3.1. Age-Hardening Response and Microstructure of the Alloys
Figure 1 illustrates the evolution of microhardness with aging time for the ZX11 alloy. The initial hardness value of the as-cast ZX11 alloy is approximately 62 HV. After solution treatment, this value decreases to around 56 HV. Subsequently, after undergoing an isothermal aging process at 180 °C for 3 h, the hardness reaches a value of about 71 HV. Beyond this point, the hardness decreases with further aging.
For the wear study, samples in four different conditions were tested (
Table 2): as-cast state; solution-treated; solution-treated and subsequently aged for 3 h (referred to as peak-aged condition); and solution-treated and subsequently aged for 24 h (referred to as over-aged condition).
Figure 2 displays the microstructure of the various tested conditions of the ZX11 magnesium alloy. The constituent phases of this alloy in the as-cast condition were characterized in depth in a previous study [
16]. In this condition (
Figure 2a), the precipitated particles formed a quasi-continuous phase at the grain boundaries (GBs). Within this phase, two types of Zn-Ca precipitates with different contrast were observed: Ca
2Mg
6Zn
3 (bright particles) and Mg
2Ca (dark particles). These precipitates are commonly found in many Mg-Zn-Ca alloys, and for the ZX11 in the as-cast condition, the Mg
2Ca appeared as discrete particles that were embraced by the Ca
2Mg
6Zn
3 phase due to the solidification process. The Mg
2Ca particles possess a higher melting point, and, during solidification, they were the first formed. Subsequently, the Ca
2Mg
6Zn
3 phase precipitated around the binary phase, embracing these particles. Meanwhile, the Mg-matrix solidified and lost alloying elements from the precipitates, which finally formed at the GBs and resulted in the microstructure shown.
Figure 3 shows an EDS mapping analysis of the Ca
2Mg
6Zn
3 and Mg
2Ca precipitates within the alloy.
The solution treatment (
Figure 2b) caused the full dilution of the Ca
2Mg
6Zn
3 phase, while only partially diluting the Mg
2Ca particles, with solely the dark contrast phase being detectable. This difference is caused by the lower melting temperature of the Ca
2Mg
6Zn
3 phase. The continuity of precipitated phases at GBs disappeared because they were provided by the ternary phase that dissolved. As a result, after the solution treatment, only the Mg
2Ca precipitates appeared at GBs.
In the aged conditions, the general morphology exhibited similarities to the solution-treated state, but with larger precipitates. In the peak-aged condition (
Figure 2c), the precipitates increased in size and number. In addition to the precipitated phases observed in the solution-treated condition, the microstructure of the alloy also revealed the presence of new particles within the grains resulting from the precipitation of the previously dissolved ternary phase. In the over-aged (
Figure 2d) condition, the particles observed in the previous condition were retained. This indicates that most of the elements that were initially dissolved in the matrix had already precipitated around the particles formed during shorter heat treatment times. The presence of these particles in the over-aged condition suggests that the alloy had reached a state where the majority of the precipitates had formed and further aging did not significantly alter their composition or distribution. Hence, the slight decrease in hardness can be attributed solely to the enlargement of the grain size within the microstructure [
20].
3.2. Wear Rate
The wear tests were conducted following the procedure and the configuration indicated in the experimental section, using a counterbody made of AZ31 magnesium alloy, while previous studies employ steel as the counterbody material. It is important to notice that in biomedical applications, it is not common to find Mg in direct contact with steel. Therefore, the steel–magnesium tribological system is not relevant to biomedical applications, apart from the fundamental knowledge that those studies provide. Continuous contact between materials with similar natures, such as two types of magnesium alloys, either identical or similar, is more common in real applications. In our study, the alloy selected as the counterbody is the AZ31 Mg alloy that is the most commonly used alloy in biomedical applications. Its composition, particularly due to the presence of Al, enables the identification of any material transfer from the counterbody to the tested samples. This selection ensures that the wear tests are closer to the biomedical scenarios.
Figure 4 shows the volumetric wear rates of the ZX11 alloy under different conditions after a sliding distance of 200 m. For most conditions, the application of heat treatments resulted in a reduction in the wear rate.
For a load of 2 N and a sliding speed of 0.05 m s
−1 (
Figure 4a), the wear rate was reduced by 32% for the solution-treated condition and by 60% for the peak-aged condition, compared to the as-cast state. However, the application of an over-aging treatment increased the wear rate, bringing it closer to the wear rate observed in the solution-treated condition, which was 32% lower than that of the as-cast samples.
