3.1. Coatings Microstructure
The Al12Si/SiC coating microstructure analyzed by SEM is shown in
Figure 1. A good metallurgical interface between the substrate and coating can be observed in the cross-section (
Figure 1a). The deposition efficiency can be calculated from the ratio between the deposited cross-section area, the scanning speed, the sprayed powder density, and the powder feed ratio [
27].
According to this equation, the deposition efficiency of the Al12Si/SiC coating was 19%. This value is similar to the efficiencies obtained by other deposition techniques of composite materials such as [
28,
29].
The microstructure was characterized in the zones that showed different compositions (
Figure 1b,c). In both cases, the composite matrix was mainly formed by dendrites of Mg
2Si and a eutectic constituent formed by Al
12Mg
17-Mg and Al
12Mg
17 [
19]. These dendrites can be observed in
Figure 1d. Some zones of the coating (
Figure 1b) present a higher percentage of magnesium than others (
Figure 1c). In
Figure 1b, it can be seen that in zone 1, the only phase apart from the matrix is the SiC reinforcement particles. However, in
Figure 1c—where a detail of zone 2 of the coatings can be observed—Al
4C
3 and Si, apart from the SiC and the matrix, can be observed. The reason for these differences observed in the coating is that during the fabrication process, the solidification was too fast and the diffusion flows were not enough to incorporate the diluted magnesium from the substrate homogenously into the coating [
19].
The addition of Si or Ti to the alloy changes the microstructure of the coating. The Al40Si/SiC and Al12Si-Ti/SiC coating presents a good metallurgical substrate–coating interface, as can be seen in
Figure 2a,b. The deposition efficiency of the Al40Si/SiC coating was 33%, and 19% in the case of the Al12Si-Ti/SiC coating. As in the case of Al12Si/SiC, for the Al40Si/SiC coating (
Figure 2c,e) the composite matrix was formed by Mg
2Si, Al
12Mg
17-Mg, and Al
12Mg
17. In addition, the SiC particles were not degraded and there was no presence of Al
4C
3. However, the SiC matrix interface showed discontinuities because the wettability decreased [
19]. The Al12Si-Ti/SiC coating (
Figure 2d,f) was characterized by a Mg
2Si, Al
12Mg
17-Mg and Al
12Mg
17 matrix. Ti only reacts with Al and Si to form TiSi
2 and AlTi
3. There was no eutectic Si due to all of the silicon reacting with titanium and magnesium [
19].
3.2. Wear Testing
Figure 3 shows the wear test measurements of the different samples using a 10 N load, a 4 mm diameter, and 200 rpm for a total wear length of 1000 m.
The average microhardness is shown in
Figure 3a. All of the coatings have a higher microhardness than the uncoated ZE41 magnesium alloy. The microhardness increase was higher than ×4 in the case of Al12Si/SiC and Al40Si/SiC, and ×3.5 in the case of the Al12Si20Ti/SiC with respect to the uncoated ZE41. The hardness is similar in Al12Si/SiC and Al40Si/SiC, but the standard deviation is lower in the Al40Si/SiC coatings because their microstructure was more homogenous than that in the Al12Si/SiC. In addition, despite Al
4C
3 increasing the coating’s hardness, these components could be hydrated, which can reduce the mechanical properties of the coating. The Al12Si20Ti/SiC coating’s microhardness is slightly lower than the others because the Al-SiC-Ti reactivity forms phases with lower hardness.
The friction coefficient is shown in
Figure 3b. The friction coefficient is higher in the Al12Si/SiC and Al40Si/SiC coatings than in the ZE41 substrate and Al12Si20Ti/SiC (which have similar friction coefficients). The Al12Si/SiC friction coefficient is 27% higher than that of uncoated ZE41. The presence of Al
4C
3 in the Al12Si/SiC coating produces a higher friction coefficient because the Al
4C
3 breaks loose and produces abrasive wear. In the Al40Si/SiC coatings, the friction coefficient is 18% higher than that of the ZE41 substrate. In this case, it is because the primary silicon particles break loose. In the Al12Si20Ti/SiC coatings, the friction coefficient is similar to the uncoated ZE41. The breaking loose of small phases is lower and, consequently, the friction coefficient is similar to that of the substrate.
