3.1. Tribological Behaviors
Friction coefficients of PEEK/GO and PEEK/MoS
2 composites are shown in
Figure 2. It can be clearly seen that the friction coefficient of PEEK and PEEK/GOs is gradually increasing and then fluctuating in a stable range (
Figure 2a). The variation of average friction coefficient gradually decreases, and then increases with increasing of the GO content (
Figure 2c). However, the average friction coefficient of all PEEK/GOs is lower than PEEK. The average friction coefficient of PEEK is 0.2319, suggesting PEEK has poor friction reduction capacity, which coincides with the results of Li et al. [
33]. The PG05 has an average friction coefficient (
μ) of 0.1966, reduced by 15% compared with PEEK. The PG07 has a minimum average friction coefficient of 0.1901, decreased by 18% compared with PEEK, indicating the best friction reduction abilities. When GO content exceeds 0.7 wt.%, the average friction coefficient begins to increase. For instance, PG09 has an average friction coefficient of 0.2171, which decreases by only 6% compared with PEEK; but the response time to a stable friction state is obviously different (
Figure 2a). The response time of PEEK is about 950 s. However, the response time of PEEK/GOs reduces with increasing of the GO content. For example, the response time of PG03 is about 500 s, which reduces by 47% compared with PEEK, and the values of PG05 and PG07 are both about 400 s. The values of PG09 and PG1 decrease by 63% compared with PEEK. It is very important to reach stable friction state of friction pair quickly, which can greatly shorten the running-in time, reduce energy consumption, and improve the performance of composites. Meanwhile, it can be observed that the friction coefficient of PEEK/MoS
2s and PEEK/GO/MoS
2 is gradually increasing at the beginning of the experiment, and then fluctuating in a stable interval in
Figure 2b. The response times of all PEEK/MoS
2s are about 350 s and less than those of PEEK, indicating the MoS
2 content has little effect on response time. It can be concluded that the MoS
2 can effectively reduce the response time of PEEK/MoS
2s. For PG07PM15, the response time is around 380 s, lower than that of PG07, but higher than that of PM15, which decreases by 60% compared with PEEK. The average friction coefficients of all PEEK/MoS
2s decrease with the increasing of MoS
2 content, i.e., the larger the MoS
2 content, the smaller the average friction coefficient (
Figure 2d). For example, the average friction coefficient of PM5 reduces 24% compared with PEEK. Moreover, the PM25 has a minimum average friction coefficient of 0.1602, which reduces by 31% compared with PEEK. However, the average friction coefficient does not result in a significant change when the MoS
2 content exceeds 15 wt.%. For example, the average friction coefficient of PM15 is 0.1626, which decreases by 30% compared with PEEK. Nevertheless, the values of PM20 and PM25 only decrease by 30% and 31%, compared with PEEK, respectively. Furthermore, the average friction coefficient of PG07PM15 is 0.1637, which decreases by 14%, compared with PG07, and increases by 2% compared with PM15 (
Figure 2d). Owing to low hardness, easy plastic deformation, and poor friction heat dissipation of PEEK, its high response time and average friction coefficient is mainly attributed to strong adhesion and plough effect on the rubbing surface [
2,
6,
34]; however, the friction heat loss is improved and the adhesive effect is reduced for composites. On the other hand, GO and MoS
2 play a key role in reducing friction coefficient of composites, owing to their self-lubrication. The better friction reduction of PEEK/MoS
2 than PEEK/GO is attributed to the better self-lubrication of MoS
2; however, the excessive GO leads to the increase of friction coefficient. This is due to the stress concentration and increased fragility, weakening the bonding strength between GO and PEEK matrix.