For the 2 N load, the increase in the sliding speed to 0.1 m s
−1 increased the wear rates (
Figure 4b). The application of a solution treatment reduced the wear rate by 45% compared to the as-cast state. In the peak-aged condition, the wear rate was only slightly below (8%) that of the as-cast state. The application of an over-ageing treatment resulted in a 30% lower wear rate.
These findings indicate that for the 2 N load, heat treatments can effectively reduce the wear rate of the ZX11 alloy, with the peak-aged condition exhibiting the lowest wear rate for a sliding speed of 0.05 m s−1 and the solution-treated condition at a sliding speed of 0.1 m s−1.
For higher loads, i.e., 20 N, the evolution of the wear rates was different from that observed at lower loads (
Figure 4a,b). At a sliding speed of 0.05 m s
−1, the wear rate of the solution-treated samples increased by 14% compared to the as-cast condition. However, the wear rate of the peak-aged samples was reduced by 18%, and the over-aged samples showed wear rates similar to the solution-treated ones.
Similarly, at the conditions of 20 N load and a sliding speed of 0.1 m s−1, the results showed a reduction in wear rates of 14% for the solution treatment and 35% for the peak-aged samples. The over-aged samples showed a wear rate equivalent to that of the as-cast samples.
The dispersion of data from the wear tests observed for the 20 N load and 0.1 m s−1 of sliding speed was much greater than for the other tests. The presence of a significant statistical deviation is typically indicative of instabilities on the wear process, as well as material transfer and loss from the samples, as shown below.
In general, wear increases with the applied load in most alloys, and it was observed for the ZX11 alloy for most of the conditions except for the as-cast samples. This suggests that the wear behavior of the ZX11 is influenced by its microstructure, which is particularly different in the as-cast condition, where the quasi-continuous network of precipitated phases at GBs provides a preferential path for crack propagation.
Additionally, the wear rates of the ZX11 alloy increases with sliding speed for the ZX11 alloy in most conditions, although wear is not affected much by the sliding speed in the solution-treated condition. In the case of the ZE41A alloy [
27], wear rates have a double dependence on sliding speed; at very low speeds, the wear rate is higher due to the predominance of oxidative wear; it decreases at medium speeds; and then it increases again at higher speeds where abrasive wear becomes dominant.
Other alloys, such as the AZ91 and AZ91 modified with up to 3 wt.% of lanthanum-based rare earths, showed wear rates similar to those of the as-cast ZX11 alloy, but lower than the measured ones in the peak-aged condition [
28]. However, they used a steel counterbody, which does not make possible a straightforward comparison of the results. In a similar system (AZ91 with 2 wt.% Y), but obtained by extrusion, they observed that the application of a T6 heat treatment resulted in a reduction in the wear rate that was greater for higher loads (15% reduction for 5 N loads and 35% reduction for 70 N) [
29]. Nevertheless, they did not provide the results for the solution-treated alloy, so the effect of the intermediate stage was not shown. Also, a Mg-Zn-Y-Zr alloy showed lower wear rates in the initial state, and the application of heat treatments reduced them even more [
22]. This observation is similar to what was observed in the ZX11 alloy, in which the as-cast samples showed a quasi-continuous phase at GBs and presented greater wear rates than the conditions with distributed precipitated particles.
Figure 5 shows the specific wear rate (
k) for all the tested conditions, which is determined by dividing the wear rate (
Figure 4) by the load used following Archard’s law (Equation (1)). In all cases, the specific wear rate was strongly dependent on the applied load. The specific wear at the load of 20 N was considerably lower across all conditions than for the 2 N condition, regardless of the sliding speed employed. The values observed ranged from 0.002 to 0.008 mm
3 N
−1 m
−1 for all 2 N conditions, while for the 20 N load test, they were one order of magnitude lower, ranging from 0.0002 to 0.0008 mm
3 N
−1 m
−1.
The
k values were lower for the sliding speed of 0.05 m s
−1 than for 0.1 m s
−1 as is extracted from the comparison of
Figure 4a,b. Therefore, the sliding speed has an impact on the specific wear rate. As observed, Archard’s law was not followed, since abrasive wear did not predominate. The condition with the lowest
k value at 0.05 m s
−1 was the peak-aged at both loads. At 0.1 m s
−1 of sliding speed, the
k value was minimal for the solution-treated condition at 2 N of load and the peak-aged one at 20 N of load. It is suggested that changes in the wear mechanisms take place.
