3.1 Characterization of natural serpentine and PNS composite coatings
Figure 1(a) and (b) showed XRD spectrum and FTIR spectrum of natural serpentine. The main absorption peaks in XRD spectrum were exactly corresponding to the standard card of PDF#00-002-0094, which depicted the monoclinic magnesium silicate hydroxide Mg3Si2O5(OH)4, and specific crystal face indexes had been properly noted in Figure 1(a). However, there also appeared some weak absorption peaks that belonged to impurities. XRD analysis indicated that the main component contained in the natural serpentine was magnesium silicate hydroxide. FTIR spectrum demonstrated the characteristic functional groups of natural serpentine, and there only appeared absorption peaks of group O-H and group Si-O [40, 41]. The reason why absorption peak of group Mg-O did not appear in the FTIR spectrum was that its wavenumber was located about 450 cm−1, but Figure 1(b) only provided infrared signals that located in wavenumber of 550 cm−1 to 4000 cm−1.
The morphologies of the finely grind natural serpentine powders were exhibited in Fig. 2. It was obvious that various morphologies were contained in natural serpentine powders, such as fibrous natural serpentine in Fig. 2(a) and (b), schistose natural serpentine in Fig. 2(c), and granular natural serpentine in Fig. 2(d), respectively. With regard to fibrous natural serpentine, its diameter was about 50 nm, and aspect ratio can achieve two orders of magnitude. Higher tensile strength of fibrous natural serpentine was conducive to reduce the self-brittleness of phosphate coatings. Diameter of the schistose natural serpentine powders reached about 3 µm, at the same time, with small particles and sticks agglomerated onto their surface. Additionally, diameter of small granular natural serpentine was about 50 nm to 800 nm.
The surface morphologies of the fabricated PNS composite coatings were demonstrated in Fig. 3. Mass percent of natural serpentine in Fig. 3(a) and (b) was 0%; mass percent of natural serpentine in Fig. 3(c) and (d) was 10%. Although, the two kind coatings were all appeared different degrees of crack, obvious disparities could still be easily discovered. For coatings without incorporation of natural serpentine, there emerged serious micro cracks and its width can achieve 1 µm to 5 µm. Some cracks were even throughout the total figure and extended to further places of phosphate coating. Along with the water evaporation, volume of the phosphate coatings was rapidly decreased during the curation process, and therefore, micro cracks were inevitably generated on the surface. When it came to the coatings incorporated with 10 wt.% natural serpentine, the situation was significantly improved. Existence of fibrous natural serpentine conspicuously narrowed the width of cracks, and also restrain its further elongation. As a result, the defects of PNS-10 composite coatings were obviously decreased. However, for the reason that the proportion of fibrous natural serpentine was relatively small, some micro cracks still appeared on the surface of PNS-10 composite coatings. More significantly, reactive hydroxyl of natural serpentine would certainly occur interfacial combination with the reactive groups of phosphate cement during high-temperature curation process, by which mechanical strength of the composite coatings can be elevated, and anti-wear performance of the composite coatings can be subsequently promoted.