Figure 2e shows the variation of wear rate (
K) of PEEK, PEEK/GOs, PEEK/MoS
2s, and PEEK/GO/MoS
2. The wear rate of PEEK is 6.213 × 10
−6 mm
3/Nm, which is in the same order of magnitude as the value obtained by Zhang et al. under a dry sliding condition [
2]. The wear rate of PEEK/GOs significantly decreases, and then slightly fluctuates in a very small scale, with increasing of GO content. When the GO content increases from 0.5 wt.% to 0.7 wt.%, the wear rate obviously decreases. However, when the GO content exceeds 0.7 wt.%, the wear rate is basically constant. Regardless of the GO content, the wear rates of all PEEK/GOs are less than PEEK, indicating the positive effect of GO on the wear resistance. The wear rate of PG03 and PG05 reduces by 10% and 15% compared with PEEK, respectively. Nevertheless, PG07 has the lowest wear rate of 4.169 × 10
−6 mm
3/Nm, which decreases by 32.9 % compared with PEEK. In other words, PG07 has the best wear resistance. The variation in wear rate of PEEK/MoS
2s obviously decreases with the increasing of MoS
2 content. When the MoS
2 content is less than 20 wt.%, the wear rate of PEEK/MoS
2s approximately reduces linearly, whereas the difference in wear rate between PM20 and PM25 is minimal; however, the wear resistance of PEEK/MoS
2s is significantly increased in comparison with PEEK. For instance, the PM5 has a maximum wear rate of
K = 5.011 × 10
−6 mm
3/Nm, which decreases by 19% compared with PEEK; whereas PM25 has a minimum wear rate of 3.333 × 10
−6 mm
3/Nm, reduced by 46% compared with PEEK. The wear rate of PG07PM15 decreased by 35% compared with PEEK, which was almost equal to that of PM15 and less than that of PG07. It can be concluded that MoS
2 plays a major role in the wear resistance of PG07PM15. It is worth noting that the abilities of friction reduction and wear resistance of the composites with both GO and MoS
2 are not necessarily better than those of the composites with single GO or MoS
2. The wear resistance and friction reduction of composite are attributed to the enhancement of its hardness, tensile strength, and thermal stability. On the other hand, the self-lubrication properties of GO and MoS
2 also have positive effects on tribological behaviors.
To facilitate the understanding of the friction and wear mechanism, representative SEM images and 3D optical morphologies of the worn surface of PEEK, PEEK/GOs, PEEK/MoS
2s, and PG07PM15 are showed in
Figure 3. The PEEK (
Figure 3a,a’) exhibits uneven grooves parallel to the sliding direction and irregular removed material adhered back to the surface, which shows signs of abrasive wear and adhesive wear mechanism. Meanwhile, there are large amount of debris on the worn surface. The typical combination mechanism of abrasive wear and adhesive wear has been previously observed under a dry sliding condition [
35]. A clear boundary is produced between the worn area and non-worn area, owing to PEEK stacking under the compressive load and friction force. The PEEK surface shows deep wear scratching. The worn surface is very rough, too. The worn surface of PG03 (
Figure 3b,b’) is smoother than PEEK. It exhibits several very shallow grooves, lamellar dissection, fatigue fracture, short stripped pits, and wear debris, indicating wear characteristics of brittle materials to a certain extent. There are obvious partition boundaries similar with PEEK for PG03, PG07, and PG1. However, compared with PEEK, their wear depths are slighter (
Figure 3c,d,c’,d’). Such characteristics may be attributed to the decrease of melting temperature of PEEK/GO, but the wear degree of the PEEK/GOs is obviously different. With increase of GO content, the fatigue wear becomes more and more dominant, and adhesive wear and abrasive wear gradually reduce. This indicates PEEK/GOs show signs of severe fatigue wear, adhesive wear, and mild abrasive wear. The PM5 (
Figure 3e) represents more even grooves parallel to the sliding direction, plastic deformation, randomly adhered material, and wear debris. Nevertheless, the worn surface of PM15 (
Figure 3f) exhibits a small quantity of shallow grooves, adhered material, and wear debris. Simultaneously, a water wave-like morphology is obviously observed. Similar to PEEK, the worn surfaces of PM5 and PM15 result from the same wear mechanism, i.e., mainly abrasive wear and adhesive wear, except the wear degree of individual components is different. Whereas in the case of PM25 (
Figure 3g), the worn surface shows a completely different morphological feature from those of PM5 and PM15. It is very smooth and covered by an almost uniform and essentially continuous light gray layer. In addition, the gray layer can avoid direct contact between composites and counterpart, which further improves the tribological properties. It can be observed that there is very little adhesive material and plastic deformation on the worn surface, but no obvious grooves and wear debris. This indicates that PEEK/MoS
2s possess better friction reduction and wear resistance compared to PEEK and PEEK/GOs. As shown in
Figure 3e’,f’,g’, there are still obvious wear area boundaries. The accumulation of boundary materials is attributed to the plastic deformation of PEEK matrix. With increasing of MoS