The specific wear rate (
k) is commonly used to assess the influence of load on the wear behavior and it is a sensitive factor to identify changes in the wear mechanisms. In the case of the ZX11 alloy, it has been observed that wear (
Figure 4) is nearly unaffected by the load, causing the
k values to reduce with the load (
Figure 5). This behavior is different from what is typically observed in many systems [
30,
31]. There are two considerations to be made on this observation.
On the one hand, a Mg-based alloy has been used as the counterbody, i.e., the hardness of the two materials involved is similar. In most wear tests, a hard material is used as the counterbody, favoring the appearance of two-body abrasive wear [
32]. However, this situation is not realistic within the human body, as no hard metals are present. The presence of similar materials implies that many other wear mechanisms could be present, as the evolution of
k indicates.
On the other hand, the heat treatments applied have an impact on the wear rate of the ZX11 alloy. The presence of a quasi-continuous phase at the GBs in the as-cast condition caused fatigue wear due to crack propagation through the precipitated phases. This wear mechanism is not observed after the different heat treatments, due to the dissolution of the Ca2Mg6Zn3 phase at the GBs. Therefore, the evolution of the microstructure through the different heat treatment conditions changes the response of the alloy during its contact with the counterbody.
3.3. Friction Coefficient
The friction coefficients obtained for the ZX11 alloy samples were measured and recorded during testing, and the results are shown in
Figure 6. In all tested samples, an unstable friction coefficient was observed for approximately 50 m of sliding distance. After this initial phase, the friction coefficient reached a relatively constant value.
In the steady state, the friction coefficients measured ranged between 0.27 and 0.46. It is important to indicate that the evolution of the friction coefficient with the heat treatments follows a similar trend to that of the wear rate.
The values observed for the friction coefficient are similar to those measured in wear tests where a magnesium alloy is tested against other metals in dry-sliding conditions. Ramesh et al. [
33] obtained a similar average friction coefficient value for different Mg-Zn alloys applying 20 N and 2000 m of sliding distance using a pin-on-disc configuration with a steel disc as the counterbody. Zhao et al. [
34] evaluated the friction coefficient of the as-cast and heat-treated Mg-6Gd-2Zn-0.4Zr (wt.%) alloy using a ball-on-disc configuration with stainless-steel balls, a sliding velocity of 20 mm s
−1, a load of 20 N, and a 60 min test duration. Under these test conditions, they observed a very slight difference in the friction coefficient value between the as-cast and the heat-treated samples, ranging from 0.30 to 0.40 in all cases. Zhou et al. [
35] showed in their work an average friction coefficient value of 0.3187 for an as-cast AZ91D alloy using a pin-on-disc configuration and steel balls, applying 10 N and 300 rpm for 15 min. Blau et al. [
36] obtained friction coefficient values in the range of 0.30 to 0.35 for the AZ91D alloy manufactured by casting and thixomolding using three tribometers with different characteristics.
This research indicates that the combination of the ZX11 alloy with the AZ31 alloy produces friction levels similar to those observed between other Mg alloys and steel. Friction can arise either from abrasion or from the formation of micro-welds. In the case of the abrasive wear mechanism, one material penetrates into the other, and this mechanism is more frequent when a harder material comes into contact with a softer one [
31]. In such cases, the friction is primarily limited by the shear resistance of the softer material, and the hardness of the harder material usually has limited impact. This mechanism causes the extraction of material from the samples and the formation of scratches and debris.
Another type of interaction between the testing material and the counterbody occurs when metallurgical unions are formed at the surface, and the ridges of both surfaces contact and weld. In this case, the upper limit of friction is associated with the lower shear resistance between the joint, the substrate, and the counterbody. This phenomenon causes material transference between the sample and the counterbody.
As a result, friction in the combination of the ZX11 alloy with the AZ31 alloy is limited by the shear resistance of both alloys. The evaluation of the composition at the surface of the samples can indicate if material transference is taking place, and the study of the surface morphology and of the debris formed during the wear test provides relevant information on the wear mechanism affecting the contact.
3.4. Determination of the Elementary Composition of the Surface
The surface composition of the ZX11 samples was examined using EDS analysis to assess any modifications caused by the wear tests. The evolution of the surface can come from the oxidation of the surface or from transference of material from the counterbody.
