3.1. Microstructure and Phases Analysis
Figure 2 shows the substrate-coating transition zone of hardfacing coatings. One of the most important features that distinguish hard coating from other coating methods is that their coating thickness is quite high and they create strong metallurgical bonds with the substrate. These features enable hardfacing coatings to be used for a long time in aggressive environments. On the other hand, porosities that will occur in the transition zone and coating layer after coating will significantly reduce the performance of the coating. SEM images show that the alloys selected as the coating material and the process are compatible and a metallurgical bond is formed between the substrate material and the coating layer. Additionally, no significant porosity formation was observed in either the transition zone or the coating layer. In the SEM images, the coating process carried out at room temperature, columnar grains were formed at the interface, extending from the substrate material towards the coating area. During repeated melting and solidification, the temperature increases and columnar grains grow from the colder substrate material towards the coating region. The interface, which develops depending on the dynamic solidification conditions of the welding process, has been reported in similar studies [
17,
23].
XRD analysis of hardfacing coatings was given in
Figure 3. According to XRD analysis that changing alloying element in the coating composition was affected the phases in the microstructure. In the XRD analysis, α-Fe, FeMo
2B
2, M
2B type boride, R-Fe
63Mo
37 and trace amounts of Fe
3B and µ-Fe
7Mo
6 phase were detected. Normally, R-Fe
63Mo
37 and Fe
3B phase is a non-equilibrium phase formed under rapid cooling conditions [
26]. However, it can be seen in trace amounts in a solidified microstructure under dynamic solidification conditions such as the welding process.
SEM images of hardfacing coatings are given in
Figure 4. Fe
16B
4 based hardfacing coating composition contains 20 atomic% boron. During solidification, the Fe2B phase is expected to solidify first as seen Fe-B phase diagram given in
Figure 5. As the eutectic temperature drops below the eutectic temperature, which is approximately ~1177
oC, the liquid phase will solidify eutectically. For this reason, primary Fe
2B phases and eutectic α-Fe+Fe
2B structure can be found in the coating microstructure. However, since the electrode composition passes very close to the eutectic point, the amount of primary phases is expected to be low. In EDS analysis, α-Fe (point 1) represents Fe
2B with a partially block structure (point 2) and α-Fe+Fe
2B eutectic structure (point 3). Normally, boron element cannot be detected quantitatively by EDS analysis. However, with EDS analysis, signals can be received from points containing boron. As expected, there were trace amount block-structured Fe
2B phases that solidify primarily and the microstructure largely solidifies as eutectic structure.
It is seen that significant changes occur in the microstructure of the Fe
15MoB
4 based coating with the addition of molybdenum in the electrode cover composition. According to EDS and MAB analysis, a molybdenum signal is received from these block-structured phases. It is also understood that these block-structured phases contain boron. It is known that for a fixed boron ratio in the Fe-Mo-B system, increasing the amount of molybdenum shifts the solidification to the hyper-eutectic side [
27]. On the other hand, the high melting temperature and thermodynamically stable phase in the Fe
15MoB
4 based coating is the FeMo
2B
2 compound [
28,
29]. For this reason, solidification begins with the formation of FeMo
2B
2 solid crystals and this phase grows with a block structure. After the formation of the FeMo
2B
2 phase, theoretically molybdenum is consumed in the melt, but there is still atomic 15% boron. Solidification continues with the formation of α-Fe, similar to a sub-eutectic composition according to the Fe-B phase diagram. For this reason, there are α-Fe islands around the FeMo
2B
2 phase. When the temperature drops below the eutectic line remaining liquid solidified as α-Fe+Fe
2B eutectic structure and continue to grow. It indicates that the phase detected as M
2B in the XRD analysis is taken from the α-Fe+M
2B structure, which solidifies eutectically around the block structured phases. Although it was assumed that this system solidifies according to the Fe-B phase diagram, the eutectic formed is morphologically different from the eutectic structure formed in the Fe
16B
4-based coating. It is known that molybdenum can be dissolved in the Fe
2B phase [
30]. Possibly, molybdenum dissolved in the Fe
2B phase causes this phase to solidify as (Fe,Mo)
2B and the morphology of the eutectic structure changes. Similar microstructure has been reported as a reticular eutectic structure in the literature [
27]. Also, molybdenum signal was received from EDS region number 2. Although the solubility of molybdenum in α-Fe is quite low under normal conditions, its solubility is ~24% at ~1450
oC. It has been reported that [
31,
32]. In a process where solidification is dynamic, such as welding, rapid regional or structural cooling might be occurred [
24]. For this reason, it was normal for the molybdenum solubility to increase regionally in α-Fe and the dark gray regions are called α-(Fe,Mo). As a result, block-structured FeMo
2B
2, α-(Fe,Mo) and α-Fe+(Fe,Mo)
2B eutectics were expected to form in the final microstructure.
