Corrosion Properties of Ti-Nb-Sn Alloys Prepared by Mechanical Alloying

Abstract

Recent efforts to develop materials for biomedical applications, such as rigid implants, have focused on improving biocompatibility, and obtaining equivalent microstructure and mechanical properties. However, due to the complex biological environment of the human body, materials often fail in one or more of their required properties. In this study, we examined the mechanical and electrochemical behavior of Ti base alloys combined with Nb and Sn. Titanium-niobium-tin alloys of composition Ti-30Nb, Ti-30Nb-6Sn, Ti-30Nb-9Sn and Ti-30Nb-12Sn were synthesized by mechanical alloy of powders and thermally sintered. The alloys were characterized microstructurally and mechanically. Subsequently, the corrosion of alloys in Hank's solution by electrochemical techniques was assessed. The characterization results showed that the alloy microstructure is composed by a major β-Ti phase rich in Nb and Sn, and minor α-Ti phase containing less Nb and Sn. The microstructural porosity observed was similar between the alloys, except for the Ti-30Nb-6Sn alloy, which presented less porosity. Electrochemical results showed polarization associated with the microstructural characteristics of alloys, particularly their porosity and density. Consequently, the Ti-30Nb-6Sn alloy presented the lowest corrosion current density (i_corr) due to having the lowest exposed surface area. However, the higher porosity present in the other alloys promoted of increase in their i_corr values. Ti-Nb based alloys showed either passive or limited current density behaviour, due to the formation of TiO₂, Nb₂O₅ and SnO₂. According to the EIS results, the corrosion mechanisms in Hank´s solution, were controlled by resistive-capacitive impedance and by diffusion through the oxide layer. Optical microscopy, scanning electron microscopy and x-ray diffraction characterizations, were used to complement the results.

1 Introduction

As life expectancy has increased, as a consequence of medical and social advances, it has led to an increase in bone diseases in the elderly, as well the need to use structural support or permanent implants to replace damaged bones (Ref 1). Titanium is currently the most important element used as a biomaterial in biomedical implants, since it is not carcinogenic, genotoxic, mutagenic, allergic, or corrodible, while being only moderately cytotoxic (Ref 2,3,4,5,6) Furthermore, it exhibits good mechanical properties (Ref 7) which can be improved by alloying it with Nb and Sn (Ref 8,9,10).

Titanium is an allotropic element which crystalizes into several phases, of which the most common phases in medical implants are named alpha and beta. Of these two phases, the beta phase is preferred because of its lower Young's modulus (Ref 3, 11, 12), which is modified with Nb and Sn additions (Ref 13,14,15). Niobium stabilizes the titanium beta phase at room temperature (Ref 16, 17) while tin acts as a controller for the alloying process, facilitating the niobium diffusion (Ref 18). Although titanium, niobium, and tin are ideal for manufacturing medical implants, it is difficult to obtain alloys of these metals due to their different melting points. Therefore, powder metallurgy is an alternative technique to obtain them, yielding a powder whose properties are enhanced. In addition, powder metallurgy has a lower cost than other methods in producing Ti alloys with desired geometry and mechanical properties (Ref 19, 20).

While there are corrosion reports on Ti-Nb alloys, as the alloy has already been studied for biomedical applications (Ref 3, 21,22,23), there has been little research on the corrosion mechanisms of Ti-Nb-Sn base alloys (Ref 24). In the galvanic series of metals, titanium has a standard potential of − 1.63 V which is close to that of aluminum. Therefore, titanium cannot be considered as being intrinsically noble, but its excellent resistance to general corrosion in most environments is recognized (Ref 25). This is because as a result of a stable protective surface layer, which consists primarily of TiO₂. This thin oxide layer will passivate the alloy surface, as long as the integrity of the TiO₂ layer is maintained (Ref 25). However, the addition of alloying elements such as Nb and Sn may influence the corrosion behavior of the Ti-Nb-Sn alloy. As the titanium alloys, the formation of a stable and protective oxide layer on the Ti-Nb-Sn alloy's surface when exposed to an oxidizing environment is expected. The composition and structure of the oxide layer are crucial in determining the corrosion resistance of the alloy.

In the current development of new Ti alloys there is a line of study which addressees the combination of titanium with other biocompatible elements. Some examples of these combinations include Ti-Ta-Sn foams (Ref 8), Ti-Nb-Sn-hydroxyapatite composites (Ref 18), of β-stabilized Ti-Nb-Sn-Cr alloys (Ref 26), Ti-10Ta-8Mo and Ti-10Ta-8Mo-3Sn alloys (Ref 27). These have shown improvement of the properties of Ti alloys for biomedical purposes. Another line of study focuses on alternative methods of manufacture to regulate the architecture of non-stochastic materials, with predicted mechanical properties which overcome the limitations of conventional manufacturing methods. Many of the published works in this field suggest that additive manufacturing is surpassing conventional methods. However, these techniques still need to be perfected (Ref 19, 20). Additionally, research is being undertaken to assess the electrochemical behavior and corrosion resistance of based Ti alloys (Ref 28,29,30).

Antecedents for the current work include the synthesis and characterization of Ti-Ta-Nb-Mn foams obtained by powder metallurgy (Ref 9) in 2016, a study of the microstructure evolution of Ti-Nb based alloy obtained by high-energy ball milling (Ref 17) in 2017, and the synthesis, characterization, and mechanical properties of Ti-Nb-Ta-Zr foams for biomedical applications (Ref 10) in 2019. Subsequently, with a focus on the crystallography of these kinds of Ti alloys, the effect of Sn on synthesis of nanocrystalline Ti-based alloy with FCC structure (Ref 15) was published in 2020. More recently, the analysis of mechanical and electrochemical properties of the Ti-Ta-Na foam obtained by sintering (Ref 8) was published in 2022.

In order to gain more robust scientific knowledge about the properties of these materials and their suitability for biomedical applications, we present a study of the mechanical and electrochemical behavior of Ti base alloys combined with Nb and Sn. In our research, we study the microstructural and mechanical characterization and evaluate the electrochemical corrosion of Ti-Nb-Sn alloys obtained by mechanical alloying.

2 Experimental Developments

The experimental procedure to obtain and characterize the Ti-Nb-Sn alloys starts with the mechanical alloying of pure metallic powders. Some samples of the obtained alloy powder were characterized by scanning electron microscopy (SEM), energy-dispersive spectroscopy (EDS) and x-ray diffraction (XRD). Some other alloy powders were compacted and thermally sintered to obtain pieces of the alloys with structure. Once the alloys were obtained, the microstructure characterization by optical microscopy (OM), scanning electron microscopy and energy-dispersive spectroscopy were performed. Additionally, the porosity, density and elastic modulus were measured in order to compare the mechanical characteristics of the Ti-Nb-Sn alloys versus the properties of human bone. Finally, the corrosion behavior of the Ti-Nb-Sn alloy samples was carried out by electrochemical techniques. Corroded sample surfaces were analyzed by SEM and EDS to complement the results.

