Introduction

Although titanium (Ti) and Ti-based alloys are the most widespread materials for dental implant fabrication and used with a successful rate [1, 2], two main issues are of particular concern: (1) a mismatch in Young’s modulus (E) between the bone and implants and (2) insufficient surface bioactivity of implants [2]. The composite, namely the one with a Ti+17 vol% Mg composition, was introduced as a promising candidate in our previous works [3,4,5,6] to minimize the above-mentioned drawbacks of currently used Ti-based materials. Its Ti matrix assures required mechanical properties under static and dynamic loading, and the biodegradable Mg component leads to a reduction in E and promotes bone deposition and osseointegration [7,8,9]. Mechanical treatment by grinding and polishing was applied for the surface finish of the Ti+17 vol% Mg composite of a cylindrical design, as was demonstrated in [6]. The smooth and less strained surface resulted in a pronounced response represented by a tolerated degradation rate and a distinct in vitro biological behavior [6]. Moreover, it eliminated the need for further surface treatments, e.g., sandblasting, chemical etching, etc., and concurrently maintained the bioactive surface nature when compared with the established Ti-based materials.

Although most orthopedic implants are susceptible to a variety of corrosion processes, galvanic corrosion is of particular concern. The galvanic corrosion occurs when two (or more) dissimilar metals are coupled in a corrosive medium. In the case of dental implants, when saliva penetrates prosthetic components, the less noble metal corrodes at an accelerated rate, resulting in a flow of electric current. This current flows through the surrounding tissues and thus causes some pain and local inflammation [10,11,12]. The overall rate of electrochemical processes involved in a dental implant structure is mainly linked to the formation of a base film at the interface and the transport of either metal or oxygen through the formed film [13]. Mg is prone to corrosion in aqueous solutions because of its high chemical activity, i.e., a low standard electrode potential (− 2.356 V) [14]. Mg and its alloys were used in conjunction with different components, such as aluminum, steel, zinc, carbon fiber, and copper [14,15,16,17,18,19]. In specific environments, such materials exhibited an accelerated rate of Mg degradation due to the galvanic coupling. For the Ti+17 vol% Mg composite material, the contact between Mg and Ti is susceptible to the galvanic corrosion in the oral environment due to a high difference in their standard electrode potential (− 2.356 vs. − 0.163, respectively) [20, 21]. Therefore, it is inevitable to study the galvanic corrosion behavior and mechanism of the proposed material.

Several techniques were used to investigate the galvanic corrosion of dental implant materials, such as potentiodynamic polarization (PDP) and electrochemical impedance spectroscopy (EIS) [22, 23]. The PDP measurements give information about the corrosion potential (Ecorr), which expresses the thermodynamic stability of the studied systems and the resistance of materials against corrosion [24, 25]. The corrosion kinetics are linked to the corrosion rate (RCorr), which can be calculated from the corrosion current density (icorr). But EIS helps in the evaluation of corrosion mechanism, corrosion resistance, and surface passivity [25,26,27,28,29]. It measures the resistance of a metal/electrolyte interface over a wide range of frequencies (from 100 mHz to 100 kHz). Moreover, it precisely characterizes the systems formed by an array of layers with different properties by employing electrical equivalent models. These models allow a correlation between the applied electrical elements and processes that may occur in the studied systems.

The purpose of this study was the evaluation of electrochemical (galvanic) corrosion properties of a dental implant (DI) fabricated from the newly designed Ti+17 vol% Mg composite material (further labeled as TiMg). Corrosion properties of DI were examined using the open-circuit potential (OCP), PDP, and EIS measurements to understand the corrosion behavior and mechanism in a simulated physiological environment. The present study was performed using a commercial design MV3.6-10 DI in Hank’s balanced salt solution (HBSS) in line with the ASTM G5-14 and ISO 10993-15 guidelines to mimic the performance upon the desired application. The obtained results were directly compared with those of the TiMg composite of a cylindrical design to evaluate how various surface characteristics, e.g., geometry, roughness, and Mg forms, influenced the degradation of Mg in the test solution.