Figure 7 represents the atomic percentage of oxygen detected on the tested sample surfaces. The results indicate that all the samples exhibited oxygen levels above 25%, but significant variations in the oxygen proportion were observed. In most conditions, a higher oxygen content was observed in tests conducted with higher loads, implying the formation of a greater quantity of oxides on the surface, which were also more stable.
Additionally, it was observed that at the low sliding velocity, the wear rates (
Figure 4a) and oxygen content on the surface (
Figure 7a) exhibited a proportional relationship. The friction coefficients of the ZX11 alloy at the low sliding velocity (
Figure 6a) show an inverse relationship with the presence of oxygen (
Figure 6a). This indicates that the oxides formed during the wear test have a lubricant effect during the wear test that resulted in a lower wear rate. Also, the as-cast sample shows a different behavior than the other systems, with a reduction in the amount of oxygen when the load applied increases, while the effect of load was lower for the samples in the peak-aged condition.
The results obtained at the highest speed (
Figure 4b and
Figure 7b) do not follow the same tendency. The amount of oxygen on the surface was lower and increased with the applied load (
Figure 7b), but there is no clear correlation with the friction coefficients or the wear rate. This suggests that there are other wear mechanisms involved in the evolution of the surface of the samples during the wear tests.
Figure 8 shows the relationship between the wear rate and the friction coefficient with the atomic percentage of oxygen. In general, the increase in the wear rate corresponds to an increase in the O
2 content found. In the case of the friction coefficient, the observed trend is the inverse. As the friction coefficient decreases, so does the atomic percentage of oxygen.
Aluminum was detected on the surfaces of the tested samples, even though the ZX11 alloy does not contain Al in its composition. However, the counterbody material, the AZ31 magnesium alloy, contains a small proportion of Al. Hence, the presence of aluminum on the surface of the tested samples indicates material transference from the counterbody to the ZX11 samples. The quantity of aluminum observed on each specimen is shown in
Figure 9, with values exceeding 1% of aluminum content. Considering that the counterbody contains 3 wt.% of aluminum, an aluminum proportion of 1% suggests that approximately one third of the sample surface is covered by material transferred from the counterbody.
For applied loads of 2 N in the as-cast and solution-treated ZX11 alloy (
Figure 9a,b), as well as for the 20 N in the as-cast 0.05 m s
−1 and solution-treated 0.1 m s
−1 conditions, the amount of content remained below 0.25%. However, in the remaining conditions, the aluminum content ranged from 0.50 to 1.00%. In general, there is a proportional relationship between the presence of aluminum and oxygen, indicating the possibility of a similar mechanism contributing to their appearance.
3.5. Wear Mechanisms
To explain the differences observed in the wear processes of the tested samples, an analysis of the wear mechanisms present in each condition has been conducted. To evaluate the predominant mechanisms, both the surface of the samples and the debris formed have been studied.
Figure 10 shows the worn surfaces of the as-cast ZX11 alloy pins under the various test conditions applied. In all conditions and samples, fine grooves parallel to the sliding direction were visible, indicating the occurrence of abrasion (red arrow). Additionally, the presence of oxygen in all sample surfaces indicates the occurrence of oxidation wear (green arrows). Oxidation wear is common in magnesium alloys due to its high reactivity and is characterized by the formation of spots or platelets on the magnesium alloy surface. However, the observed behavior exhibits distinct characteristics among the different samples studied.
Figure 10a–d show two types of oxides with different contrast. The darker oxide is enriched in aluminum, indicating the presence of material transferred from the AZ31 counterbody to the pin surface. This oxide is more visible in
Figure 10a–c because it does not cover the entire surface. However, its proportion is significantly higher in the 20 N–0.1 m s
−1 condition (
Figure 10d), where it completely covers the surface.
At the lower speed and with an increase in applied load, in addition to abrasion and oxidative wear, deformation layers of the alloy become noticeable, and delaminated zones were observed (inset in
Figure 10b—blue arrow). At higher speeds, a more continuous surface is observed, indicating the occurrence of plastic deformation (yellow arrow) [
17,
18]. The 2 N load-tested sample (
Figure 10c) showed more oxides than the 20 N tested one (
Figure 10d).
Figure 11 presents evidence of the main wear mechanisms observed in the ZX11 solution-treated samples under various test conditions. Abrasion grooves and the presence of oxygen were observed on all worn surfaces. Additionally,
Figure 11a reveals cracks perpendicular to the sliding direction, indicating the occurrence of delamination wear. In this wear mechanism, subsurface cracks propagate and extend until they reach the surface as cracks that are perpendicular to the sliding direction.