2.1 Mechanical Alloying of Pure Metallic Powder

Four alloys were obtained by mechanical alloying. The Ti, Nb and Sn metals quantities in atomic percentage (at. %) were: 30 of Nb; 0, 6, 9 or 12 of Sn; and Ti balanced. Samples are identified as: S1 = Ti-30Nb, S2 = Ti-30Nb-6Sn, S3 = Ti-30Nb-9Sn, and S4 = Ti-30Nb-12Sn. On one hand, the percentage of Nb was chosen to ensure the preferential formation of the beta phase titanium (β-Ti) at relative low temperature, according to phase equilibrium thermodynamics diagrams (Ref 16). Specifically, atomic percentages of Nb close to 30 promotes preferential formation of β-Ti phase around 1200 °C (Ref 16). On the other hand, the percentage of Sn was chosen at low values to see the sensibility of the α/β-Ti equilibria in the Ti-Nb matrix.

All these samples were obtained by milling powders for 10 hours, using yttrium-stabilized zirconium oxide balls, highly polished with a hardness 60 times greater than those of regular steel which are highly resistant to wear and resistant to acids and solvents. Two sizes of balls were used, of 5 and 10 mm. The ratio between the number of balls was 1 to1 by each size and the powder-ball weight ratio was 10 to 1. The milling speed was 250 rpm for all samples.

A planetary ball mill RETSCH PM 400 was used to carried out powder alloying. The purity of all three powders was greater than 99% and the mesh size was 100, 325 and 325 for Ti, Nb and Sn, respectively. To control the milling process 2% weight of stearic acid was used as combustion suppressor. Before mechanical alloying of the samples S1, S2, S3 and S4, the grinding media and the grinding jar were coated by prelaminar milling of pure titanium powder to mitigate contamination of the samples.

2.2 Characterization of the Powdered Alloys

The microstructure of the powder alloys was analyzed by SEM, EDS and XRD. Firsts the SEM + EDS were used to analyze the microstructure of the powders and the distribution of the elements within them. XRD was used to obtain the characteristic diffraction patterns of the powders, which were subsequently analyzed by means of the Rietveld method. This procedure minimizes the differences ${S}_{y}$ between the calculated ${Y}_{i}\left(c\right)$ and the estimated ${Y}_{i}\left(0\right)$ x-ray intensities, weighting a factor ${w}_{i}$ to each point $i$ in the diffraction patterns (eq.1) and giving information of the phase distribution within the samples.

$$min{S}_{y}=\sum {w}_{i}{\left|{Y}_{i}\left(0\right)-{Y}_{i}\left(c\right)\right|}^{2}$$

(1)

The phase distribution of the samples was measured in transmission mode using a step of 0.12° and 10 seconds per step, over a 2θ angular range of 20° to 100° of a multipurpose powder diffractometer. Then, to obtain the diffractograms for the four samples, the results were processed and analyzed in the software MATCH^® to identify the corresponding peak to each phase. Once all peaks were identified, their crystallography information files (CIF) were obtained from Crystallography Open Database (COD). These files are COD 9016190 for alpha titanium, COD 9008554 for beta titanium, COD 9008546 for Nb, COD 9008570 for Sn, and COD 1521478 for yttrium-stabilized zirconia (YSZ). The CIFs were then loaded to Maud^®, which is a general diffraction/reflect analysis program based on the Rietveld method and the LaB₆ standard was used to correct instrumental error (a = 4.1565915(1) Å).

The software MAUD^® was used to estimate, from the adjustment of the peak intensity for the diffractogram of each alloy (Eq. 1), the weight percentage of elements within the samples including beta phase titanium $\left(\%\beta Tiw\right)$ and alpha phase titanium $\left(\%\alpha Tiw\right)$. Once these percentages had been obtained, the distribution of the beta phase $\left(\%\beta Ti\right)$ and alpha phase $\left(\%\alpha Ti\right)$ of titanium was calculated.

The micrographs of the powder alloys were obtained by using a scanning electron microscope JSM-IT30 and the x-ray diffraction patterns were obtained by using a multipurpose powder diffractometer STOE STADI MP, equipped with a DECTRIS MYTHEN 1 K detector with K_α1 of Couper radiation (λ = 0.15406 nm).

2.3 Compaction and Powder Sintering

The alloys were obtained using a compaction machine ZWICK/ROELL model E030. The compaction pressure was 1193 MPa, and the compaction time was 12.5 minutes divided in 5 stages of 2.5 minutes using polyvinyl PVA with 10% of water as a lubricant and binder agent to avoid breakage of samples. Last settings were chosen from previous calibration tests.

After compaction, the samples were sintered in a tubular oven NABERTHERM. The sintering procedure was carried out in three stages; the first one at 220 °C for 3 hours to volatilize the binder agent, the second one at 800 °C for 10 minutes as a gradual step to the final sintering temperature to avoid abrupt changes and the third one was at 1250 °C for 4 hours to sinter the alloys to ensure the grain necks formation. The geometry of the obtained samples was cylindrical shape, sizing in average 10 mm of length and 8 mm of diameter. Last parameters were selected to obtain the specific size and form of the fabricated samples from previous calibration test.

The total number of sintered samples of the alloys was 12, three replicates of each combination S1, S2, S3 and S4. Owing the mechanical characterization, the metallography studies and the electrochemical studies were carried in different places, one set of the replicates was used to the mechanical characterization, another was used to the metallography studies and the last one was used to the electrochemical studies. It should be highlighted that, in this work, there was no chance to verify the repeatability of the response of each set of samples used in all three kinds of studies, however very clear tendencies were identified from results, giving this work an only exploratory scope and leaving the rigorous statistical analysis for studies where the application of the material is already well defined.

2.4 Characterization of the Alloys

Some samples of the sintered alloys were polished and etched with Kroll´s reagent (solution of 6% HNO₃ and 2% HF in H₂O) (Ref 31), to reveal their microstructure. These samples were analyzed by OM and SEM to see the accommodation of the powder grains and to compare their microstructure and the distribution of the metallic elements in the alloys on the etched surfaces.