2 content, the wear depth reduces and the coverage rate of MoS
2 film increases, which are due to the unique lubrication characteristics of MoS
2. Furthermore, it is worth noting that MoS
2 content has a positive contribution to wear resistance, i.e., the higher the MoS
2 content, the better the wear resistance (
Figure 2c). The PG07PM15 (
Figure 3h,h’) represents very even, shallow, narrow grooves parallel to the sliding direction, a small quantity of microcracks, a water wave-like scratch, spalling pit, and plastic deformation. Its worn surface is smoother than those of PG07 and PM15. It can be observed that the worn surface shows neither the severe sheet spalling in the case of PG07, nor the deep grooves and wear debris in the case of PM15. Compared with other composites, a typical wear area boundary is observed in
Figure 3h’. Such morphology features indicate that the wear mechanism of PG07PM15 is the combination of slight fatigue wear, mild abrasive wear, and adhesive wear. It can be concluded that the synergism between GO and MoS
2 helps to form a good and tenacious lubricant film on the interface between composites and counterpart.
In order to better explain tribological behaviors of composites,
Figure 4 presents a schematic diagram under a dry sliding condition. In the case of PEEK (
Figure 4a), the grooves of worn surface are mostly due to the stiffness of counterpart of steel pin being much higher than that of PEEK. During the friction and wear process, the hard protuberances on the steel pin penetrate into the PEEK, and remove softer PEEK under applied contact load and friction shear. On account of the combination of the plastic deformation, delamination, and friction shear, PEEK is transferred to the counterpart surface to form a lubricant film. For the PEEK/GOs (
Figure 4b), the wear characteristics may be due to the fact that the hardness and tensile strength of the composites is improved by addition of GO with large specific surface area and wrinkle, leading to less flow stress and considerably less plastic deformation for the sliding contact surface, as well as the decrease in the toughness of composites. Moreover, GO may agglomerate in the PEEK matrix, which is more likely to lead to the initiation of microcracks. Under the repeated stresses, the GO will be exposed gradually on the worn surface, carrying most of the applied load, promoting the formation of protective and in situ tribolayer on the interface between the composites and steel pin. These morphology features indicate that the addition of GO changes the main wear mechanism of composites from adhesive wear and abrasive wear to fatigue wear. The strong interface bonding of the composites make it difficult for GO to be detached from the PEEK matrix, and effectively improves the wear resistance of the composites. However, the lubricant films on the interface will be destroyed by desquamate caused by fatigue wear, as well as wear debris generated by fatigue wear. The abrasive wear is produced on the surface of composites by wear debris, which results in the shallow groove marks. Simultaneously, the wear debris plays a certain role in grinding composites, which makes its surface smooth. For PEEK/MoS
2 (
Figure 4c), the MoS
2 is more easily released from the composites and exposed on the worn surface, owing to weaker bonding strength of MoS
2 and PEEK matrix. More importantly, the MoS
2 layered structure formed via weak van der Waals interactions between each layer gives better tribological performance, which makes it easier to form a continuous and complete lubricant layer between the interfaces. At the same time, the softer lubricant films can effectively cover the worn surface, fill the wear marks and the wear debris. This corresponds to the better friction reduction and wear resistance of PEEK/MoS
2s in
Figure 2b,c. However, for PG07PM15, GO and MoS
2 would be concurrently exposed on the worn surface during the tribological test in
Figure 4d. The GO with high mechanical strength can withstand load applied, and plays a certain role in lubrication and reduce composites from being further worn off, and then the MoS
2 mostly plays the lubricant effect. In a word, the formed lubricant films provide a low-strength junction at the interface, resulting in lower friction coefficient and wear rate than PEEK. So, GO and MoS
2 can effectively avoid direct touching of the friction pair surfaces, thus further inhibiting the wear of composites. However, the generated heat, which cannot be taken away as soon as possible, or dissipated during the process of friction, will lead to the further softening of the composites and the enhancement of adhesive wear. The effects of heat distribution and dissipation on performance are important [
36]. In the meantime, the abrasive debris, difficultly discharged from the worn surface, will result in the inevitable abrasive wear under a dry sliding condition. On the contrary, the phenomenon of wear surface adhesion and a large amount of debris deposition are almost not observed, due to the presence of water lubrication [
3].