3.2 Tribological performance of PNS composite coatings
The tribological performance of phosphate composite coatings incorporated with various mass percent of natural se rpentine was demonstrated in Fig. 4, of which the mass percent of natural serpentine increased from 0–25% with an increasing rate of 5%. Fig. 4(a) exhibited the specific friction coefficient and wear rate of each coating, and Fig. 4(b) represented concrete frictional curve of each coating. All experiments were conducted under same friction settings. Specifically, the friction coefficients of PNS composite coatings slightly ascended with the increase of natural serpentine, and when the mass percent of natural serpentine was 20%, the friction coefficient of the composite coatings was in maximum. The same conclusion could also be reflected in Fig. 4(b). However, when it referred to wear rates of PNS composite coatings, its variation trend emerged distinct discrepancy to that of friction coefficients. With the mass percent of natural serpentine increased from 0–25%, wear rate of each coating was firstly decreased and then rapidly increased when content of natural serpentine exceeded 10 wt.%, which illustrated that with incorporation of 10 wt.% natural serpentine, phosphate coatings owned the most excellent anti-wear property, and anti-wear performance was promoted about 55.8%. When mass percent of natural serpentine was less than 10%, the amount of natural serpentine was relatively small, and its strengthening effect was insufficient to phosphate coatings. In contrast, when the mass percent of natural serpentine was more than 10%, the bonding strength between phosphate cement was largely wakened by excessive natural serpentine. Therefore, mechanical strength of the corresponding phosphate coatings was decreased, and as a result, the phosphate coatings with excessive incorporation of natural serpentine behaved deficient anti-wear performance under the same experiment condition. Only when the mass percent of natural serpentine was 10%, anti-wear performance of phosphate coating can be mostly promoted.
The 3D morphologies and contour profiles of wear tracks were exhibited in Fig. 5, of which the mass percent of natural serpentine in Fig. 5(a) and (b) was 0%; in Fig. 5(c) and (d) was 10%; in Fig. 5(e) and (f) was 25%. It was obvious that when phosphate coating was incorporated with 10 wt.% natural serpentine, its wear track in Fig. 5(c) was relatively flat, and abrasion extent was apparently mitigated when compared with that of Fig. 5(a) and (e). This conclusion could also be confirmed by the specific contour line of each wear track. For PNS-10 composite coating, both the width and depth of wear track were simultaneously achieved the minimum value. The experimental results secondly proved that with incorporation of 10 wt.% natural serpentine, anti-wear performance of the phosphate coatings was significantly promoted.
Although anti-wear performance of phosphate coatings was effectively promoted by incorporation of 10 wt.% natural serpentine powders, friction coefficient of PNS-10 composite coatings was still relatively high. Especially, accompanied by the increased mass percent of natural serpentine, friction coefficient of the composite coatings also slightly increased. The possible reason was may attributed to that relatively small friction force under the friction condition of low applied load and sliding speed was insufficient to completely overcome the interlayer force in natural serpentine, which made inferior friction-reducing property of natural serpentine on phosphate coatings. Therefore, performance of PNS-10 composite coatings under different applied load and sliding speed was investigated, and specific result was shown in Fig. 6. Fig. 6(a) was the tribological performance under different applied load from 2 N to 14 N, in which sliding speed was 8 cm/s, and sliding distance was 150 m. Fig. 6(b) was the tribological performance under different sliding speed from 4 cm/s to 16 cm/s, in which applied load was 11 N, and sliding distance was 150 m. When applied load increased to 8 N, friction coefficient of composite coatings decreased about 21.3%, and when the sliding speed exceeded 8 cm/s, friction coefficient of composite coatings decreased about 15.6%. At the same time, more obvious impact was imposed on the wear rate of composite coatings. Accompanied by the increase of applied load, wear rate was conspicuously declined, which indicated that the anti-wear ability of composite coatings was more significantly promoted when the applied load was increased. In Fig. 6(b), accompanied by the increased sliding speed, wear rate of composite coatings was firstly decreased and then increased. Superhigh sliding speed of 16 cm/s may exceed the speed resistant capacity of the composite coatings and serious abrasion was occurred on the composite coatings, Therefore, increased wear rate was finally exhibited.
Figure 7 exhibited the optical morphologies of wear spots under various friction condition, and the specific experimental settings had been noted on each picture. An obvious variation law could be perceived was that accompanied by the increasing of applied load and sliding speed, rubbing degree appeared on the wear spot was apparently alleviated. And simultaneously, transfer and adhesion of the composite coating on surface of the counterpart steel ball was obviously intensified. What need to be pointed out was that the minimum wear spot area was obtained in Figure 7(f), in which the applied load was 11 N and the sliding speed was 8 cm/s. Such experimental results demonstrated that excellent anti-wear performance of PNS-10 composite coatings under the friction condition of increased applied load and sliding speed in Fig. 6 may related to the adhesion and transfer of composite coatings onto the counterface.