Experimental

Materials

The TiMg composite was fabricated from plasma atomized (PA) Ti Grade 1 (AP&C, Canada) and gas atomized (GA) Mg (IMR Metalle, Austria) powders by a powder metallurgy approach, which consisted of blending, cold isostatic pressing (CIP) at a pressure of 200 MPa, uniaxial hot vacuum pressing (HVP) at a pressure of 500 MPa, a vacuum of 8 Pa, and a temperature of 410 °C and eventually direct extrusion (DE) at a reduction ratio of 16:1, a ram speed of 10 mm s−1, and the extrusion temperature was set to 420 °C. The TiMg composite material was tested in two forms: (1) cylinders (labeled as TiMg cyl) and (2) dental implants (labeled as TiMg DI). Cylindrical samples had a diameter of 4 mm and a height of 8 mm and were machined longitudinally to the extruded rods. The as-machined cylinders were ground using abrasive grits (up to 4000# grit size) and then polished using abrasive diamond paste with a size of 1 µm to obtain a fine and smooth surface, as was described in [6]. The TiMg DI samples were a commercial design MV3.6-10 DI [30] with a thread diameter of 3.6 mm and a length of 10 mm. The TiMg DI samples were produced using a common computer numerical controlled (CNC) machining approach that is typically used to produce commercial Ti-based dental implants without any surface treatments. All samples were ultrasonically cleaned in ethanol, air-dried, and finally stored in a desiccator for coveted tests.

Surface roughness measurement

The surface roughness of the TiMg cyl and TiMg DI samples was measured using a ZEISS LSM 700 confocal microscope (Carl Zeiss Microscopy GmbH, Germany) according to the ISO 25178 guidelines. 3D-colored maps of the investigated surfaces were processed using ConfoMap ST software. Two surface areas of interest (SA1 and SA2) on the lateral side of TiMg cyl with an area of 320 × 920 µm2 for each one were examined (Fig. 1a). But five areas of interest on a thread pitch (P1 and P2), a thread tip (T1 and T2), and apical cutting edge (A) of TiMg DI were analyzed (Fig. 1c). The scanned areas were 480 × 320 µm2 for P1 and P2, 96 × 64 µm2 for T1 and T2, and 298 × 480 µm2 for A. The surface roughness parameters that included arithmetical mean height (Sa), root mean square height (Sq), and maximum height (Sz) were determined for the measured areas. The obtained data were expressed as mean ± standard deviation (SD).

Figure 1
figure 1

a SEM micrograph of the lateral surface of TiMg cyl that displays areas of interest (SA1 and SA2) for surface roughness measurement. b 3D-colored map of the representative surface area (SA1) of TiMg cyl. c SEM micrograph of MV3.6-10 TiMg DI that displays areas of interest on the thread pitch (P1 and P2), the thread tip (T1 and T2), and the apical cutting edge (A) for surface roughness measurement. 3D-colored maps of the representative surface areas d T1, e A, and f P1 of TiMg DI.