It is well known that the nanofillers have an inherent tendency to form agglomerates, due to the large surface area in the polymer matrix, and lead to decrease of the reinforcing effects [
3,
11,
25,
37]. The uniform distribution has an important effect on its hardness, tensile strength, and tribological properties. In order to evaluate the dispersion of GO and MoS
2, the EDS analysis of element mapping was carried out for PG07PM15 in
Figure 5. It can be clearly observed that MoS
2 was evenly dispersed in the PEEK matrix, but, as PEEK and GO both are mainly composed of carbon, the distribution of GO in PEEK cannot be visually observed through EDS mapping images. It is hoped that another way or method can be explored in future research. This indicates that the hot isostatic pressing method used to prepare PEEK composites is reasonable and effective in this work, and the prepared composites could meet the experimental requirements.
3.2. Mechanical Properties
The tribological properties of a material are closely related to its hardness, which is always considered to represent the scratch resistance, abrasive resistance, and cutting resistance of the material surface [
38,
39]. To investigate the reason for the super wear performance of composites, the Vickers hardness and tensile strength of PEEK, PEEK/GOs, PEEK/MoS
2s, and PEEK/GO/MoS
2 are shown in
Figure 6. As shown in
Figure 6a, the Vickers hardness of PEEK/GOs increases with the increasing of GO content. The Vickers hardness of PG03 is 25.65 HV, and the value of PG1 reaches 28.39 HV, which is about 15% higher than that of PEEK. Interestingly, the Vickers hardness of PG07 is around 13% higher than that of PEEK, which is only 2% lower than that of PG1. This variation trend of the Vickers hardness of PEEK/GO is consistent with the reported results of Puértolas et al. [
3] and Lin et al. [
38]. It can be seen that the Vickers hardness of PEEK/MoS
2s is gradually increased firstly, and then the Vickers hardness decreases after the MoS
2 content beyond 15 wt.%. The PM15 is the largest, which is 25.26 HV, and it is about 2% higher than that of PEEK. PM20 and PM25 are almost the same as that of PEEK. This coincides with the results of Zalaznik et al. [
25,
40]. The Vickers hardness of PG07PM15 is around 1.1 times that of PEEK, and it is 26.42 HV. Interestingly, its Vickers hardness is 5% higher than that of PM15, but 6% lower than that of PG07, so this confirms that GO has a better effect on improving the Vickers hardness of composites than MoS
2. The addition of GO or MoS
2 could reinforce the formation of cross-networked structures and produce reorganization of amorphous polymer chains in composites, which provide a higher resistance against the normal load applied. The higher hardness of PEEK/GOs is related to the high hardness of GO [
41]. The excellent hardness is beneficial to the tribological properties of composites, which coincides with the results of
Figure 2.