Figure 3c shows the friction coefficient evolution as a function of the wear distance. A stabilization of the friction coefficient value from the first 100 m of the wear test can be seen.
Figure 3d shows the wear rate for the substrate and coatings on a ZE41 magnesium alloy. The coated materials’ wear rates were 40% lower than the uncoated substrate. Generally, the wear rate is inversely proportional to the microhardness. The differences in the wear rate of the coatings are due to the matrix-reinforcement interface resistance, apart from the microhardness [
30]. The absence of Al
4C
3 in the Al40Si/SiC and Al12Si20Ti/SiC coating microstructure produces a higher matrix–reinforcement interface resistance and a lower wear rate (10.8% and 7.1% lower, respectively, in comparison with the Al12Si/SiC coating) despite the lower microhardness (1.25% and 12.85%, respectively).
In addition, the Archard coefficient (
Figure 3f) allows the determination of the presence of different wear mechanisms. The wear mechanism is different between the substrate and the coatings.
3.3. Wear Mechanisms
In order to identify the different mechanisms that take place during the wear of the substrate and the coatings, the worn surfaces and the wear debris were observed by SEM. In addition, the evolution of the wear mechanism was evaluated by comparing the worn surface to the sliding distance (200, 500, and 1000 m).
Figure 4a shows a comparison of the SEM micrographs of the worn surfaces of the uncoated ZE41 and the different coatings (Al12Si/SiC, Al40Si/SiC, and Al12Si20Ti/SiC) against the sliding distance.
In all cases, many fine grooves parallel to the sliding direction can be observed on the worn surface, which is proof of the presence of an abrasion mechanism. These lines are usually caused by the presence of hard particles that plough into the pin. The movement of these particles over the surface causes the removal of the material along its path on the surface. Embedded particles in the worn surface can be observed in some cases (
Figure 4b). In some conditions, the loss of the surface material was observed due to the delamination mechanism. Abrasion and the delamination wear mechanism predominate for a 200 m sliding distance, but the abrasion wear mechanism is also present in nearly all of the conditions tested mixed with other wear mechanisms. For higher sliding distances, the oxidative wear mechanism’s importance increases. In all of the samples, there was evidence of oxidation, as the EDX peak of oxygen appeared for every sample.
The SEM micrographs of the cross-section worn surfaces shown in
Figure 5a show the different wear mechanisms compared to the sliding distance and the analyzed material.
In the uncoated ZE41, as the sliding distance increases, the formation of a continuous oxidative wear product layer can be observed.
In the Al12Si/SiC coating, SiC particle detachment was not observed, but Al4C3 could be crack starting points. Cracks perpendicular to the sliding direction can be observed. During the wear process, subsurface cracks grow and are aroused as cracks perpendicular to the sliding direction, causing the detachment of sheet-like fragments of the worn material and producing delamination. The cracks increase with the sliding distance and form a crackle layer formed by small particles, which produces an abrasive wear mechanism and an oxidation mechanism.
In the Al40Si/SiC coating, for a 200 m sliding distance, the wear mechanisms are mainly abrasion and delamination, as in the case of Al12Si/SiC, in which there were cracks perpendicular to the surface that can produce a delamination wear mechanism. Mg2Si could act as starting crack point. For higher sliding distances, an oxidative wear product layer on the wear surface can be observed, which indicates the increases of the oxidative wear mechanism.
In the Al12Si20Ti/SiC coating, for a 200 m sliding distance, the wear mechanism is mainly abrasive. In contrast with the other coatings, cracks were not observed in the cross-section. As in the other cases, for higher sliding distances, the oxidative mechanism increases, and an oxidative product layer on the wear surface can be observed.
Figure 5b–f shows the EDX analysis performed on the uncoated ZE41, Al12Si/SiC, Al40Si/SiC and Al12Si-Ti/SiC coatings’ worn surfaces, respectively. There was evidence of oxidation, as the EDX peak of oxygen appeared for every sample. In addition, an Fe peak from the pin can be observed in the Al12Si/SiC EDX analysis. In the case of the Al12Si-Ti/SiC coatings, an oxygen peak can be detected in addition to a titanium peak.