Full size image

Electrochemical measurements

To mimic the in vivo environment, HBSS was used in all experiments. The chemical composition of HBSS is reported in Table 1. The experiments were performed at 37 ± 0.5 °C and the pH was adjusted to 7.4. Electrochemical measurements were performed on a VersaSTAT3 potentiostat (AMETEK, Inc., USA), controlled by VersaStudio software, following the ASTM G5-14 and ISO 10993-15 guidelines with minor changes. A three-electrode cell (K47, Princeton Applied Research) containing 900 mL HBSS was used. In the cell, twin graphite counter electrodes and a saturated calomel reference electrode (SCE) (Hg/HgCl2/saturated KCl) were employed. The top and bottom flat sides of the TiMg cyl samples were covered with a silicone adhesive, leaving a 1-cm2 surface area exposed to the test solution. But only the top side of the TiMg DI samples was covered, leaving an exposed surface area of 1.8 cm2. The sample holder was inserted into a threaded hole on the top side of the TiMg cyl and TiMg DI samples. Both samples were vertically positioned and entirely immersed in the electrolyte as a working electrode. The OCP variation was recorded over the first 2-h immersion in HBSS. A scan rate of 0.5 mV/s was applied for PDP, and each test was carried out in the range of − 250 mV below OCP to 500 mV above. The PDP scans were performed after the tested samples were immersed in HBSS for 2 and 24 h. The parameters βa, βc (anodic and cathodic Tafel slopes, respectively), ECorr, iCorr, and RCorr were obtained from the PDP curves using CView 3.5 software program (Scribner Associates Inc., USA). For EIS, a frequency range of 100 kHz to 100 mHz was applied, and the perturbation amplitude was 10 mV. Similar to PDP, the scans were performed after the samples were exposed to HBSS for 2 and 24 h. For the PDP and EIS measurements after 24-h exposure, the OCP value was detected automatically by the potentiostat at the moment at which the measurements started. Equivalent electrical models were proposed by ZSimpWin 3.20 software to fit the obtained EIS data. All measurements were repeated three times to assure reproducibility. After corrosion testing, the surfaces of TiMg cyl and TiMg DI and the formed base film were analyzed using a scanning electron microscope (SEM, JEOL JSM-6610) equipped with an energy-dispersive X-ray spectrometer (EDS, Oxford).

Table 1 Chemical composition of Hank’s balanced salt solution
Full size table

Results

Evaluation of surface roughness

Figure 1 presents SEM micrographs of surfaces of the TiMg cyl and TiMg DI samples along with 3D-colored maps of the representative surface areas of both samples. The mean values of surface roughness parameters for TiMg cyl were slightly smaller compared with those of TiMg DI. The surface roughness parameters for the representative areas of TiMg DI were different. The surface of the thread tip had the smoothest roughness parameters, while the surface of the thread pitch was rather rough, as shown in Fig. 1d and f. Furthermore, the surface of the apical cutting edge had almost similar roughness parameters to those of the lateral surface of TiMg cyl (Fig. 1b, e). The corresponding roughness parameters, sa, sq, and sz, for the scanned surface areas of TiMg cyl (SA1 and SA2) and TiMg DI (T1, T2, A, P1, and P2) are given in Table 2.

Table 2 The corresponding surface roughness parameters for the scanned areas of TiMg cyl (SA1 and SA2) and TiMg DI (T1, T2, A, P1, and P2)
Full size table

Open-circuit potential (OCP)

Figure 2 shows curves of OCP versus exposure time of the TiMg cyl and TiMg DI samples immersed in HBSS for 2 h at 37 °C. The 2-h interval was selected since it was sufficient time for the surfaces of the studied samples to form a base film. The OCP variation with time was almost similar for both samples without any significant differences. In the initial stages of immersion (first ⁓ 15 min), an abrupt displacement of OCP occurred toward the positive (noble) direction. Afterward, OCP increased slowly with extending the exposure time. Apart from that, some considerable fluctuations in the OCP curve occurred for TiMg cyl.

Figure 2
figure 2

The open-circuit potential (OCP) variation with exposure time for the TiMg cyl and TiMg DI samples immersed in HBSS for 2 h at 37 °C along with the OCP values at the moment at which the experiments started, and 2 h later.

Full size image

Potentiodynamic polarization (PDP)

Figure 3 shows PDP curves of the TiMg cyl and TiMg DI samples after 2- and 24-h immersion in HBSS at 37 °C. The parameters βa, βc, ECorr, iCorr, and RCorr that were obtained from the PDP curves are presented in Table 3. After 2-h immersion, both samples exhibited high corrosion tendencies in the test solution, and TiMg cyl had slightly higher values of iCorr and RCorr compared with those of TiMg DI. Moreover, TiMg cyl had a higher value of βc compared with that of TiMg DI, indicating a huge release of H2 bubbles due to the dissolution of Mg. After 24-h immersion, a reduction in the values of iCorr and RCorr was observed, and ECorr shifted positively. The reduction was significant, particularly for TiMg DI, whereby RCorr of TiMg cyl and TiMg DI was 0.09841 and 0.01768 mmpy, respectively. For both samples, βa had higher values than those of βc. This revealed an anodic control during the corrosion process.