3.3 Strengthening mechanism of natural serpentine on tribological performance of phosphate coatings
For the purpose that explore strengthening mechanism of natural serpentine on tribological performance of phosphate coatings, detailed analyses on the wear spot were conducted. The selected counterpart steel ball was under the friction condition where applied load was 11 N, and sliding speed was 8 cm/s. Elemental analyses on wear spot were demonstrated in Fig. 8, of which Fig. 8(a) was the morphology of the wear spot, Fig. 8(g) was the EDS result of the wear spot, and others was the mapping result of each element on the wear spot. Aside from the counterface element of Fe and C, appearance of element P, O, Mg, Si, and Cr was attributed to the transfer of PNS-10 composite coatings onto the counterface during the friction process. Among them, the mass percent of element O was in maximum, which was because both the phosphate binder and natural serpentine were rich with O. Element P and Cr were from transfer of phosphate cement. Element Mg may come from the transfer of phosphate cement or natural serpentine, and element Si was from the transfer of natural serpentine.
In order to explore the specific role of natural serpentine during the friction process, XPS analysis was conducted on the same wear spot. The concrete results were shown in Fig. 9, of which Fig. 9(a), (b), (c), and (d) was the XPS fine spectrums of Fe2p, P2p, Si2p, and Mg1s, respectively. Chemical bond of Fe-O-Si in Fig. 9(a), indicated that natural serpentine happened isomorphic replacement reaction with the counterface:
① 4Fe + 3O2 → 2Fe2O3
② Fe2O3 + Mg3Si2O5(OH)4 → Mg2Fe2SiO5(OH)4 + MgO + SiO2
③ Fe2O3 + Mg3Si2O5(OH)4 → Fe2Si2O5(OH)4 + 3MgO
and at the same time, generated the microcrystalline ceramic phase of MgO and SiO2 on the counterface [28, 32, 35, 36]. Because of the special micro-structure and high chemical activity, natural serpentine possessed high cation exchange capacity. Especially, bigger Mg2+ in octahedron layer can be easily replaced by Fe3+, Cr3+, Al3+ and etc. What’s more, lattice bending and structure distortion of natural serpentine were also inevitably occurred at the friction interface, through which unsaturated chemical bonds such as O-Si-O, Mg-O, and Si-O-Si were released. Under the stimulation of friction force and local friction heat, natural serpentine was prone to generate interfacial interactions with the counterface. Chemical bond of Fe-O-P indicated that phosphate cement also occurred chemical reaction with the counterface. Physical adsorption and tribo-chemical reaction between counterface and natural serpentine as well as phosphate cement greatly intensified adhesion and transfer of composite coating onto the counterface, through which tribo-film can be in-situ formed on the counterface during the friction process [42–48]. What should be noted was that the chemical bond of P-O-M (Metal) in Fig. 9(b) may is P-O-Mg, P-O-Cr, or P-O-Fe. Chemical bonds of Si-O, Si-O-Si, and Mg-O in Fig. 9(c) and (d) may attributed to the transferred natural serpentine, but also may attributed to the newly generated microcrystalline ceramic phase of MgO and SiO2.
The morphology of tribo-film formed on the counterpart steel ball was shown in Fig. 10. It was clear that a continuous protective tribo-film had generated on the counterface after the friction experiment. At the same time, because of tribo-chemical reaction between counterface and natural serpentine as well as phosphate cement, tribo-film was well combined with the counterface. The thickness of the tribo-film was from 73 nm to 146 nm. The total frictional surface of the counterpart steel ball was completely protected by the generated tribo-film and no exposed area could be observed reflected in Fig. 10(a). However, there were numerous write and black spots contained in the tribo-film exhibited in Fig. 10(b) and (c). For the sake of identifying the concrete chemical composition of these unknowing spots, higher resolution of STEM analysis was conducted. What should be emphasized was that the diffraction contrast between the TEM and STEM was exactly opposite. That was to say, substance with black appearance in TEM images was turned white in STEM images, and vice versa.