Figure 6 shows the debris obtained in the wear tests. For a 200 m sliding distance, in the case of uncoated ZE41, the debris is formed by 100 µm short wires. In the case of the Al12Si/SiC and Al12Si20Ti/SiC coatings, the debris is formed by 100 µm platelets. This debris presents an oxidized surface. In addition, in the case of Al40Si/SiC, the debris is constituted of 130 µm particles. In all cases, the debris size decreases as the sliding distance increases. The oxidized debris particles break during the wear test and form debris agglomerates that increase the abrasive and oxidation wear mechanisms.
3.4. Corrosion Products and the Corrosion Mechanism
Figure 7 shows the corrosion properties for Al12Si/SiC, Al40Si/SiC, Al12Si-Ti/SiC coatings, and uncoated ZE41 in a 3.5% NaCl solution for different immersion times (1, 24, 96, and 168 h).
The open-circuit potential (OCP) variations for the uncoated substrates and Al/SiC coatings are represented as a function of the immersion time in the aggressive solution (
Figure 7a). Two main groups of materials can be discerned by their different behaviors. The first one consists of the uncoated ZE41 magnesium alloy, and the second group consists of the composite coatings. Uncoated ZE41 showed 5% lower
Ecorr values than any of the coated systems. In this sample,
Ecorr gradually increased until 50 h, and for the lower immersion times the
Ecorr remained constant.
In all of the coated systems, Ecorr increases for the first hour of immersion, followed by a decrease before the first 48 h.
For low immersion times, the Al12Si/SiC coatings’ Ecorr remained constant. However, the Al40Si/SiC and Al12Si20Ti/SiC coatings gradually decrease. In all cases, the Ecorr after 168 h was similar, and they were more active than the ZE41 substrate.
The corrosion current flows at OCP were measured under potentiostatically controlled conditions.
Figure 7b shows the variation of the corrosion current density with the time for the uncoated and Al/SiC coatings.
The uncoated substrate has a lower current density than the coated ones. In the first 24 h, an increase of the current density can be observed following a decrease of this factor, indicating the protective oxide layer’s formation.
At the start of the experiment, Al12Si/SiC had a higher current density than the uncoated ZE41, but lower than the other coated systems. The Al12Si/SiC current density gradually increased for the first 75 h, indicating the occurrence of a rapid anodic reaction; however, for lower immersion times, the current density decreases due to the protective oxide layer’s formation. Al40Si/SiC and Al12Si-Ti/SiC coatings have the highest current density; however, for the first 24 h, they follow a similar behavior: the current density decreases due to the protective layer’s formation. However, for higher immersion times, their behavior changes as a function of their composition. The Al12Si-Ti/SiC coating current density remains constant for all of the time, but at higher values than other systems. On the other hand, from 24 h to 96 h, the Al40Si/SiC current density is lower than the other coatings, but the current density increases quickly with time, indicating the occurrence of a rapid anodic reaction.
The corrosion behavior of the uncoated substrate and Al/SiC coatings was also studied in terms of the evolution of the polarization resistance (
Rp) with the immersion time (
Figure 7c). Uncoated ZE41 Rp gradually increases after 24 h immersion due to the superficial oxide protective layer’s formation. The coated systems follow similar behaviors, but the increase is 85% lower than the uncoated sample and, in some cases, the
Rp value decreases due to the breaking of the protective superficial layer. The noblest coating is Al12Si-Ti/SiC, followed by Al40Si/SiC.
Figure 7d shows the mass gain evolution (mg) with the immersion time. In all cases, from 75 h, the mass gain increases. The Al12Si/SiC coating shows the highest mass gain (81% higher than the other samples). The other samples show much lower mass gain, and it is similar in all of them.
3.5. Corrosion Products and the Corrosion Mechanism
The uncoated and coated samples were analyzed by SEM and EDS in order to identify the morphology and composition of the corrosion products when the materials were immersed in 3.5 wt.% NaCl at 25 °C for 168 h. The corrosion attack occurred in different ways in the different systems.
The surface morphology of the uncoated ZE41 immersed in the test solution for 168 h is shown in
Figure 8a.
Figure 8b shows a detail of the acicular morphology of the corrosion products formed.