Figure 3
figure 3

Potentiodynamic polarization (PDP) curves of the TiMg cyl and TiMg DI samples after a) 2- and b) 24-h immersion in HBSS at 37 °C.

Full size image
Table 3 The parameters βa, βc (anodic and cathodic Tafel slopes, respectively), ECorr (corrosion potential), iCorr (corrosion current), and corrosion rate (RCorr) that were obtained from the potentiodynamic polarization (PDP) curves of the TiMg cyl and TiMg DI samples after 2- and 24-h immersion in HBSS at 37 °C
Full size table

Electrochemical impedance spectroscopy (EIS)

Figure 4 shows Nyquist plots of the TiMg cyl and TiMg DI samples after 2- and 24-h immersion in HBSS at 37 °C. The Nyquist plots that represent the EIS data were analyzed by employing electrical equivalent models, as shown in Fig. 5. The proposed models consisted of the resistance of the test solution (Rs), a constant phase element (CPE1) that represented the capacitance response of the formed film, the dispersion coefficient of the formed film (n1), the resistance of the formed film (R1), the double layer capacitance (CPEdl), the dispersion coefficient of the double layer (ndl), the charge transfer resistance (Rct), and Warburg impedance (W) that represented the impedance arising from the mass transfer and chemical diffusion processes. Model (a) and model (b) were used to fit the EIS data obtained after 2- and 24-h immersion, respectively. The EIS experimental data and the proposed equivalent models matched very well (Chi-square χ2 < 10−4). The Nyquist plots (Fig. 4a) show capacitive loops that tended to close for both samples immersed in HBSS for 2 h. After 24-h immersion, non-closing capacitive loops were observed, and the EIS data showed higher corrosion resistance (i.e., higher values of R1). Moreover, n1 had lower values for TiMg cyl and TiMg DI, and the values of ndl for TiMg cyl were higher than those of TiMg DI. Furthermore, with extending the exposure time, TiMg cyl and TiMg DI had an extremely low Rct, and the Warburg impedance was visible in the Nyquist plots as a straight line in the low frequency region (Fig. 4b). The corresponding values of parameters of the equivalent models are presented in Table 4. It is worth pointing out that the EIS data were consistent with the results obtained by the PDP measurements.

Figure 4
figure 4

Experimental and simulated Nyquist plots of the TiMg cyl and TiMg DI samples after a 2- and b 24-h immersion in HBSS at 37 °C.

Full size image
Figure 5
figure 5

Electrical equivalent models, a and b, employed for fitting the EIS experimental data obtained after 2- and 24-h immersion, respectively.

Full size image
Table 4 Parameters of the equivalent models obtained from fitting the EIS experimental data
Full size table

Figure 6 shows the mechanism of Mg degradation of the TiMg composite. The Mg degradation resulted in the formation of charge-deficient sites at the surface of the studied sample. As a result, distinct species, such as Ca, P, and carbonate, were incorporated from the test solution into the surface. The corrosion by-products are mainly carbonated apatite species CAp (Ca5(PO4/CO3)3(OH)), magnesium phosphate (Mg3(PO4)2), and magnesium hydroxide Mg(OH)2, as was confirmed in our previous studies [3,4,5,6].

Figure 6
figure 6

Systematic diagram of the proposed mechanism of degradation of the TiMg composite.

Full size image

Examination of surfaces after corrosion

Figure 7 shows SEM micrographs of surfaces of the TiMg cyl and TiMg DI samples after 2- and 24-h exposure to HBSS at 37 °C. The surfaces of the studied samples were almost fully covered in corrosion by-products after 24-h immersion when compared with 2-h immersion. The base film that formed on the surface of the TiMg cyl sample was thicker and grew uniformly after 24-h exposure when compared with 2-h exposure, as shown in Fig. 8. The formed film had a thickness of 59.8 ± 4.2 and 88.4 ± 3.2 µm after 2- and 24-h exposure, respectively. This behavior was consistent with the results of Rcorr mentioned above. The EDS analysis confirmed that the surface of TiMg cyl after 24 immersion was passivated by a film enriched with Ca/P species (CAp). Moreover, the low percentage of Ti revealed that its surface was also covered in corrosion by-products. These SEM observations were in line with the electrochemical results outlined above.