Figure 11 was the STEM images of micro particles captured in tribo-film and its composition analyses. Figure 11(a) and (b) exhibited the specific morphology of the particles. Particles in Figure 11(a) was more concentrated and its diameter was only about 3 nm, and diameter of the particle in Figure 11(b) was about 20 nm. Elemental mapping result of Figure 11(b) was reflected in Figure 11(c), which confirmed that the elemental composition of the particle was Ni. Simultaneously, lattice fringe spacing of this particle was also corresponding to the spacing of (111) in cubic Ni. Obviously, because the friction condition was relatively harsh, partial exposure of the substrate was happened during the friction process. Therefore, wear debris of the substrate was embedded in tribo-film. However, for extremely small diameter of particles in Figure 11(a), resolution of the elemental mapping was too poor to identify their elemental composition. But lattice fringe spacing of those particles was exactly corresponding to the spacing of (400) in magnesium silicate hydroxide, which conformed that the particles in Figure 11(a) was super-finely grind natural serpentine powders.
Figure 12 was composition analyses of the black spot reflected in STEM image of tribo-film, which corresponded to the white spot in Fig. 10. Combined with its specific morphology in Figure 12(a) and elemental mapping results of Mg, O, Ni, Cr, and Fe, it could be found that signal strength of each element at the bleak area were apparently reduced. Obviously, elemental composition of the bleak area was not from the phosphate cement, substrate, or counterpart steel ball. Therefore, the only possible result of the bleak area was a hole, which was normal to the formation of tribo-film during the friction process. The reason why mapping result of element Si was not exhibited in Figure 12 was that signals of element Si in STEM analysis was uncharacteristic.
Based on the above analyses of wear spot and tribo-film. The strengthening mechanism of natural serpentine on tribological performance of phosphate coatings can be demonstrated as follows. When applied load and sliding speed were increased, under the stimulation of friction force and local friction heat, natural serpentine was activated and occurred isomorphic replacement reaction with the newly friction-polished counterface, which triggered the formation of tribo-film on the counterface [32, 37, 39]. As the friction process continued, the abrasion and generation of tribo-film finally achieved dynamic equilibrium, and as a result, a certain thickness of tribo-film was in-situ formed on the counterface. Although, numerous micro-holes were inevitably produced during the accumulation process of tribo-film, the captured micro wear debris and super-finely grind natural serpentine powders can effectively fill in these holes, which definitely elevated the compactness of tribo-film. What’s more, the newly generated microcrystalline ceramic phase of MgO and SiO2 not only elevated mechanical strength of the tribo-film, but also definitely promoted anti-wear ability of the tribo-film. Just because existence of such tribo-film at the friction interface, anti-wear performance of the phosphate composite coatings was significantly promoted. At the same time, serious furrows generated on the wear spot during the friction process were also effectively self-repaired by the tribo-film, through which further abrasion on the counterface was greatly restrained.
As for friction-reducing property of phosphate composite coatings, when applied load and sliding speed was increased, layer slip in natural serpentine was boosted, and friction coefficient of the composite coatings was firstly reduced. But in the later stage of friction experiment, with consumption of natural serpentine as well as accumulation of the hard ceramic phase (MgO and SiO2) at the friction interface, friction coefficient of the composite coatings was subsequently increased. The conclusion could be proved by the specific friction curves of PNS-10 composite coatings exhibited in Fig. 13, of which Fig. 13(a) was friction curves under increased applied load, and Fig. 13(b) was friction curves under increased sliding speed. As a result, friction-reducing ability of natural serpentine was finite to phosphate composite coatings throughout the total friction process, reflected in Fig. 6.