There is a significant increase in the ratio of block-structured FeMo
2B
2 phases, in the SEM image of the Fe
15MoB
4 based coating given in
Figure 4. The thermodynamic analysis performed with PANDAT software (
Figure 5) that similar phases occur in the Fe
14MoB
4 based coating. In XRD analysis, it was determined that the Fe
14Mo
2B
4 based hardfacing coating contained α-Fe, FeMo
2B
2 phase, M
2B type boride and trace amounts of Fe
3B borides in its microstructure. Solidification is expected to begin with the formation of the FeMo
2B
2 phase, similar to the Fe
15MoB
4-based coating. However, in the microstructure of the Fe
14Mo
2B
4 based coating, it is seen that the block structured phases and the eutectic structure change morphologically. It also has been reported in the literature that the FeMo
2B
2 phase formed in the Fe-Mo-B system can solidify in different forms and the morphology of this phase can change depending on the Mo/B ratio [
29]. In similar studies, it has been reported that a eutectic structure consisting of FeMo
2B
2 and Fe,Mo solid solution can be found around the FeMo
2B
2 phase [
30]. However, this eutectic structure solidifies into a morphologically floriform structure. Another phase detected in the XRD analysis, the M
2B type phase, indicates the eutectic solidified Fe
2B phase. As stated in previous studies, Fe
2B can dissolve some molybdenum during solidification and form (Fe,Mo)
2B phase [
30]. The Fe
3B phase detected in trace amounts in the XRD analysis was not visible in the microstructure in a distinguishable. Yin et al. [
27] stated that the Fe
3B phase was distributed within the Fe-M
2B eutectic structure. In the final microstructure, consist of FeMo
2B
2, α-(Fe/Mo)+FeMo
2B
2 eutectic (floriform structure that solidifies around FeMo
2B
2), α-(Fe/Mo) and α-(Fe/Mo)+(Fe,Mo)
2B (reticular eutectic structure).
A structure consisting of light-colored phases, eutectic-like structures with the same contrast as these phases, and a matrix phase can be seen in the SEM image of Fe
14Mo
4B
2 based coating. According to the thermodynamic analysis performed with Pandat software, α-Fe, FeMo
2B
2 and phase were expected to be present in the microstructure under stable solidification conditions. However, in XRD analysis, in addition to α-Fe and FeMo
2B
2 phases, µ-Fe
7Mo
6 and R-Fe
63Mo
37 phases were detected. According to the Fe-Mo phase diagram, the R-Fe
63Mo
37 phase is an unstable phase below 1200
oC and indicates solidification at a high cooling rate [
17]. Molybdenum signal was detected from white island-shaped phases in the EDS analysis. The first phase expected to form under balanced conditions was the FeMo
2B
2 phase. However, the surface of the island-shaped phase shown at EDS point number 1 has a faceted structure, unlike the FeMo
2B
2 phase. Possibly, this phase is the R-Fe
63Mo
37 phase, which solidifies out of equilibrium, as stated in the literature [
17]. The µ-Fe
7Mo
6 phase, which was detected in trace amounts in the XRD analysis, could not be clearly detected in the microstructure. Previous studies also show that this phase is not clearly distinguishable [
17,
33]. Eutectic regions indicate the FeMo
2B
2 phase detected in XRD analysis. The amount of molybdenum affects not only the morphological properties of FeMo
2B
2 phases but also the morphology of the eutectic structure (α-(Fe,Mo)+FeMo
2B
2). For this reason, increasing the amount of molybdenum and decreasing the boron rate affects the distance between the lamellar in the eutectic structure in the Fe
14Mo
4B
2 based coating.
