Galvanic corrosion occurs when dissimilar alloys are placed in direct contact within the oral cavity or within the tissues. When saliva penetrates into prosthetic components in contact with implants, the metal dissolution generates currents, due to a potential difference created by the formation of a galvanic cell (Figure 1
). In the case of dental implants, complicated electrochemical processes related to implant and suprastructure are linked to galvanic corrosion [1
] which leads to a clinically relevant situation due to two main reasons: the biological effects that may result from the dissolution of alloy components, and the bone destruction caused by the current flow that results from galvanic coupling.
The alloy in the couple that corrodes would be the less noble or more active alloy. Coupling could result in an electropositive local environment along the implant interface, which could directly influence bone resorption. Hence, these galvanic couplings should be avoided [3
Olmedo et al. (2003) [4
] have observed that ionic release induced by corrosion could be responsible for peri-implantitis and treatment failure. Particles dissolved as a result of corrosion process are phagocytosized by macrophages (host response) and release mediators of inflammation in the form of cytokines (host defense), which inhibit the production of osteoblasts and promote osteolytic activity, leading to implant loosening. They found macrophages loaded with metal particles as indicators of the corrosion process in the soft peri-implant tissue of failed human dental implants.
The corrosion process may limit the metal’s resistance to fatigue, which may eventually cause the fracture of the implant. Guglielmotti and Cabrini (1997) [5
] found metal particles included in the osseointegrated bone tissue of implants that failed due to metal fatigue, providing evidence of the corrosion of metallic structure.
The corrosion products can be distributed throughout the entire body, and may even cause allergic reactions (generally of the delayed type; Type IV) or a hypersensitivity reaction [2
], with release of inflammatory mediators (cytokines). It has not been proved yet whether hypersensitivity to metal is the cause of implant failure or vice versa [6
]. Metals like nickel, cobalt, beryllium, gold and palladium rank high in the allergy hit lists. Nickel leads the position on this list [7
]. Therefore, the ultimate goal must be to use only those alloys with minimal ion release.
Commercially pure titanium (c.p. Ti) and its alloys have been widely used for dental implants due to their mechanical properties, good corrosion resistance in biological fluids and biocompatibility [8
]. Although is widely accepted that titanium alloys are good materials for endo-osseous implantation [9
], the choice of a suitable alloy for the suprastructure still remains an open question. Gold alloys are employed for suprastructures because of their excellent biocompatibility, corrosion resistance and mechanical properties. However, the increasing cost of precious alloys used in restorative dentistry has led to the development of low cost metallic materials for dental prostheses [3
]. Major components of these alloys include nickel, cobalt or titanium. These alloys have good mechanical properties and are cost-effective, but their biocompatibility and corrosion resistance are of concern. It has been reported that failures of some implants was due to corrosion [4
]. Thus, the design of suprastructures has to be made considering the corrosion resistance when the alloy is coupled to titanium, the biocompatibility, and the clinical studies of the relationship between the metal and the surrounding tissues (epithelium, connective tissue, and bone tissue) [10
There are several factors such as the presence of fluoride ion, difference in oxygen concentration, dental plaque, microorganisms and mechanical stress that could increase further the corrosion rate, and should be also taken into account [11
An overview of the existing knowledge on the galvanic corrosion mechanism in oral environment has been carried out through a literature search shown in Table 1
, which summarizes the published studies on galvanic corrosion of dental implants/suprastructures.
Generally, the potential of the anode and the cathode is measured using a three-electrode configuration [10
] and the galvanic effect quantified through the mixed potential theory. However, this approach has some limitations due to the initial hypothesis required for the application of the mixed potential theory [19
]. Indeed, Al-Ali et al. [14
] simulated the galvanic couple by welding different materials and they observed that the potential and current density of the welded couple did not correspond with the values obtained by the simple combination of the potentiodyinamic curves of the individual material. These values do not agree with experimentally obtained ones because the crevice corrosion is not taken into consideration in the simplified mixed theory, so it cannot be applied. Significant differences in corrosion behaviour were observed between welding joint and adhesive joint, finding by adhesive joining method higher values of coupled ICORR
than those for uncoupled alloys, and crevice associated with adhesive joint made the galvanic couple more corrosive than welding joint. Ciszewski et al. [15
] also observed that the metal ion release determined through the galvanic current values obtained from accelerated electrochemical tests and applying the mixed potential theory were different from the direct measurements (adsorptive stripping voltammetry) of the metal ion release in the same couples. Indeed, metal ions from the cathode of the pair were also detected after 30 days of immersion.
To avoid the limitations of the mixed potential theory, specially obtained when studying passive materials, the galvanic effect was directly measured through a zero-resistance-ammeter (ZRA) [1
]. Some other authors also tested the system superstructure/implant [18
] and the measure of the corrosion damage was directly measured by quantifying the metal ion release after immersion tests of superstructures and implants couples by different analytical techniques such as the Inductively Coupled Plasma-Atomic Emission Spectrometry [23
]. Yamazoe et al. [23
] observed that titanium release in titanium/dental alloy is highly influenced by the titanium microstructure and in a titanium/titanium pair the metal ion release increases when differences in microstructure also increased.