2.5 Density and Porosity Measurement

The estimation of density and porosity of the Ti-Nb-Sn alloys was carried out following the standard ASTM C373 (Ref 32). This procedure requires the dried weight $\left(D\right)$, the immersed weight $\left(S\right)$ and the saturated weight $\left(M\right)$ of the samples. Once, these data were obtained, the density (ρ), the apparent porosity), and the absorption of water $\left(A\right)$ of the samples were calculated by Eq 2, 3 and 4, respectively.

$$\rho =\frac{M-D}{M-S}\times 100$$

(2)

$$B=\frac{D}{M-S}$$

(3)

$$A=\frac{M-D}{D}\times 100$$

(4)

2.6 Estimation of Elastic Modulus

The elastic modulus of the sintered alloys was measured with a Tektronix TDS 2012B oscilloscope, following the standard ASTM D2845 (Ref 33), which describes the procedure to measure the speed of the compression waves ${{\varvec{v}}}_{{\varvec{L}}}$ in a known density media ρ. Then, with these data, the ultrasonic elastic constant $E$ was calculated using Eq 5.

$$E=\rho \times {v}_{T}^{2}\times \frac{3{v}_{L}^{2}-4{v}_{T}^{2}}{{v}_{L}^{2}-{v}_{T}^{2}}$$

(5)

where, the strength limit of the alloys was not measured, since samples size was not compatible with traction destructive test. However, it is well known, there is an unfavorable relationship between porosity, elastic modulus and strength limit (Ref 34) which is left for future research.

2.7 Study of the Electrochemical Behavior

Before the electrochemical tests were assessed, two pieces of each compacted and sintered samples were joined to a cupper insulated wire for electric connection, and then were embedded in epoxy-resin, leaving one flat face (cross section) uncovered in order to be polished and tested in the corrosion evaluations. Particular care was taken during the preparation of the encapsulated samples to avoid crevice corrosion along the edges of the sample, verifying and correcting any separation or pore at the alloy/epoxy-resin interface of each sample. The uncovered cross section of each specimen was mechanically polished with 3000 grit size sandpaper, then washed in an ultrasonic bath for 1 hour to remove contaminants from the surface. It is important to mention that roughness and porosity of the alloys is not uniform across the complete surface of the specimens due to defects resulting from the sintering process. However, the effects of this in the corrosion behavior are only addressed qualitatively in this study.

The electrochemical tests of the alloys were carried out by using a potentiostat/galvanostat ACM Instrument Gill-AC connected to a computer (with a sweep and measurement resolution of: 25 µV and 1.0 µV ± 0.0015% nonlinearity, respectively). Two techniques were used to assess the electrochemical behavior of samples. The potentiodynamic technique (TF) was used, applying a polarization potential range from − 400 mV to + 2000 mV vs open corrosion potential (OCP) with 1 mV/s scan rate. The anodic over-potential of 2000 mV was applied in order to identify any occurrence of passive behavior. The electrochemical impedance spectroscopy (EIS) technique, applying an alternating current (AC) potential of 30 mV of amplitude from the 30 kHz to 0.01 Hz frequency range. 50 readings per tests logarithmically spaced between the frequency range were taken. Two sets of EIS´s measurement of each sample were taken, but one experimental result set was considered for plotting. Before each electrochemical measurement, the samples were subjected to the open circuit condition until a steady-state potential was reached. This last between 15 and 20 minutes, and potential value measured was taken as the OCP value. The EIS data for each set up were collected in two experimental times: first one was measured after steady-state potential was reached, and the second was measured after 24 hours of continuous immersion, approximately. The electrochemical cell was composed of a saturated calomel electrode (SCE) as reference electrode (RE), a platinum wire as auxiliary electrode (AE), and the Ti-Nb-Sn samples as the working electrodes (WE), whose flat exposed and tested area was 0.5 cm². The electrolyte used was Hank’s solution at 37 °C. Hank’s solution (with pH 7.4), has the chemical composition of 8 g/L NaCl, 1 g/L glucose, 0.4 g/L KCl, 0.35 g/L NaHCO₃, 0.14 g/L CaCl₂, 0.1 g/L MgCl₂·6H₂O, 0.06 g/L Na₂HPO₄·2H₂O, 0.06 g/L KH₂PO₄ and 0.06 g/L MgSO₄·7H₂O (Ref 35). After the electrochemical tests, the corroded surfaces samples immersed in Hank´s solution for 24 hours were analyzed by SEM and EDS to identify corrosion results. Also, EIS results were analyzed by fitting the data (using the ZView software) of just one set of representative experimental results, applying the equivalent circuit model (ECM), described in the results and discussion section.

3 Results and Discussion

3.1 Microstructure and Crystallography of the Powder Alloys

The results of the SEM analyses of sintered alloys shown that the distribution of all elements, Ti, Nb, and Sn, in all alloys was uniform (Fig. 1), it was found that the particle shape had changed from a flake, for the base alloy, to a circular shape (Fig. 1a), for the alloy with 12 at.% of Sn (Fig. 1d). In addition, from the image analysis of the four samples of powder alloys, the mean and standard deviation of the grain diameter, as well as, the most frequent range of the grains was estimated; the means of the grain diameter were 3.15, 3.10, 1.82 and 1.22 µm for S1, S2, S3 and S4, respectively, the standard deviation of the grain was 1.89, 2.19, 1.89 and 1.22 µm, respectively, and the most frequent range of diameter were 2-4 µm for S1, S2 and S3, and 0-2 µm for S4. From these data are evident the increase in the atomic percentage of tin reduces the mean of the grain diameter of the powders, which may modify the final structure of the sintered alloys (section 3.3).

Fig. 1

Morphology by SEM of the alloy powder: (a) S1, (b) S2, (c) S3, and (d) S4

The XRD results presented in Fig. 2 show the characteristic peaks of alpha and beta phases within alloys of which the most important peaks are 38.4° (101), 55.6° (200), 69.6° (211), 82.37° (202), 94.8° (301) corresponding to beta phase titanium (Ref 23, 36, 37) and 35° (100), 38.4° (002), 40.15° (101), 53° (012), 62.8° (110) as the alpha phase titanium. The other peaks appearing in diffractograms which are not indexed belong to Nb and Sn crystal structures, in addition, the intensity of all peaks increases as the amount of Sn incorporated in the samples raises, being the peaks of the S1 alloy powders, the shortest and the peaks of S4 alloy powders, the highest.

Fig. 2

X-ray diffraction patterns of the powder’s alloys

From x-ray diffraction patterns, the Ti phases distribution for the obtained alloys S1, S2, S3 and S4 was computed by mean of the Rietveld method (section 2.2). The percentage of β-phase is 75.24% for the S1 alloy, 77.12 % for the S2 alloy, 82.60 % for the S3 alloy and 80.30 % for the S4 alloy. The difference in the percentages for each alloy is the α-Ti phase. From these distributions, it is observed that Sn increases the percentage of β-Ti phase within the alloys, reaching the maximum percentage of this phase at S3 alloy, with a slight reduction for S4 alloy. It should be highlighted that this behavior may indicate the limit of the Sn effect in the promotion of β-Ti phase formation. Similarly, there have been other studies (Ref 26, 27), which also incorporated Sn and others elements in Ti-based alloys to promote phase transformation, and their results concur with the current findings.