3.2. Hardness and Wear Tests
Figure 6 shows the macro hardness test results of hardfacing coatings. In the coatings carried out on AISI 1020 steel substrate, the lowest and highest hardness change in the substrate material hardness was measured as 58 % in the Fe
16B
4-based coating and 73 % in the Fe
14Mo
2B
4-based coating, respectively. Additionally, it was observed that an increase in macrohardness as the amount of molybdenum added to the Fe-B based hardfacing electrode composition increases. With 5 % molybdenum added to the electrode cover composition, an increase in hardness of 22.8 % was observed, and with the addition of 10 % molybdenum, an increase in HRC hardness of 30.48% was observed compared to the Fe
16B
4 based coating. On the other hand, the macro hardness value of Fe
14Mo
2B
4 based hardfacing coating was measured as 56.5 HRC, while the hardness value of Fe
14Mo
4B
2 based hardfacing coating was measured as 49.9 HRC. In other words, the hardness value of Fe
14Mo
2B
4 based hardfacing was approximately ~13 % higher than that of Fe
14Mo
4B
2 based hardfacing. This shows that although the hardness increases with molybdenum for Fe-Mo-B based coatings, the amount of boron in the coating composition has a significant effect on the increase in hardness. In addition, the top values of hardness were related to the phases in the microstructure and especially the FeMo
2B
2 phase plays an important role in increasing the hardness.
In the study, the lowest microhardness value was measured from the substrate material as seen
Table 3. Among the coated samples, the lowest matrix phase hardness was measured in Fe
16B
4 based hardfacing coating. The matrix phase hardness of the Fe
16B
4 based coating gives almost the same results as the substrate material. The solubility of boron, an interstitial element, in the α-Fe matrix was quite low. For this reason, the boron in the composition was found in Fe
2B or the eutectic phase formed by this phase. With the addition of atomic 5 % molybdenum into the alloy, there was a significant increase in the hardness of the matrix phase. Additionally, it was observed that this increase was not only in the matrix phase but also in the eutectic structure. Due to the nature of the welding process, molybdenum may dissolve to some extent in the matrix phase and the Fe
2B phase. This situation causes an increase in hardness in the matrix phase as a result of solid solution hardening [
34]. Similarly, molybdenum dissolved in the Fe
2B phase causes the hardness of this phase to increase. Although the hardness of the other phases was similar, trace changes were observed depending on the composition, especially in the matrix phase. On the other hand, the matrix hardness of the Fe
14Mo
4B
2 based coating was less than that of other coatings containing molybdenum. The hardness of the R-Fe
63Mo
37 phase in the microstructure of the same coating was also measured to be lower than the boron-containing phases. As can be seen from previous studies, changes in both microstructural and atomic dimensions change the hardness of the coating [
35,
36]. The difference in the matrix phase between the compositions might cause distortion in the lattice, causing the hardness to be slightly different. Microstructurally, each phase prevents dislocation movement, but the physical properties of the phase determine the limit values of plastic deformation. As stated in the literature, boron-containing phases play a decisive role on the hardness of the coating [
37]. The hardness values measured from the phases in the microstructure have an effect on the macro hardness. Although the hardness values measured in the study depend on the composition and production method for Fe-Mo-B based alloys, they appear to be compatible with various studies [
38,
39,
40].