Figure 7
figure 7

SEM micrographs of surfaces of the TiMg cyl and TiMg DI samples after a, c 2- and b, d 24-h immersion in HBSS, respectively, along with results of elemental EDS analysis of the formed base film from the rectangle area. The arrows refer to the uncovered surface of Ti.

Full size image
Figure 8
figure 8

Representative cross-sectional SEM micrographs of TiMg cyl after a 2- and b 24-h exposure to HBSS.

Full size image

Discussion

When a dental implant is placed in the oral environment, an electrochemical interaction between the implant material and the surrounding environment may occur. This interaction results in a release of metallic ions into the physiological environment and thus leads to the formation of corrosion by-products on the implant surface. This behavior may both result in damage to the surrounding tissues and impaired healing and trigger some adverse effects, such as allergy [31]. In an aqueous medium, Mg undergoes two chemical processes; dissolution of Mg [Eq. (1)] and hydrogen (H2) evolution [Eq. (2)] [32].

$${\text{Anodic}}\;{\text{process:}}\;{\text{Mg}} \to {\text{Mg}}^{{{2} + }} + {\text{2e}}^{ - }$$
(1)
$${\text{Cathodic}}\;{\text{process:}}\;{\text{2H}}_{{2}} {\text{O}} + {\text{2e}}^{ - } \to {\text{2OH}}^{ - } + {\text{H}}_{{2}} \uparrow$$
(2)
$${\text{Overall}}\;{\text{reaction:}}\;{\text{Mg}} + {\text{2H}}_{{2}} {\text{O}} \to {\text{Mg}}\left( {{\text{OH}}} \right)_{{2}} + {\text{H}}_{{2}} \uparrow$$
(3)

According to [Eq. (3)], the surface of Mg is typically covered in Mg(OH)2 with some micropores and defects caused by the H2 evolution [33]. When Mg is coupled with Ti, the galvanic corrosion arises from a high difference in their standard electrode potential (− 2.356 vs. − 0.163, respectively) [20, 21]. The galvanic coupling intensifies the degradation of Mg, resulting in the H2 evolution, which relatively curbs the protectiveness of the base film toward further degradation. Hence, it was significant to study the galvanic corrosion of the TiMg dental implant [22, 23].

Mg-containing materials generate an electrical potential when exposed to an electrolyte. This potential (OCP) reaches after a certain time of exposure to a value at which the tested sample is in a relatively steady state with its environment. For the OCP measurements, the initial abrupt increase was attributed to the formation and thickening of the base film formed on the surfaces of TiMg cyl and TiMg DI. Afterward, the slow increase in OCP revealed the growth of the formed film on the surfaces of the studied samples that promotes their corrosion resistance. The fluctuations in the OCP curves of TiMg cyl are difficult to interpret, but they were likely linked to surface characteristics and/or the occurrence of localized corrosion. For the PDP measurements, the cathodic branch of the polarization curve corresponds to a hydrogen evolution reaction due to the dissolution of Mg. TiMg cyl had a higher value of βc after 2-h immersion compared with that of TiMg DI, indicating a huge release of H2 bubbles and a higher Rcorr ensued. A prominent reduction in anodic current densities with increasing immersion time was observed, particularly for TiMg DI. This behavior revealed a stable passive film and high corrosion resistance. The difference in the Mg degradation of both samples after 24-h immersion was attributed to different geometries of the tested samples (cylindrical vs. threaded outer surface), surface roughness parameters, and the forms of Mg at the samples’ surfaces. Although TiMg DI had slightly higher values of the surface roughness parameters than those of TiMg cyl, it showed lower RCorr. This behavior was mostly attributed to a variation of the Mg component at the surface of TiMg DI compared with the one at the surface of TiMg cyl. In other words, TiMg DI of a complicated threaded design had two types of Mg, one type was arrayed along the direction of extrusion in the longitudinal direction, and the other one was in the form of an interconnected skeletal structure in the transversal direction. This behavior was consistent with [34], which revealed that the corrosion resistance of a rolled AZ80 magnesium alloy is very sensitive to the texture of Mg at the surface. Apart from that, as the exposure time extended, TiMg DI with a partially rough surface made the formed film relatively bound to its surface (Fig. 7d) and thus reduced the degradation of Mg.