3.2 Microstructure of Sintered Alloys

The four cylindrical samples of alloys obtained by heat treatment sintering are shown in Fig. 3. The effect of the incorporation of Sn to the Ti-Nb based alloy was evident, since the top flat surface of the samples presented surface defects. Therefore, the Ti-Nb alloy without Sn (S1), presented the most regular surface (Fig. 3a). The S2 alloy (with 6 at. % of Sn) exhibited an irregular edge and an elongated depression in the center of its top face (Fig. 3b). The S3 alloy (with 9at. % of Sn) presented a rough surface with grooves in different directions (Fig. 3c). But, the Ti-Nb alloy with 12 at. % of Sn(S4) showed the most imperfect and irregular surface characteristics, with elongated grains and lack of microstructure cohesion (Fig. 3d).

Fig. 3

Top images of the sintered Ti-Nb-Sn compressed powders: (a) S1, (b) S2, (c) S3, and (d) S4 alloys

This etching procedure revealed the grain limits, and, at the same time, it shown the presence of the two phases within the alloys (Fig. 4). A homogeneous morphology surface with some pitting and well-defined grain boundary is easily identifiable (Fig. 4). The optical image of the S1 alloy surface shows a microstructure surface with some defects and pitting caused by the etching, (Fig. 4a), as well as some lighter areas. The etched S2 alloy exhibits a surface with less defects than S1, and similar evidence of the dissolution grain by etching, albeit less, and lighter areas are also shown, (Fig. 4b). In Fig. 4(c) the optical micrograph of the etched S3 alloy is shown, where a surface with more defects and etching damage are observed, however, grain boundary is not well defined. Similarly, the microstructure of the S4 alloy (Fig. 4d) it is not well defined, but grain size of both phases is longer, and the defects density increased (Fig. 4), moreover the number of defects coming from sintering and their diameters increases as the atomic percentage of Sn increases. On the other hand, the number of pits coming from etching and their diameters decreases as the atomic percentage of Sn increases.

Fig. 4

Surface images by OM of the etched sintered alloys at 500x magnification: (a) S1, (b) S2, (c) S3, (d) S4

3.3 Distribution of Phases Within Surface of the Sintered Alloys

The morphological characterization of alloys by SEM is shown in Fig. 5. Two grain phases in the microstructure were identified in the four alloys. The brighter phase, represented as the matrix one, contains the dark phase (the last, indicated by the yellow asterisk), that with the titanium phase distribution, the brighter phase and the dark grains correspond to the β-Ti and α-Ti phases, respectively. The dark grain diameter decreases as the proportion of Sn increases in the Ti-Nb based alloys. Moreover, the pits produced by the etching (Fig. 4) match with dark grains. Therefore, the pits on the surface of the alloys were caused by anodic dissolution of the dark grains, which are more susceptible to Kroll reagent (Ref 31) than the brighter phase of the alloys. However, the S4 alloy, shows two clear different regions, the refined grain like the other alloys, and depressed dark zones associated with the coalescence of defects. Thus, the voids size is wider, due to the defects cluster formation. However, the pits resulting from etching in this alloy are smaller than those of the other three alloys samples.

Fig. 5

Morphological images by SEM of the etched alloys at 5000x magnification: (a) S1, (b) S2, (c) S3, (d) S4

Results of the EDS analysis presented a distribution of Ti, Nb and Sn within the two Ti-phases identified in the four alloys. Table 1 shows the at. % of the Ti, Nb and Sn elements of the brighter phase, the β-Ti, which increases with the at. % of Sn. Here, the proportion of Nb is higher than the nominal quantity of alloys. Table 1 shows the quantity (in at. %) of the metallic elements in the dark grains, the α-Ti phase. The proportion of Nb is closer to the initial one in alloys (10 at. %) although, the proportion of Sn increases.

Table 1 Elemental chemical analysis of the sintered Ti-Nb-Sn samples, and corroded surface of samples immersed in Hank’s solution at 37 °C during 24 hours

Although the elements were uniformly distributed in the samples (Fig. 1), after sintering heat treatment of the alloys, two phases were formed, one of them, enriched with Nb and Sn and the other impoverished of these elements, but with high amount of Ti. This behavior is evidence of the segregation of elements due to Sn the maximum sintering temperature (1250 °C) is much higher than the fusion temperature of Sn (232 °C). The density of Nb (8.4 g/cm³) is slightly higher than the density of Sn (7.3 g/cm³) and it is almost twice the density of Ti (4.5 g/cm³). Thus, the melted Sn covers the niobium and expels the Ti, due to gravity. A close look at the grains of the S4 alloy (Fig. 6) gives stronger evidence of the phase segregation, identified as the large dark grains and large brighter phase grains, as well as regions where both phases are partially mixed (Fig. 6b), the segregation of the phases in the S1 alloy should be caused by the molten Ti.

Fig. 6

Morphology by SEM of the segregated grains of the S4 alloy at: (a) 100x and (b) 500x magnifications

3.4 Mechanical Properties of the Sintered Alloys

The estimation of the voids, density and elastic modulus was carried out to be compared with human-bone characteristics. In accordance with the porosity results, the four alloy samples exhibit a voids density range between 20% and 40% (Fig. 7a). However, these values are within the minimum and maximum interval of human-bone porosity (Ref 34). The density measurement results of the four samples presented values in the range of 4 g/cm³ and 5 g/cm³ (Fig. 7b), which is more than the double that of the density average in human bone (Ref 34). Similarly, results of the elastic modulus indicated that the sintered Ti-Nb-Sn alloys exhibit values of elastic modulus between the range of 38 GPa and 48 GPa (Fig. 7c), as before reported in research on Ti-Nb-Sn alloys with similar composition to the S2 alloy evaluated here (Ref 2, 3, 37,38,39). Furthermore, the Young's modulus was slightly higher than the average elastic modulus of human bone (Ref 34). It is noteworthy that this corresponds to a very similar profile of the void fraction, density, and the elastic modulus, due to their very close relationship between each other. Therefore, density and elastic modulus can be enhanced by creating porosity in alloys.

Fig. 7

Mechanical properties of the alloys: (a) Porosity, (b) Density and (c) Elasticity modulus

3.5 Electrochemical Behavior of the Ti-Nb-Sn Alloys

Figure 8 presents the dynamic polarization curves (potentiodynamic) of the four alloys (S1 to S4) immersed in Hanks’ solution at 37 °C. Results show that the S4 alloy sample presented the most negative corrosion potential (E_corr) of about − 564 mV_SCE, and corrosion current density (i_corr) of 0.0126 mA/cm², which represents the most active behavior at the first immersion time. Similar behavior was presented by the S1 alloy, as such was reported before (Ref 21, 23, 36). The S2 alloy presented the most positive values of the E_corr, of around − 358 mV_SCE and the lowest i_corr of 0.0009 mA/cm². Increasing Sn content from 6 to 12 at. % promoted a displacement curve to more negative potentials and higher current density values (Fig. 8). Preliminary observation of the potentiodynamic behavior was associated with the compactness of the samples after the sintering treatment, i.e., the porosity as well as the density of the alloys shown in Fig. 8, where the S2 alloy showed the lowest defects density and, therefore, the lowest i_corr value. Correspondingly, the S2 sample should present the lowest corrosion rate due its lower exposed surface area, so the increase in porosity increases this surface area and the site densities were attributed to the concentration changes in solution promoting an increasing of current density. Table 2 presents the obtained parameters from potentiodynamic results applying the method of intercepts (Ref 40). Additionally, the Tafel anodic and cathodic constants were obtained.