Figure 7 shows the wear rate graph of hardfacing coatings depending on load and distance. According to the graph, it is understood that the wear rate increases with both increasing distance and increasing load. In the study, the highest wear rate was measured from the substrate material. Among the coated samples, the highest wear rate was measured in Fe
16B
4 and the lowest wear rate in Fe
14Mo
2B4 based coating. Wear rates exhibit a similar behavior to the hardness results. As Archard stated [
41], the wear rate varies inversely with the hardness. The basic philosophy of hardfacing coatings is to cover the surface with a harder layer than the substrate material. As mentioned above, each hard phase in the microstructure of this hard layer formed on the substrate surface changes the effect of the coating on wear. In addition, not only the hardness of these phases, but also their compatibility with the surface, morphology, etc. properties also affect the wear behavior of the coating. In this study, the hard eutectic α-Fe+Fe
2B structure, which forms the microstructure of the Fe
16B
4-based coating, increases the wear resistance of the substrate material. The high hardness FeMo
2B
2 phase formed in the microstructure with the addition of molybdenum and the different morphologies of this phase cause further increase in wear resistance. On the other hand, although the wear resistance increases with increasing molybdenum amount, the wear resistance of the Fe
14Mo
2B
4 based coating was measured to be higher at all loads and distances than the wear resistance of the Fe
14Mo
4B
2 based coating. Boron is a critical element in the formation of hard phases, and the volumetric fraction of the phases formed by boron affects the wear resistance of the coating. In addition, although boron provides the formation of the hard phase, the matrix phase is critical in terms of both hardness and wear resistance. The fact that molybdenum is soluble in the matrix phase and increases its hardness has an effect on wear. Likewise, both the hardness and wear resistance of the Fe
16B
4-based coating, where the matrix phase is soft, were measured to be lower than the Fe
14Mo
4B
2-based coating.
Figure 8 shows the friction coefficient graphs obtained after the wear test at different loads and distances applied to hardfacing coatings. Friction coefficient values vary between ~0.4-0.8 µ. According to the friction coefficient data of the coated samples, it can be said that the highest friction coefficient value was obtained in Fe
16B
4-based hardfacing and the lowest friction coefficient values were obtained in Fe
14Mo
4B
2-based hardfacing. The overall friction coefficient is closely related to the surface profile and phases in the microstructure [
42,
43,
44]. In composite materials such as hardfacing, each phase has a different effect on wear. On the other hand, the hardness, morphology, volumetric ratio and distribution of these phases are other parameters that affect the friction coefficient. The friction coefficients given in
Figure 8 act independently of the surface hardness. However, it can be said that the smoothest surface among the samples is on the Fe
14Mo
4B
2 based hardfacing coating after the substrate material. On the other hand, the composite structure seen in the microstructures of other coatings is likely to create resistance during sliding. These composite structures might cause an increase in the coefficient of friction. Similar studies in the literature have stated that the coefficient of friction increases with the increase in the amount of hard phase in the metal matrix composite [
45]. In the study, no correlation was found in the change of friction coefficient according to load and distance.
In the EDS analysis of the substrate material shown in
Figure 9, an intense amount of oxygen signal was received from the dark areas. The formation of a FeO-based oxide layer during wear on the carbon steel surface has been reported in similar studies [
46]. On the other hand, the microstructure of the substrate material was not containing factors such as grain boundaries, secondary phases and eutectic structure that would create resistance during wear compared to coated samples. For this reason, friction occurs on a smoother surface. In the friction coefficient graph given in
Figure 8, it reached the steady state wear phase after a short running-in period. The homogeneous oxide layer formed on the surface might acted as a solid lubricant, causing the friction coefficient to be relatively lower than other samples. On the other hand, it is seen that this oxide layer forms debris at some points. Local adhesions may occur as a result of micro-weld in softer and easily oxidized substrate material. Increasing heat during friction and the change in surface topography due to the oxide layer formed may also cause fluctuations in the friction coefficient [
47]. On the other hand, oxide-based debris formed as a result of the delamination of the oxide layer was seen in the SEM image of the substrate material. It was understood that wear was dominated by oxidative and adhesive wear mechanism. The worn surface appearance of the hardfacing coated samples shown in
Figure 9 shows that the oxidized parts remained more localized compared to the substrate material.