In vivo studies suggest that polymetallism leads to corrosion process. Foti et al. [24
] showed in animal models after 2 months, that the presence of a precious alloy superstructure leads to titanium migration towards the area around the cervical region of the implant, but without apparent modification of osseointegration, whereas with titanium superstructures this phenomenon was not observed. In humans, a wide range of galvanic currents resulted from electrical contact of metallic restorations in mouth has been noted. These currents are influenced by the area ratio, the total surface area of the galvanic couple, the particular conditions of each individual, chewing …. and the consequences were observed to vary depending on the location in the mouth, due to the degree of oral mucosa keratinization among others [25
The aim of this study is to analyse the electrochemical behaviour in artificial saliva (AS) of cobalt-chromium alloys (CoCr), nickel-chromium-titanium (NiCrTi), gold-palladium alloy (Au) and titanium alloy (Ti6Al4V), used in the manufacture of the implant prosthesis structures and to evaluate the galvanic effect produced by the contact (physical or electrical through the electrolyte) of those with the commercially pure titanium grade 2 (TiG2). The influence of fluorides present in artificial saliva on the electrochemical processes and galvanic corrosion between dental implants and suprastructures are also studied.
The conclusions obtained were:
Passive dissolution is the main corrosion mechanism of titanium and CoCr alloys and the corresponding passive dissolution rate was not accelerated by the presence of fluorides. NiCrTi alloy is the less corrosion resistant alloy among the studied materials in artificial saliva and fluorides critically accelerate its corrosion rate due to the susceptibility of nickel towards fluorides.
Titanium alloys starts actively dissolving when the pH of the artificial saliva is below 3 and the fluoride content is 1000 ppm. Under these conditions, HF concentration is sufficiently high to form soluble titanium complexes in all studied titanium alloys.
Measurement of galvanic corrosion of TiG2 implant coupled to different materials can only be carried out by zero-resistance ammeter and the direct measurement of the galvanic current and potential due to the passive nature of the biomedical alloys (CoCr, NiTiCr and Ti6Al4V).
Acceleration of corrosion due to galvanic effects was only observed between titanium alloys and CoCr suprastructures in fluoride-containing acidic solutions. This galvanic effect is highly dependent on the solution chemistry and the coupled material, increasing when the suprastructure is Ti6Al4V.
2.1. Open Circuit Potential (OCP)
OCP values of the tested materials were measured for 30 min in both electrolytes (Figure 2
a,b). In all cases an abrupt increase of OCP was measured at the beginning of the OCP measurement due to the spontaneous formation of the passive films. The OCP stabilizes after 600 s and constant values were obtained at the end of the test. Those values are reported in Table 2
The CoCr-c alloy exhibited the most negative potential (−611 mV) in AS while Au showed the highest one (121 mV). The other materials had intermediate OCP values, between −309 mV and −203 mV. In the fluoride content solution, Figure 2
b, titanium alloys showed an intensive OCP decrease at around −1000 mV while the other tested materials established at OCP similar than the values in AS. Only the CoCr-c experienced an increase in OCP when immersed in ASF−
2.2. Potentiodynamic Curves
shows the potentiodynamic curves of the studied materials in AS, ASF−
pH3. In general, all polarization curves can be divided in four potential domains. In the cathodic domain the current density is due to the reduction of water and partially dissolved oxygen. The transition domain from cathodic to anodic current takes place at the corrosion potential (Ecorr
), at which the corrosion current density (icorr
) is obtained. The third domain corresponds to the passive zone due to oxide film formation on the metal surface and where the current density remains stable at the passive current density (ip
). Finally, the tanspassive domain, corresponds to the zone comprised above the breakdown potential (Eb
), characterized by the increase in current due to the dissolution of the protective oxide film as well as water oxidation. Only titanium alloys do not show this transpassive region.
Electrochemical parameters obtained from the potentiodynamic curves for each material in electrolytes AS and ASF−
pH3 are summarized in Table 3
All the tested materials are metals and alloys spontaneously passivated in AS since their OCPs lied within their respective passive domains. In the acidic AS fluoride content solution (ASF−pH3) titanium alloys showed an active dissolution mechanism, thus presenting a significant decrease in the OCP and Ecorr, and an increase in icorr and ip.
In AS, CoCr and CoCr-c alloys showed the lowest Ecorr values (−738 mV) in good agreement with the lowest OCP values reported in the previous section. However, they presented a large passive domain. For potential values above 700 mV an increase in current density took place and began the dissolution of the oxide layer (chromium oxide). The NiCrTi alloy (Ecorr −295 mV) shows higher Ecorr than the CoCr alloys, but its passive domain is very narrow, showing transpassive dissolution at very low potential value (+158.50 mV) compared to the other materials. Au showed the highest Ecorr and TiG2 and Ti6Al4V the broadest passive domain.