Fig. 8

Tafel curves of the alloy from the corrosion tests

Table 2 Obtained parameters from Tafel’s curves by mean of the standard practices (Ref 51)

During the first + 200 mV of anodic polarization, approximately, alloys presented activation process, consequently, the corrosion products formation covering the alloys surface occurs. With increasing applied potential, Ti-Nb-Sn alloys showed semi-passive or limited current density for about + 300 mV potential range: from − 400, − 293 and − 324 mV to around of − 20 mV_SCE for S1, S3 and S4, respectively, with a current density increase about one order of magnitude; and from − 190 to + 200 mV_SCE for S2 sample. Therefore, the dissolved Ti (Ti ions) resulting from anodic dissolution may form the Ti complex (such as ${\left[{TiCl}_{6}\right]}^{2-}$) by reaction with Cl⁻ in the Hank´s solution (Ref 6). With higher concentration of ${\left[{TiCl}_{6}\right]}^{2-}$, the passive film could be formed. Chen et al. (Ref 6) propose that dissolution of the titanium matrix occurs simultaneously with the establishment of the TiO₂ passive film, by the presence of enough Cl⁻in solution to form the ${\left[{TiCl}_{6}\right]}^{2-}$ complex, in accordance with:

$${\text{Ti}}\xrightarrow{{6{\text{Cl}}^{ - } }}[{\text{TiCl}}_{6} ]^{{2 - }} \overset {2{\text{H}}_{2} {\text{O}}} \longleftrightarrow {\text{TiO}}_{2} + 6{\text{Cl}}^{ - } + 4{\text{H}}^{ + }$$

(6)

In addition to the TiO₂ forming the passive film on the Ti-Nb-Sn alloys in Hank's solution, the Nb₂O₅ compound and a small amount of SnO₂ oxide have been identified (Ref 5). At higher over-potential values, increasing current density with an average of 2 to 4 mA/cm² per each 1000 mV were observed, where S1 and S4 shown the lowest and highest corrosion density increasing rate. It is indicated that the TiO₂-Nb₂O₅-SnO₂ oxide layer is in an unstable state above + 1000 mV_SCE and susceptible to pit nucleation on the surface (Ref 28) with an increase in anodic current density due to the water decomposition. This phenomenon has been observed for Ti and Ti alloys (Ref 5). By increasing the over-potential, the S1 alloy showed re-passivation region from + 420 to + 750 mV_SCE, with a current density around 0.26 mA/cm² due to the migration of ions into the passive layer, as well the oxygen evolution reaction promoting the oxide reformulation. Eventually, dissolution of this oxides layer occurs at upper potentials, following similar behavior than S2 sample. Sn additions promoted higher anodic current density shown by S3 and S4 samples.

The electrochemical impedance spectroscopy (EIS) results of the alloys in Hank’s solution at 37° are presented in Fig. 9 and 10. The EIS experimental results and fitting plots (dashed-line) of the Ti-Nb-Sn alloys at the first hour and 24 hours of immersion are shown through the Nyquist and Bode plots. The Nyquist plots obtained at both immersion times (1 and 24 hours) show the characteristic mechanisms associated with three time constants, one was associated with charge transfer, capacitive and diffusion process through an irregular surface, which is composed by the β-Ti and α-Ti metallic phases (Fig. 5), principally. The two other were associated with the convergence of capacitive and diffusive mechanisms in the frequency range from the middle to the lowest. According to the Nyquist curves, at the highest frequencies, from 3000 to 100 Hz approximately, the Ti-Nb based alloy and the Ti-Nb-Sn alloys show charge transfer process as the principal corrosion mechanism, with |Z| values at 100 Hz around 50 Ohm·cm². (angle phase plots). Additionally, in this high frequencies region between 1.7 and 30 kHz, lower |Z| values of about Ohm·cm² with values of phase angle close to 5° were associated with the electrolyte resistance, as was described by Della Rovere et al. (Ref 41). While, at the frequency range from middle to the lowest (100-1 Hz, and 1-0.01 Hz, respectively, with a gradual increase in the phase angle), both resistive and diffusive mechanisms were observed, due the presence of stable TiO₂-Nb₂O₅-SnO₂ oxide layer and the diffusion of ions trough the scale thickness. That was associated with phase angle values higher than 23° at 0.1 Hz, indicating that their impedance became higher. These results suggest that the passive film formed on Ti-Nb based alloys offer promotes corrosion resistance.

Fig. 9

EIS results of the alloys in the Hank’s solution at 37 °C, at the beginning of immersion: (a) Nyquist plots, (b) Module impedance and Phase Bode plots. Here, the plots with symbols are the experimental measurements, while, the dashed-line plots correspond to the fittings applying the EEC

Fig. 10

EIS results of the alloys in the Hank’s solution at 37 °C, after 24 hours of immersion, (a) Nyquist plots, (b) Module impedance and Phase Bode plots. Here, the plots with symbols are the experimental measurements, while, the dashed-line plots correspond to the fittings applying the EEC

The S2 sample showed the highest impedance module values (|Z|: at 0.01 Hz) of about 732 Ohm·cm² at the start of immersion (Ref 23), decreasing 14 % by around 631 Ohm·cm² after 24 hours of immersion (|Z| module plots in Fig. 9b and 10b). S2, like the other alloy samples, presented charge transfer process at high frequencies (3000-100 Hz), both at the beginning and after 24 hours of immersion (zoom in Fig. 9a and 10a), but the S1 sample presented the broadest semicircle at high frequencies. The corrosion mechanism was associated mostly with the TiO₂ film formation and microstructural porosity, as shown by the results of the microstructural and mechanical characterizations.