As a result of the SEM analysis performed on the coated samples, different mechanisms played an active role during wear. It can be seen that debris and delaminations were formed on the surface of the Fe
16B
4-based hardfacing coating shown in
Figure 10-a after wear. Debris might be occurred due to the sawing effect created by the eutectic phase during friction. It was also observed that delamination occurs in the soft (matrix) phase during friction in the eutectic structure. This indicates that it might be pulled out due to adhesion formed in the soft matrix phase with increasing heat in the later stages of friction. With the addition of molybdenum, the delamination of hard phases formed in the microstructure decreased by reducing the oxide detachment on the surface. However, the abrasive wear increases with increasing hard phase fraction and micro abrasive lines appear on the surface. One of the basic mechanisms expected to develop during the wear of hardfacing coatings is microabrasive wear, which was expected to occur due to three-body wear mechanism. As a result of the micro-abrasive wear mechanism, small pieces pulled out from the friction surfaces were plastered on the surface, causing wear called smeared [
48,
49,
50]. The SEM image and MAP analysis shown in
Figure 9 indicate this situation, where cracks occur directly in oxidized areas. Similarly, the cracks formed in
Figure 10-b and c indicate removal from the surface after smearing. On the other hand, block-structured phases formed by the addition of molybdenum caused a large hardness difference between the matrix and the particle. The difference in hardness between the matrix and the hard phase might be caused fatigue in the matrix phase. Fatigue occurring during wear might cause fatigue cracks and subsequent ruptures after a while. In addition, in the SEM image of the Fe
15MoB
2 based coating given in Figure. As a result, hard phases such as M
2B and FeMo
2B
2 play an active role during friction. These hard phases and eutectic structures were in direct contact with the abrasive. For this reason, the increasing amount of block-structured or eutectic hard phase acts in the direction of increasing wear resistance. Similar results have been reported in previous studies [
19]. In the study, it was determined that micro abrasive, adhesive and oxidative wear were dominant wear mechanism.
3.3. Corrosion Behavior
OCP curves were given in
Figure 11-a. The OCP curves become stable in a short time. As stated in the literature, when the open circuit potential reaches equilibrium, it indicates that the film formation and dissolution on the surface are in balance [
51]. In all OCP curves, there was first a slight movement towards the anodic side, and then the curves became stable. The OCP value of the substrate material was measured to be slightly more positive than the coated samples. This might be due to the oxide layer that forms more easily on the surface of the substrate material. According to the OCP test, anodic corrosion tendency decreased with increasing molybdenum amount in the coating composition. The OCP value of the Fe
16B
4-based coating was measured to be similar to molybdenum-containing coatings.
Tafel curves obtained from samples subjected to electrochemical corrosion test in 0.5 M NaCl solution were given in
Figure 11-b. The corrosion potential obtained from these curves gives information about the direction of corrosion [
19,
52]. According to the curves, there was no big difference in corrosion potentials. However, the most noble sample in terms of corrosion potential is the AISI 1020 steel substrate. In other words, it was the sample substrate material with the least corrosion tendency according to its corrosion potential. According to the corrosion parameters given in
Table 4, it is seen that the addition of molybdenum to the molybdenum Fe-B based coating composition does not cause a linear change on the corrosion potential. Among the coated samples, the noblest corrosion potential was measured in the Fe
14Mo
4B
2 based coating. On the other hand, current density, which is a critical parameter in the kinetic interpretation of corrosion, is directly related to the rate of corrosion. According to corrosion current density values, the lowest current density was measured in Fe
16B
4 based coating. Many researchers have reported that Fe-B alloys near the eutectic point have very good corrosion resistance [
53,
54,
55,
56]. There was a decrease in the current density values with the addition of molybdenum into the Fe
16B
4 based coating. In other words, the corrosion resistance of the coating has increased.