EIS is a powerful method for evaluating the corrosion mechanism and the formed base film and its protectiveness [35,36,37]. To interpret the obtained EIS plots, it was necessary to propose an equivalent electrical circuit model that described all involved electrochemical reactions. The equivalent models consisted of different elements that precisely characterized the corrosion behavior and mechanism. The employment of CPE instead of an ideal capacitor was necessary due to the inhomogeneous nature of the studied systems (i.e., rough surfaces and/or a variation of reaction rates). An electrical double layer (dl) represented the interface between the sample surface and its surrounding medium. This double layer was formed as ions from the medium were adsorbed onto the sample surface. Mg can electrolytically dissolve into the medium according to (Eq. 1); therefore, a charge transfer reaction occurred. Moreover, Mg2+ that diffused into the surrounding medium resulted in forming by-products on the sample surface, and such chemical diffusion processes can create a Warburg impedance, which depends on the frequency of the potential perturbation. At high frequencies, the Warburg impedance has small values since diffusing reactants don’t have to move very far, whereas it has high values at low frequencies since the reactants have to diffuse farther [38]. The resistance of the formed film (R1) represented a barrier against ions migration, which reflected its protectiveness level [37, 39]. The capacitive response in the high frequency range was related to the base film, while the response in the low frequency range was linked to the double layer capacitance and the charge transfer resistance. After 2-h immersion, the capacitive loops that tended to close demonstrated the formation of a base film on the surfaces of the tested samples. The reaction was then controlled by a charge transfer process. As the immersion time extended, the H2 evolution resulted in the formation of a relatively loose film, so the n1 values decreased. Moreover, TiMg cyl had higher values of ndl compared with those of TiMg DI. This was likely attributed to the variation of the geometry and surface roughness of the TiMg cyl and TiMg DI samples and the diffusion process through the formed film. After 24-h immersion, the high values of R1 for TiMg cyl and TiMg DI reflected the passive behavior and high corrosion resistance. The Warburg impedance was visible in the Nyquist plots after 24-h exposure that arose from the formation and dissolution of chemically non-stable by-products caused by the H2 evolution. Therefore, as the exposure time extended, the polarization was due to a combination of kinetic and diffusion processes. The involved reaction was then controlled by both charge transfer and semi-infinite length diffusion processes, as shown in Fig. 5b.

As shown in Fig. 7, the formed by-products almost fully covered the surfaces of the tested samples after 24-h immersion in HBSS. The Mg component present at the surface obviously deteriorated, while Ti was protected because of the galvanic coupling. The TiMg cyl sample exposed to HBSS for 24 h was covered in a film composed of bioactive Ca/P species (CAp), (Mg3(PO4)2), and Mg(OH)2, as was confirmed by the EDS analysis and in our previous works [3,4,5,6]. The EIS results along with the SEM observations demonstrated that CAp formed the stable passive base that attached directly to the surface, while the other by-products started to dissolute as the exposure time extended and thus led to an increase in the mass transfer and chemical diffusion processes. As was reported in our previous works [3, 6], the biomimetic CAp boosted the cell survival and adhesion, and it is expected to promote the proliferation and osseointegration criteria [40].

This investigation revealed that the implant design, surface roughness, and forms of Mg at the surface crucially influenced the corrosion behavior and surface passivity. The accomplished results provided evidence that TiMg DI of a complicated threaded design had more stable passive film and higher corrosion resistance in HBSS compared with those of TiMg cyl of a cylindrical design. These results, along with the outstanding findings that were demonstrated in our previous works, revealed that TiMg DI has immense potential as a promising material for the fabrication of DI.