The Sn content increasing promoted the formation of more defined semicircle at the highest frequencies (Nyquist plot). That was associated with charge transfer process at the metal/solution interface (zoom in Fig. 9a) at highest frequencies, but at middle frequencies (100-1 Hz), a capacitive and diffusive mechanism at the bottom of pores was observed (Fig. 9b). At the lowest frequencies ((1-0.01 Hz), diffusion mechanism was mostly observed, which was associated with the diagonal line with angle phase values from 30° to 50° and 70° at both initial and 24 h of immersion, respectively. At the initial immersion, the |Z| values at 0.01 Hz presented by the S1, S3 and S4 alloys were below to 540 Ohm·cm², where the S4 alloy showed the lowest impedance value (264 Ohm·cm²) showing a |Z| reduction of around 18%. These alloys displayed higher activity than the S2 alloy, as was observed by the potentiodynamic results (Fig. 8). At 24 hours of immersion, S1, S3 and S4 alloys showed similar behavior than at the beginning of the immersion, a charge transfer process at the highest frequencies was observed, and mixture of capacitive and diffusion mechanism at the middle and lowest frequencies, respectively, occurs (Fig. 10a), but the |Z| values at 0.01 Hz were lower than at the beginning of immersion, about: 317, 436 and 220 Ohm·cm², respectively (Fig. 10b).

One of the governing mechanism for the corrosion resistance of Ti-Nb-Sn alloys was by diffusion due to the formation of a protective Ti and Nb oxides layer on the surface of the alloy, limiting the diffusion of metallic and/or oxygen ions principally. The formation of the oxide layer was associated with the high content of Nb and Sn in the alloy, which promoted the formation of an oxide layer of TiO₂ and Nb₂O₅ enhancing surface passivation (Ref 22, 24). However, the corrosion resistance of the alloys decayed as a result of the formation of corrosion products scale (as TiO₂) on the phases richer in titanium which form galvanic micro-sites with the surrounding phases with higher contents of Nb and Sn. These form anodes which oxidize in the Nb₂O₅ and SnO₂. Consequently, impedance resistance decreased 18.93%, 13.80%, 19.11% and 16.35% after 24 hr of immersion for S1, S2, S3, and S4, respectively.

The proposed corrosion mechanisms were associated with the microstructural characteristics on the alloys with and without Sn content in the Ti-Nb based alloy, which have two principal phases, forming similar characteristics to an electrode with a superimposed porous microstructure layer (Ref 37, 39). The galvanic micro-cell formation at the dissimilar phases promotes the corrosion increasing. However, the corrosion of the alloys was associated with the activation mechanism at the oxide-free surface and the diffusion through the Ti, Nb and Sn oxides with major or minor protective effect. Figure 11 shows the equivalent electric circuit (EEC) that was used as an analogy of the corrosion mechanisms at the active surface with porous characteristic. The EEC, as reported before (Ref 5, 6, 23), is composed of the electrolyte resistance (R_s), set up in series with the parallel arrangement of a constant phase element CPE₁ with the passive film resistance (R_f) which grows at the alloy’s surface interphase through the activation mechanism in series with second arrangement composed with a constant phase element CPE₂ in parallel with the charge transfer resistance (R_ct) in serial with the diffusive element (Z_D) presented by the protective oxide layer inherent in the porous sites of alloys (Ref 37). The diffusion element (Warburg diffusion impedance; Z_D) was included in the ECC before R_ct, associated with the increase in corrosion products on the porous/electrolyte interface. R_f represents the dense film inside the passive film (Ref 28). As a result, a corrosion mechanism governed by transport of mass was also observed.

Fig. 11

Equivalent electric (EEC) of the corrosion mechanisms observed at the interfaces of Ti-Nb based alloys

In the EEC used here, Z_D represents the Finite Length Warburg (W_o) element. In general, the EEC described here represents the analog equivalent electrical circuit of the impedance for the defects of the surface of alloys, as a representation of porous surface (Ref 37). The elements of the EEC are defined with the following equations:

$${Z}_{CPE}={\left[Q{(j\omega )}^{n}\right]}^{-1}$$

(7)

$${Z}_{{R}_{i}}={R}_{S},{R}_{f},{R}_{ct}$$

(8)

$${Z}_{D}=\sigma \frac{coth{\left(j{T}_{D}\omega \right)}^{P}}{{\left(j{T}_{D}\omega \right)}^{P}},$$

(9)

where Q is a parameter related with the interfacial capacitance (of the constant phase element). Power n, where 0 < n ≤ 1, represents the dispersion coefficient of the capacitance loop, $j=\sqrt{-1}$ is the complex number, and ω is the angular frequency (ω = 2πf: rad/s). R_s, R_f, and R_ct are resistance elements. Hence the term T_D represents the ratio of scale thickness L and the effective diffusion coefficient D_eff of that scale (s); ${T}_{D}={L}^{2}{D}_{\text{eff}}^{-1}$, power P is between $0<P<1$, and σ is the constant of diffusion or the modulus of the Warburg resistance. The CPE is the constant phase element of the double layer, which substitutes the double layer capacitor (C_dl), due the capacitive behavior of the electrode is not ideal in real system (Ref 42). Due to corrosion products formed on porous microstructure surface, which represents a non-homogeneous surface, the CPE element was applied instead of the perfect capacitance [${Z}_{{C}_{dl}}={(j\omega {C}_{dl})}^{-1}$] for better fitting. Table 3 shows the obtained fitting values for each equivalent electrical element of the ECC used in the fitting analysis of the experimental data.

Table 3 EIS Parameters obtained by the experimental data fitting of the Ti-Nb-Sn alloys with analogous EEC (Fig. 11)

3.6 Characterization of Corroded Samples

Corroded surfaces of the Ti-Nb-Sn alloy samples after long-term electrochemical tests (24 hours of immersion) were characterized by using SEM and EDS to identify the corrosion products formed on alloy surfaces during the electrochemical tests, respectively (Fig. 12 and Table 1). Results of the SEM characterization of the corroded surfaces did not show high levels of corrosion, but the uniformity of the corrosion was evident, with a homogeneous thin scale layer, which promoted passivation behavior (Fig. 12). Additionally, Table 1 contains the report of the EDS analysis of the corroded specimens. It should be noted from data that the corrosion products are composed mainly of metallic oxides, such as TiO₂, Nb₂O₅ and SnO₂. Unsurprisingly, Sn was not found in the S1 alloy, in which appears in similar quantities both the light color phase material and dark grains. However, in the cases of the S2, S3 and S4 alloys, a tendency is identifiable; in the light color phase material, the percentage of oxygen decreases as the percentage of Sn increases in the alloys, and in the dark grains, the percentage of oxygen increases as the Sn content increases.