The icorr of CoCr and NiCrTi are one order of magnitude higher than those obtained for the titanium and titanium alloy in AS, while in ASF−pH3, the NiCrTi and CoCr-c showed an increase of their icorr, while both CoCr showed a decrease in its corrosion rate. Analogously, the ip of the NiCrTi and CoCr-c in ASF−pH3 increases two orders of magnitude with respect to the values measured in AS. The CoCr and Au did not show any influence of the fluorides in its ip (passive dissolution rate) although a decrease in Eb was observed.
b) the potentiodynamic curves of the studied materials are similar to those obtained in AS (Figure 3
a), but produced a potential reduction in the passive domain of all materials and a slight increase in the current densities.
2.3. Zero Resistance Ammeter
, Figure 5
and Figure 6
show the average values at the end of every hour of immersion of the galvanic current and potential of the pairs TiG2/CoCr (Figure 4
), TiG2/CoCr-c (Figure 5
) and TiG2/Ti6Al4V (Figure 6
) measured by Zero-Resistance ammeter.
The TiG2 coupled to both CoCr alloys in AS does not show any acceleration of its corrosion as a consequence of their galvanic coupling. Galvanic currents are very low and galvanic potentials remained constant around −200 and −400 mV for the TiG2/CoCr and TiG2/CoCr-c respectively. In the ASF−pH3 solution, the galvanic potentials decreased with respect to the values measured in AS, thus increasing the galvanic current, thus increasing the corrosion rate of the anodic member of the pair, the TiG2. Slightly higher galvanic current were obtained for the TiG2 in the TiG2/CoCr-c pair.
The TiG2/Ti6Al4V couple does not show any galvanic effect in AS but a huge increase in the corrosion rates in ASF−pH3. In the fluoride containing solution there are changes in the polarity of the pair depending on the immersion time. During the first hour and after 4 h of immersion Ti6Al4V alloy acted as the anode while it behaved as cathode between 2 and 4 h of immersion. The average galvanic currents of the pair were one order of magnitude than those measured in the other studied couples. Independently of the polarity of the pair, the galvanic potential remained constant at around −1000 mV.
4. Materials and Methods
4.1. Electrolytes and Materials
The solution employed to carry out the electrochemical tests is based on the the Fusayama artificial saliva (AS) which has the following composition: 0.4 g NaCl, 0.4 g KCl, 0.6 g CaCl2, 0.58 g Na2HPO4, 1 g urea and distilled water until 1 L. The pH of the AS solution was 6.5. Fluorides, 2.21 g/L of NaF (1000 ppm of fluorides), were added to the AS in order to analyse their influence on the corrosion behaviour of the metallic alloys (ASF−). The fluoride containing solution was prepared at pH 6.5 (ASF−) and pH 3 (ASF−pH3). Temperature of the solution was kept at 37 °C.
The selected alloys were provided by LAFITT (Valencia, Spain) and they correspond to materials commercially used for implants and suprastructures. They include two CoCr alloys, a NiCrTi alloy, a titanium alloy, one pure titanium grade 2, and a gold alloy. The chemical composition of the materials is given in Table 3
(the nomenclature of the materials used in the paper is given in brackets).
The materials were provided in form of metallic cylinders, which were embedded into non-conductive resin for the electrochemical tests thus leaving a working area of 0.28 cm2 in contact with the solution in the case of CoCr-c and NiCrTi alloys and of 0.5 cm2 for the rest of alloys.
Before each experiment, the samples were mechanically polished (500, 1000 and 4000 grit SiC paper), degreased with acetone, washed with pure water and dried with compressed air before use.
4.2. Electrochemical Equipment and Experiments
A double-wall three-electrode cell (volume 50 mL) in aerated conditions was used for all the electrochemical measurements. An Ag/AgCl 3 M KCl reference electrode and a platinum wire as counter electrode were used. All potentials were referred to the Ag/AgCl 3 M KCl electrode (205 mV versus SHE). All electrochemical measurements were carried out using a potentiostat Solartron 1287 (Solartron Group Ltd., Solatron Enterpraises, Torrance, CA, USA). Open Circuit-Potential (OCP) was measured for 30 min. The potentiodynamic polarization tests were performed by scanning the applied potential from −1200 mV and moved in the anodic direction to 1500 mV at a scan rate of 1 m Vs−1. The corrosion potential Ecorr as well as the corrosion current density icorr were automatically extracted from the polarization curves by the CorrView Version 3.0 software through Tafel slope extrapolation. Passivation current density (ip) and breakdown potential (Eb) at a current density of 100 µA/cm2 were also obtained from the polarization curves.
Galvanic current and potential was measured as a function of immersion time using a zero-resistance ammeter (ZRA). TiG2 was coupled to the Ti6Al4V and both CoCr alloys and connected to the Solartron 1287 potentiostat used as a ZRA. The galvanic current and potential established between the pairs were measured every 0.5 s during several hours depending on the couple. The same reference electrode described previously was used. The current sign was positive when the direction of the electrons was from WE1 to WE2, thus WE1 was corroding. Current values were negative when the electrons flowed in the opposite direction, that is, the WE2 was corroding. The assays were designed with TiG2 as WE1 and the other alloys as WE2.
The reproducibility of the measurements was determined through three repetitions of each test.