An increase in applied load (
Figure 11b,d) resulted in rougher surfaces, which correspond to higher wear rates, as illustrated in
Figure 3, and a greater amount of material transferred from the counterbody (
Figure 9a,b). A more homogenous surface was observable in the 2 N–0.1 ms
−1 condition (
Figure 11c). No signs of plastic deformation were observed in the solution-treated alloy, unlike the as-cast alloy.
Figure 12 shows that the peak-aged ZX11 alloy samples tested under various load and sliding velocity conditions exhibit a rough surface because of the presence of abrasion and oxidation, with oxidation being the dominant mechanism. These samples also exhibited the highest hardness. High contents of O and Al at the surface (
Figure 7 and
Figure 9) indicate oxidation and material transfer from the counterbody to the pin surface.
According to several studies [
17,
37,
38,
39], a mechanically mixed layer (MML) of oxides forms at the surface, incorporating material from the counterbody. These oxide layers possess high hardness values and remain stable within certain ranges of load and sliding velocity. If the oxide layer formed is detached during the test, the wear rate increases, as a new wear mechanism appears that favors the material removal from the sample surface. However, the presence of a stable MML layer reduces both the wear rate and the friction coefficient. Under more energetic conditions, deformation of the material at the surface is observed, which is attributable to the high temperatures reached in the contact zones between the materials.
Figure 13 shows the worn surfaces of the ZX11 alloy in the over-aged condition. Oxidation wear was also the predominant mechanism, along with abrasion. The wear rates of the over-aged samples (
Figure 4a,b) can be attributed to their hardness, as explained for the peak-aged ZX11 alloy samples. The over-aged samples present a hardness value of about 69 HV, which is higher than the hardness of the counterbody employed in the wear tests (64 ± 3 HV). This difference in hardness likely contributes to the observed wear rates.
Figure 14 shows the debris generated in the wear tests of the as-cast ZX11 alloy. The debris found presents two different morphologies with differentiated compositions (marked with the letters A and B in
Figure 14). The EDS mapping in
Figure 14 reveals that the debris marked as A exhibits a higher Al content, indicating that it comes from the AZ31 alloy counterbody. These debris particles are formed through abrasion from the counterbody during the wear test, resulting in an elongated, curved, or spiral morphology. On the other hand, debris marked as B has a higher O content and is associated with the oxide layer formed during the wear test. These oxidized particles initially have a small size and irregular morphology, and agglomerate due to repeated sliding, forming larger particles and even a surface layer with some protective properties against wear. These two types of debris were observed throughout the different heat-treated tested samples.
Figure 15 and
Figure 16 provide information on the O and Al (atomic %) content of the debris formed in each test under the different load and sliding velocity conditions. The results suggest that at lower loads and sliding velocities, the oxidative phenomenon dominates, as was observed in other alloys such as AM50B and AZ91 [
17,
18].
Figure 16 indicates that the analyzed debris from all samples presents a similar Al content. This indicates that material from the AZ31 alloy used as the counterbody was removed in all test conditions for all the heat-treated ZX11 samples.
In general, the microstructural changes of the ZX11 alloy after the heat treatments applied do not always result in improved wear resistance. However, the heat treatments prevent the plastic deformation wear mechanism in the ZX11 magnesium alloy. Shanthi et al. [
39] found that grain refinement of the recycled AZ91 alloy significantly improved the mechanical properties of the alloy but did not necessarily improve its dry sliding wear resistance. The increased wear rate of the new samples, compared to the as-cast AZ91 alloy, was attributed to the brittleness, which promotes the abrasion mechanism. The presence of an MML oxide layer and its protective nature level of protection did not seem to be correlated with the grain size or the pin properties.
Additionally, the presence of a quasi-continuous phase at grain boundaries in the studied ZX11 alloy appears to facilitate wear evolution through the crack propagation within the material. This morphology, which is interesting for the control of corrosion in chloride media [
16], does not inhibit the formation of oxides at the surface. Consequently, both mechanical and oxidative wear mechanisms are favored in the presence of the as-cast morphology. The breakdown of the continuous morphology of precipitated phases at the GBs into discrete precipitates reduces the effect of fatigue wear mechanisms, while the increase in hardness reduces abrasion. Therefore, the peak-aged condition exhibits the best wear behavior among the tested conditions.