Fig. 12

Micrographs 5000X of the corroded alloy samples. (a) S1, (b) S2, (c) S3, (d) S4 alloys

In accordance with the electrochemical behavior results and chemical composition analysis of the corrosion products; changes in the corrosion behavior of the four Ti-Nb based alloys are attributed to the microstructural and componential changes in the corrosion layer, with passive behavior. The protective properties of passive film for Ti-Nb based alloys in Hank’s solution was attributed to its relatively high Nb content, promoting the high concentration of Nb₂O₅ present on the oxide layer of their surfaces (Ref 5). Similarly, Metikos-Hukovic et al. (Ref 43), attributed the improvement of the passive properties of the surface film by Nb addition to a decrease in the concentration of anion vacancies present on the TiO₂ layer. These anion vacancies are generated by the presence of lower titanium oxidation states. Furthermore, the addition of Nb improved resistance to dissolution, increasing the corrosion resistance of Ti-based alloys due to the formation of an Nb-enriched surface layer by the preferential dissolution of Ti (Ref 44). Therefore, Nb is significantly nobler than Ti, so that with the formation of the Nb-enriched surface layer, the corrosion resistance of the material would certainly be improved (Ref 5, 6).

The corrosion mechanism of the Ti-Nb-Sn alloys in Hank´s solution (without deaeration) was initially associated with the anodic reaction dissolution of the metal occurring at anodic sites (α-Ti phase first, followed by β-Ti phase). By contrast, the natural dissociation of the water (H₂O → H⁺  + OH^–) or by electrolysis, with the dissolution salts in the Hank’s solution, improved the redox reactions, limited by the ions diffusion thought the corrosion layer. Since, the solution was not deaerated, some dissolved oxygen content was present. Therefore, the accompanying cathodic reaction is the reduction in oxygen. The associated oxidation and reduction reactions are written as follow (Ref 45):

oxidation reactions

$${\text{Ti}} \to {\text{Ti}}^{{{2} + }} + {\text{2e}}^{ - } {1}.{63}0\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(10)

$${\text{Ti}} \to {\text{Ti}}^{{{3} + }} + {\text{3e}}^{ - }\quad {1}.{37}0\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(11)

$${\text{Nb}} \to {\text{Nb}}^{{{3} + }} + {\text{3e}}^{ - } {1}.0{99}\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(12)

$${\text{Sn}} + {\text{4OH}}^{-} \to {\text{SnO}}_{{2}} + {\text{2H}}_{{2}} {\text{O}} + {\text{4e}}^{ - } \quad 0.{945}\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(13)

$${\text{Ti}}^{{{2} + }} \to {\text{Ti}}^{{{3} + }} + {\text{e}}^{ - } 0.{9}00\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(14)

reduction reactions

$${\text{2H}}_{{2}} {\text{O}} + {\text{2e}}^{ - } \to {\text{H}}_{{2}} + {\text{2OH}}^{-} \quad - 0.{827}\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(15)

$${\text{2H}}^{ + } + {\text{2e}}^{ - } \to {\text{H}}_{{2}}\quad 0.000\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(16)

$${\text{O}}_{{2}} + {\text{2H}}_{{2}} {\text{O}} + {\text{4e}}^{ - } \to {\text{4OH}}^{-} \quad 0.{4}0{1}\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(17)

$${\text{O}}_{{2}} + {\text{4H}}^{ + } + {\text{4e}}^{ - } \to {\text{2H}}_{{2}} {\text{O}} \quad {1}.{229} \quad \left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(18)

In accordance with the Ti-Nb based alloys microstructure, (Fig. 5) the β-Ti and α-Ti phases are present. At the near neutral pH solution and E_corr potentials between (− 320 to − 100 mV _{vs SHE}) (Fig. 8), the α-Ti phase (which has higher content of Ti) followed by the β-Ti phase, react forming corrosion, with the TiO₂ amorphous layer (hydrated TiO₂), Nb₂O₅ (Ref 43, 46, 47) and SnO₂ (Ref 24, 47) oxides at the surface. In this regard, D. Sri Maha Vishnu et al. (Ref 22) suggested that the formation of the protective oxide layer should occur gradually, by first transforming the metals into their sub-oxides and then further into their highest-valance oxides. For Ti (Ref 24, 47), Nb (Ref 22, 43) and Ti metal oxides, the proposed general electrochemical reactions, are the following:

$${\text{Ti}} + {\text{H}}_{{2}} {\text{O}} \to {\text{TiO}} + {\text{H}}_{{2}} \quad {1}.{43}0 \quad \left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(19)

$${\text{2TiO}} + {\text{H}}_{{2}} {\text{O}} \to {\text{Ti}}_{{2}} {\text{O}}_{{3}} + {\text{H}}_{{2}} \quad 0.{886} \quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(20)

$${\text{Ti}}_{{2}} {\text{O}}_{{3}} + {\text{H}}_{{2}} {\text{O}} \to {\text{2TiO}}_{{2}} + {\text{H}}_{{2}} \quad 0.{564} \quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(21)

$${\text{Nb}} + {\text{H}}_{{2}} {\text{O}} \to {\text{NbO}} + {\text{H}}_{{2}} \quad 0.{8}0{7}\quad \left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(22)

$${\text{NbO}} + {\text{H}}_{{2}} {\text{O}} \to {\text{NbO}}_{{2}} + {\text{H}}_{{2}} \quad 0.{575}\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(23)

$$2{\text{NbO}}_{{2}} + {\text{H}}_{{2}} {\text{O}} \to {\text{Nb}}_{{2}}{\text{O}}5 + {\text{H}}_{{2}} \quad 0.266\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(24)

$${\text{Sn}} + {\text{2H}}_{{2}} {\text{O}} \to {\text{SnO}}_{{2}} + {\text{2H}}_{{2}} \quad 0.{117}\quad\left( {{\text{E}}^{0} /{\text{V}}} \right)$$

(25)

At higher over-potentials the evolution of oxygen occurs, and the oxygen evolution on Ti and Nb occurs at the same potentials (Ref 48). Chemical analysis of the corroded surfaces showed quantities of oxygen associated with the oxides of Ti and Nb as: TiO₂ (Ref 28,29,30) and Nb₂O₅ principally, and minor quantities of SnO₂ (Table 1). At the early corrosion state, α-Ti phase promotes the formation of titanium oxides preferentially, due to its higher concentration in the phase (Table 1), in accord with Petr Vanýsek (Ref 45). This is in keeping with the corrosion mechanism by activation at the metal/solution interface observed from EIS measurements at the highest frequencies (zoom in Fig. 9a), which has been attributed to the passive film resistance (R_f) on the porous interphase growing over the α-Ti phase of the alloy microstructure. Where the highest resistance value was below 42.3 Ω cm², it was presented by the S3 sample alloy (Table 3).

Once the first titanium oxide layer was formed on the α-Ti phase, the Ti and Nb oxides start growing over the β-Ti phase, due to an evident change in the galvanic potential between the TiO₂ (over the α-Ti phase) and the surrounding β-Ti phase. Thus, the TiO₂ and Nb₂O₅, as well as the SnO₂ oxide layers are formed and begin to increase, which was attributed to the sum of the resistance elements (R_f + R_ct) and the resistive parameter of the diffusive element [Z_D = f(ω)]. Corrosion grown on both the Ti phases, promoted the passivation behavior presented by the potentiodynamic measurements, as well as by the EIS results. Once corrosion products had covered the surface of both Ti-phases, passivation or limited current density behavior was observed (Fig. 8, 9, 10). Thus, the corrosion mechanisms of Ti-Nb based alloys in Hank´s solution were governed by the resistive-capacitive impedances at middle frequencies, and by diffusion observed at lower frequenciesbecause of the formation of the TiO₂-Nb₂O₅ protective oxides scales.