Another method used to understand corrosion in kinetic terms is the EIS method. The similarity of the EIS curves seen in
Figure 11-c indicates that similar corrosion mechanisms have developed in the coating layers. On the other hand, the diameter of the capacitive loop varies in direct proportion to the corrosion resistance. In other words, the corrosion resistance of the coating with a large capacitive loop diameter is high [
57,
58]. EIS test parameters obtained by equivalent circuit modeling are given in
Table 5. According to the model, Rs is a measure of solution resistance and Rct is a measure of polarization resistance. A higher Rct value indicates higher corrosion resistance. The fact that both the capacitive loop diameter and the Rct value are large indicates that the corrosion resistance of the Fe
16B
4 based coating is higher. This can be associated with the microstructure of the Fe
16B
4-based coating. Factors such as phases and grain boundaries within the microstructure increase the polarization resistance [
59]. Rct resistance increases especially in Fe
16B
4 based coatings that contain a high fraction of eutectic structure. According to EIS data, AISI 1020 steel substrate, which does not contain grain boundaries in the microstructure, has low corrosion resistance. Although the microstructure of the Fe
14Mo
2B
4 based coating is complex, the Rct value was low. This may be related to the morphology and composition of the eutectic structure.
In the post-corrosion SEM image of the substrate material shown in
Figure 12, it is seen that a like cotton layer is covered on the surface. In the EDS analysis carried out in this region, intense signals belonging to Fe and Cl elements were received. This indicates that the iron on the surface is dissolved as a result of the anodic reaction and ferrous salts are formed on the surface.
The reaction of anodic dissolution occurring on the substrate surface is the oxidation reaction of α-Fe and is expected to occur according to the reaction given below. In the SEM image, it is understood that this reaction develops homogeneously on the surface. This type of homogeneous corrosion is a less dangerous corrosion mechanism and generally the homogeneous oxide layer formed on the surface polarizes the surface. However, the breakdown of this layer accelerates corrosion.
Cathodic Reaction:
direct oxygen reduction of cathodic reaction
Figure 13 shows SEM images of hardfacing coatings after corrosion. In the images, there was no corrosion damage in the parts formed by FeMo
2B
2, R-Fe
63Mo
37 and these phases within this eutectic, but there was a significant amount of dissolution in the α-Fe/(Fe,Mo) regions between the matrix phase and the eutectic phase. This situation indicates the formation of galvanic corrosion caused by the potential difference between the boride phase and α-Fe. Also, this results in selective corrosion of the matrix phase.
In Fe
16B
4 based coating, the matrix phase consists of α-Fe and dissolution can be occurred according to the reactions given above. The molybdenum added to the Fe
16B
4-based coating might have increased the corrosion resistance of the matrix by dissolving it in the matrix phase. Similarly, there was a significant dissolution in the phases around the FeMo
2B
2 phase. As a result of the potential difference between the FeMo
2B
2 phase and the surrounding phases, it may have an accelerating effect on dissolution. Similarly, it has been reported in the literature that the low-molybdenum-containing phase in the region adjacent to the high-molybdenum-containing phase in the eutectic structure will corroded preferentially and cause intergranular corrosion [
62]. It was also clearly seen in SEM images. In the SEM image and MAP analysis given in
Figure 14, the iron signal received from the eutectic region shows that there was no dissolution in these regions. That is, both M
2B and FeMo
2B
2 phases behave as noble with respect to the matrix phase.