Figure 8c shows a cross-section of the specimen. We can observe a detail of the layer cross-section in the
Figure 8c inset. The uncoated ZE41 surface presents generalized corrosion. The continuous oxidation product layer is 10 µm in the thinnest zone and 30 µm in the thickest zone. However, in some zones, the surface presented localized attacks. In these zones, there were signs of the nucleation and growth of a new layer of oxidized products. We can observe the accumulation of corrosion products on the coating surface. The formation of a protective superficial oxide layer produces the
Ecorr constant after the first 50 h immersed, as can be observed in
Figure 7a. The EDS (
Figure 8d) formed on the surface shown in
Figure 8a reveals that the corrosion products are formed mainly of O and Mg.
The Al12Si/SiC coating corrosion surface is shown in
Figure 9a. As is the case of the uncoated ZE41, an oxidation product layer was observed. In this case, the oxidation product layer is thicker (between 100 µm and 150 µm). In addition, the surface presented pitting corrosion and a cracked surface.
Figure 9b shows a detail of the acicular corrosion products formed that are similar to the corrosion products formed on the ZE41 sample.
Figure 9c shows a cross-section of the specimen, and a detail of this zone is shown inset. We can observe that the corrosion progresses along the interdendritic micro-constituent.
Figure 10a shows the Al40Si/SiC corrosion surface. As in the other cases, there is an acicular oxidation product layer. In some zones, this layer was cracked and pitting corrosion was observed (
Figure 10b).
Figure 10c shows an SiC particle partially covered by the oxidation product layer. The cross-section of the sample is shown in
Figure 10c. The oxidation layer is thicker than in the other cases (around 200 µm) (
Figure 10c inset). In addition, pitting corrosion is observed in the coating–substrate interface (
Figure 10c). The pitting corrosion occurred from the contact between the primary Si particles and the matrix. The EDS (
Figure 10e) shows that the corrosion products are formed mainly by Al, Mg and O.
The Al12Si-Ti/SiC surface (
Figure 11a) presents the same corrosion product layer but in this case, the layer is less continuous than in the others as shown in
Figure 11b, where the cross-section of the sample can be observed. The oxidation product layer thickness varies from 1 mm to uncoated zones. The corrosion products are acicular, with a shape that is similar to the other products observed before (inset
Figure 11a).
Figure 12c shows that the corrosion product layer progressed between the Ti-rich zones (TiSi
2, TiAl
3, and TiC), as evidenced by the cracks formed because of the pitting in the vicinity of Ti. Therefore, it seems that, in these zones, micro-galvanic couples were formed. The EDS (
Figure 10e) shows that the corrosion products are formed mainly by Al and O.
In all of the coated systems, the formation and subsequent breaking of the superficial oxide layer means that the
Ecorr increases for the first immersion hour, followed by a decrease for the first 48 h, as can be observed in
Figure 7a.
Figure 12 shows the different corrosion mechanisms of the coatings. The corrosive attack on the Al12Si/SiC surface (
Figure 12a) occurs principally through the MgO surface layer discontinuities due to the presence of SiC and intermetallics. Additionally, pitting corrosion occurs in the interface between Al
12Mg
17-Mg. The Al
12Mg
17-Mg dissolution takes place following the reactions (1) and (2):
Then, the Mg(OH)2 layer is degraded by the following reaction.
At the same time, Al
2O
3 is formed in the interface of Al
12Mg
17-Mg and Al
12Mg
17. The main reactions that take place are the following:
When these oxides surround an Al12Mg17-Mg eutectic or a Mg2Si dendrite, Mg(OH)2 is formed around the adjoining dendrites, causing a massive material detachment.
The Al40Si/SiC corrosion mechanism is shown in
Figure 12b. SiC particles act as pitting zones, because in these zones there are discontinuities on the corrosion product layer. In these coatings, Al
4C
3 needles were not observed; for this reason, there are fewer zones for the easy advance. The main reactions that take place are the same as those in the Al12Si/SiC coating.
Figure 12c shows the Al12Si20Ti/SiC corrosion mechanism. The oxide surface layer does not cover all of the coating’s surface (the TiSi
2, TiAl
3, and TiC are zones of discontinuities on the oxide product layer). The attack occurs through the surface discontinuities (TiSi
2, TiAl
3, and TiC). Cl
− ions, because of their small size, enter through the matrix interface and form galvanic microcells with the aluminum, and act as the start of pitting points.