Although, the S2 alloy sample showed the highest |Z| values, it was not much higher than that presented by the other alloys, but those |Z| values were attributed to the passivation behavior of the oxide film grown, with lower of defect concentration in the titanium oxide with the addition of Sn in 6 at.%. As well, the highest corrosion resistance shown by the S2 alloy was attributed to its microstructure and mechanical properties on the microstructure cohesion, lower superficial porosity, higher density and higher elasticity module than the Ti-Nb-based alloys with higher Sn addition, here assessed (Fig. 7). These microstructural characteristics allowed a cohesive and stable oxide layer formation with lower structural defects. Consequently, a continuous passive layer was formed, increasing the value of the sum of the resistance element (R_f + R_ct) to about 755.7 Ω·cm² at the beginning of immersion, which decayed after 24 hours to 390.9 Ω·cm² (Table 3, Fig. 9b, 10b). That was associated with anodic dissolution of the oxide layer during the immersion. This tendency was also observed in the other alloy samples.

However, the increase Sn content was not shown to increase corrosion resistance. Ti-30Nb base alloy, and the Ti-30Nb-9Sn and Ti-30Nb-12Sn, showed low the sum of the resistance element (R_f + R_ct) values (less than 260 Ω·cm²) associated with more active behavior. Similarly, resistance of the diffusive element [Z_D = f(ω)] becomes representative below the frequency of 1 Hz (Fig. 13), starting from impedances of 40 Ω·cm² (just higher than R_s values) to values close to 1000 Ω·cm² at 10 mHz generally for the four alloys at both times of immersion. Z_D curves were obtained using ZView software, simulating Warburg (W_o) element (Eq. 11) with the data shown in Table 3. The defects density observed in these alloys in porous form (obtained by powder sintering) was responsible for these lower resistance values of the S1, S3 and S4 alloys. The increase in surface area represented an increase in the electrochemical activity inside of the pores or voids, which could be useful for biomedical applications such as bone implants for further study. Similar studies of the performance of Ti-Nb-Sn alloys in SBF (Ref 49) suggested adequate mechanical, ion release and corrosion performance for use as biomaterials for orthopedic implants. Additionally, the development of Ti-Nb-Sn scaffolds and solid forms (Ref 50) show potential for orthopedic and tissue engineering applications. In regard to Torres-Sanchez et al. (Ref 51), their results confirm its suitability as a biocompatible and bioactive material for bone regeneration applications.

Fig. 13

Plots of frequency vs impedance of the diffusive element Z_Wo (Z_D) of the alloys in the Hank’s solution at 37 °C, (a) 1st and (b) after 24 hours of immersion

Similar alloys synthetized by Aguilar et al. (Ref 9), showed a non-homogenous distribution of irregularly-shaped macro pores and micro pores. Similar shapes were observed in the foams presented in the current work (Fig. 1). Moreover, the specimens obtained in this work show defects resulting from the melting of the Sn (Fig. 3), indicating a disadvantage of the powder metallurgy method compared with the additive manufacture method (Ref 19, 20). The milling procedure promotes the formation of β-Ti and other phases which coexist in the resulting alloys due to the MA process which increases the solubility of the Ti, Nb and Sb elements. This can be verified from XRD analyses (Fig. 2), which has been reported also (Ref 9, 10, 15, 17)

The results presented in this study support new findings about Ti-Nb-Sn alloys obtained by powder metallurgy and sintering methods. The main advance of this work, compared with its antecedents (section 1), is the addressing of corrosion behavior and mechanisms, which both correspond and contrast with previous research, such as corrosion of the Ti-6Al-4 V alloy within acid, salt and biological fluids reported by Cui et al. (Ref 28,29,30), which resulted in the pitting of the passivating layer formed in the alloy surface, in the contrast to the corrosion seen in Ti-Nb-Sn alloy which did not show any evidence of pitting.

4 Conclusions

In this research, Ti-30Nb-xSn alloys (x = 0, 6, 9, 12) were obtained by mechanical alloying and powder sintering. The obtained alloys displayed mechanical properties desirable for medical implants, as well as increased anti-corrosion performance due to the addition of Sn. These alloys show potential as biomaterials in biomedical applications.

The distribution of elements within the powder alloys by mechanical alloying was homogeneous, and the addition of Sn promoted a change in microstructure, increasing the spherical grains slightly. The addition of Nb caused the formation of β-Ti phase, since Nb is a stabilizer of this phase, while Sn increased the β-Ti phase formation. However, it was not possible to increase the proportion of this phase by adding Sn beyond 9 at.%.

After sintering, the alloy microstructure was composed of two Ti phases, one rich in Nb and Sn (β-Ti phase) and another lacking in these elements (α-Ti phase). The presence of Sn promoted the refining of alpha grain and defect formation due to the accumulation of melted Sn. In addition, the porosity of theses alloys is between acceptable levels of human bone porosity and, although their density and elastic modulus are larger than the average values of these properties, the obtained values are acceptable compared to current generation materials for medical implants. Furthermore, these properties can be improved by modifying the porosity of the alloy.

Preliminary observation of the potentiodynamic behavior was attributed to the compactness of the samples after the sintering treatment, i.e., the porosity as well as the density of the Ti-Nb-Sn alloys. The four alloys, displayed similar corrosion behavior, where the sample S2 presented the lowest current density due its lower surface porosity, consequently, it showed highest corrosion resistance in Hank´s solution. However, the increase in the surface porosity shown by the other alloys promoted an increase in current density. However, increase in porosity concentration, allows potential application in biomedical implants. All alloys showed semi-passive or limited current density due the film composed of TiO₂, Nb₂O₅ and SnO₂, although, the content of Sn did not present an observable effect on the anodic polarization. In accordance with EIS results, the corrosion mechanisms of Ti-Nb based alloys in Hank´s solution were governed by charge transfer process at highest frequencies, at the frequency range from middle to the lowest, both resistive and diffusive mechanisms were observed, due the presence of stable TiO₂-Nb₂O₅-SnO₂ oxide layer and the diffusion of ions trough the scale thickness.

For further applications, it is contemplated to carry out evaluations of the corrosion kinetics of Ti-Nb base alloys, as well as, the development and study in similar conditions of the Ti-Nb-Sn base alloys with porous microstructure, for biomedical applications as implants in bone structure, and as a possible application in bone tissue regeneration.