’s Law. There
was a reading of 10

\ coulombs, which corresponds to

B. Grosgogeat et al. / Biomaterials 20 (1999) 933— 941

939

Fig. 11. Ti6Al4V before (a) and after (b) coupling test in AFNOR-type saliva.

an increase of approximately 8 nm. This measurement
confirms the results obtained by Auger spectrometry.

5. Conclusion

All the results obtained, regardless of the electrolytic

environment or the techniques used (electrochemical,
scanning electron microscopy, and Auger spectrometry)
show that the intensity of the corrosion process is low in
the case of the Ti/dental alloys and Ti6Al4V/dental
alloys.

In our study, the anodic and cathodic parts had the

same surface area, whereas in vivo, the surface areas can
be considerably different, possibly modifying the inten-
sity of the galvanic corrosion current. The most unfavor-
able situation is when a small anode is linked to a large
cathode.

Other possible types of corrosion are to be considered

in addition to galvanic corrosion, for example pitting
corrosion and crevice corrosion. These types of corrosion
are specific to each alloy and are to be taken into account
in our clinical choices. Moreover, the biological charac-
teristics of each individual represent a variable which
cannot be reproduced easily in vitro (composition and
acidity of saliva, dental hygiene, eating habits, adminis-
tration of medicine & c). For this reason, the most favor-
able suprastructure/implant couple is the one which is

capable of resisting the most extreme conditions that
could possibly be encountered in the mouth.

References

[1] Schutz RW. Titanium. Process industries corrosion—The theory

and practice. National Association of Corrosion Engineers.
1986:503—27.

[2] Schutz RW, Thomas DE. Corrosion of titanium and titanium

alloys. Metals Handbook, 9th edn. Metals Park, Ohio: ASM
International, 1978, vol. 13, Corrosion, 669—706.

[3] Hille GH. Titanium for surgical implants. J Mat 1966;1:373—83.
[4] Rosenberg HW. Titanium: properties. In: Michael BB, editor.

Encyclopedia of materials science and engineering, vol 7. Oxford:
Pergamon Press, 1986:5109—10.

[5] ITO Yoshimasa. Titanium and its applications. Proc. 2nd Int

Symp on Titanium in Dentistry, Kyoto. 18—19 January 1992:27—9.

[6] Pesch HJ, Glueckert K, Tuemmler HP, Thull R. Tissue reactions

of titanium and its alloys. Titanium science and technology. In:
Lu¨tjering G, Zwicker U, Bunk W, editors. Proc 5th Int Conf on
Titanium, vol 2. Munich, 10—14 September 1984:1387—92,
Oberusel, Deutsche Gesellschaft fu¨r Metalkunde EV, 1985.

[7] Branemark PI, Hansson BO, Adell R, Breine U, Lindstrom J,

Hallen O, Ohman A. Osseointegrated implants in the treatment
of the edentulous jaw. Experience from a 10-year period. Scand
J Plast Reconstr Surg 1977; 11 (Suppl. 16) (and as a monograph
from Almqvist & Wiksell International, Stockholm).

[8] Branemark PI, Adell R, Albrektsson T, Carlsson G, Haraldson T,

Lekholm U, Lindkvist L, Lindstrom J, Lundkvist S, Rockler B.
Osseointegrated titanium implants in the rehabilitation of the
edentulous patient. Adv Biomaterials 1982;4:133—41.

940

B. Grosgogeat et al. / Biomaterials 20 (1999) 933— 941

[9] Adell R, Lekholm U, Rocker B, Branemark PI. A 15 year study of

osseointegrated implants in the treatment of the edentulous jaw.
Int J Oral Surg 1981;6:387—416.

[10] Ericsson I, Lekholm U, Branemark PI, Lindhe J, Glantz PO,

Nyman S. A clinical evaluation of fixed bridge restorations sup-
ported by the combination of teeth and osseointegrated titanium
implants. J Clin Periodontol 1986;13:307—12.

[11] Hoar TP, Mears DC. Corrosion-reistant alloys in chloride solu-

tion: materials for surgical implants. Proc R Soc (London) 1966;
SCRA294 (1439):486—510.

[12] Revie RW, Greene. Corrosion behaviour of surgical implant ma-

terials: I. Effects of sterilization. Corros Sci 1969;9:755—62.

[13] Revie RW, Greene. Corrosion behaviour of surgical implant ma-

terials: II. Effects of surface preparation. Corros Sci 1969;9:
763—70.

[14] Mueller HJ, Greener HE. Polarization studies of surgical mate-

rials in Ringer’s solution. J Biomed Mater Res 1970;4:29—41.

[15] Fraker AC, Ruff AW, Yeager MP. Corrosion of titanium alloys in

physiological solutions. Proc 2nd Int Titanium Conf., New York:
Plenumm press, 1973:2447—57.

[16] Solar JR. Corrosion resistance of titanium surgical implant alloys:

a review. In: Syrett BC, Acharya A, editors. Corrosion and degra-
dation of implant materials, ASTM STP 684. American Society
for Testing Materials, 1979:259—73.

[17] Solar JR, Pollack SR, Korostoff E. Titanium release from im-

plants: a proposed mechanism. In: Syrett BC, Acharya A, editors.
Corrosion and degradation of implant materials, ASTM STP 684.
American Society for Testing Materials, 1979:161—72.

[18] Solar JR, Pollack SR, Korostoff E. In vitro corrosion testing of

titanium surgical implant alloys: an approach to understanding
titanium release from implants. J Biomed Mater Res 1979;
13:217—50.

[19] Kruger J. Fundamental aspects of the corrosion of metallic im-

plants In: Syrett BC, Acharya A, editors. Corrosion and degrada-
tion of implant materials, STP 684. American Society for Testing
Materials, 1979:107—27.

[20] Schutz RW, Covington LC. Guidelines for corrosion testing of

titanium In: Kleefisch EW, editor. Industrial application of tita-
nium and zirconium, ASTM STP 728. American Society for
Testing and Materials, 1981:59—70.

[21] Lucas LC, Lemons JE, Lee J, Dale P. In vitro corrosion charac-

teristics of Co—Cr—Mo/Ti—6Al—4V/Ti alloys. Quantitative charac-

terization and performance of porous implants for hard tissue
applications, STP 953. American Society for Testing and Mate-
rials, 1987.

[22] Reclaru L, Meyer JM. Study of corrosion between a titanium

implant and dental alloys. J Dent 1994;22:159—68.

[23] Fusayama T, Katayori T, Nomoto S. Corrosion of gold and

amalgam placed in contact with each other. J Dent Res 1963;
42:1183—97.

[24] Meyer JM, Nally JN. Influence of artificial salivas on the cor-

rosion of dental alloys. J Dent Res 1975;54:678 (abstr. No. 76).

[25] Mansfeld F, Kenkel JV. Laboratory studies of galvanic corrosion

of aluminium alloys. In: Baboian R, France Jr WD, Roew Lc,
Rynewicz JF, editors. Galvanic and Pitting Corrosion—Field and
Laboratory Studies, ASTM STP 576. Philadelphia: American
Society for Testing and Materials, 1976:20—47.

[26] Mansfeld F, Myers RL, Lauer G. Instrumentation for corrosion

studies. Low conductivity media. 1973, Technical Report, Rock-
well International.

[27] Baboian R. Electrochemical techniques for predicting galvanic

corrosion. In: Baboian R, France Jr WD, Roew Lc, Rynewicz JF,
editors. Galvanic and Pitting Corrosion—Field and Laboratory
Studies, ASTM STP 576. Philadelphia: American Society for
Testing and Materials, 1976:20—47.

[28] Wagner C, Traud W. Uber die Deutung von Korrosionsvorga¨n-

gen durch Uberlagerung von elektrochemischen Teillvora¨ngen
und u¨ber die Potentialbildung an Mischelektroden. Z Electro-
chem 1938;44:391—454.

[29] American Society for Testing and Materials (ASTM). Standard

guide for development and use of galvanic series for predicting
galvanic corrosion performance, vol 03.02. 1991:334—40.

[30] Dalard F, Montella C, Sohm JC. Adhe´sion des peintures sur le

titane anodise´. Etude des films d’oxyde. Surf Techn 1976;4:367—80.

[31] Fraker AC, Ruff WA, Sung P, Van Orden AC, Speck KM. Surface

preparation and corrosion of titanium alloys for surgical im-
plants. In: Luckey HA, Fred Kubli Jr, editors. ASTM STP 796.
American Society for Testing and Materials, 1983:206—19.

[32] Souto RM, Burstein GT. A preliminary investigation into micro-

scopic depassivation of passive titanium implant materials in
vitro. J Mater Sci Mater Med 1996;7:337—43.

[33] Mozhi TA, Jha SC, Ray R. Technical note: Corrosion behavior of

rapidly solidified Al—Ti base alloys in chloride environment. Corr
Sci 1989;4:811—3.

B. Grosgogeat et al. / Biomaterials 20 (1999) 933— 941

941