In general, the tensile strength of a material has a close relationship with its tribological behaviors [
2,
3]. As shown in
Figure 6b, the tensile strength of PEEK is 108.3 MPa, which is in agreement with Yang et al. [
42]. The tensile strength of PEEK/GOs increases with increasing of GO content. When the content is lower than 0.7 wt.%, the tensile strength and GO content has a linear relationship. After that, the increment of tensile strength begins to decrease. The tensile strength of PG07 is 6% higher than that of PEEK, while PG1 is only 0.5% higher than that of PG07. Moreover, the introduction of GO could drastically affect the toughness of composites. The contribution of GO to the improvement of tensile strength is due to GO having good compatibility with PEEK and strong binding force, which improves the crystallinity and uniformity of PEEK matrix. However, when the GO content exceeds 0.7 wt.%, the tensile strength of the PEEK/GOs increases very little, due to the stress concentration caused by GO agglomeration. The tensile strength of PEEK/MoS
2s increases firstly, and then decreases with the increasing of MoS
2 content. The tensile strength of PM10 is 111.1 MPa, and it is 3% higher than that of PEEK. Interestingly, when the MoS
2 content exceeds 15 wt.%, the tensile strength of PEEK/MoS
2 is lower than that of PEEK. The tensile strength of PM25 decreases 12% compared with that of PEEK. It is worth noting that the MoS
2 has a positive effect on the tensile strength of composites, but excessive content will have a negative effect. The lamellar structure of MoS
2 causes shear slip of the composites under stress. The tensile strength of PG07PM15 is smaller than PG07 and higher than PEEK and PM15; its value reaches to 112.8 MPa, which coincides with the results of hardness. As stated above, GO is superior to MoS
2 in improvement of tensile strength of PEEK matrix, which coincides with the results of Mittal et al. [
43] and Puértolas et al. [
3].
In order to investigate the fracture properties of composite, SEM images of fracture surface of the PEEK, PG07, PM15, and PG07PM15 after the tensile test are shown in
Figure 7. The fracture surface of PEEK (
Figure 7a) exhibits a relative smooth river and tongue pattern, in which ravines and penetrating cracks appear, but there is no obvious fracture dimple, fracture gully, and tearing shape, reflecting that PEEK has a typical brittle fracture mode due to high strain rate, and its tensile resistance is likely the fiber cracking and pill-out. For PG07 (
Figure 7b), the fracture surface shows a relatively rough squamous-like morphology, on account of its brittleness, which is consistent with the decrease in toughness. The GO has good compatibility with PEEK matrix interface, which would cause crack deflection and disproportionation, could absorb tensile fracture energy, and hinder the crack propagation. Therefore, the tensile strength of PEEK/GO can be improved. In the case of PM15 (
Figure 7c), the fracture surface exhibits a rougher morphology, with wrinkle and a large amount of sheet faults, in comparison with PEEK, which indicates more effective reinforcement of the toughness. The MoS
2 are embedded in PEEK in its original layered structure, which is more likely to cause interlaminar slippage of the PEEK/MoS
2. Meanwhile, the addition of MoS
2 would destroy the continuity of PEEK matrix, and there are many microcracks around it, in which crack initiation and crack propagation are promoted. It coincides with the tensile strength of PEEK/MoS
2. As shown in
Figure 7d, the fracture surface of the PG07PM15 shows a relatively rougher morphology in comparison with that of PEEK, exhibiting different features in comparison with PG07 and PM15, which indicates that the presence of GO and MoS
2 can dissipate much energy and hinder the crack propagation during the fracture process. It can be observed that there are wrinkles, stacked GO sheets (blue arrows in
Figure 7), microvoids, and microcracks on the fracture surface. However, the obvious agglomeration phenomenon is not observed on the fracture surface, indicating a homogeneous dispersion of GO/MoS
2 in the PEEK matrix. This coincides with the results of Vickers hardness and tensile strength.
3.3. Thermal Properties
In order to further understand the super tribological performance of composites, the thermal behaviors are estimated.
Figure 8 illustrates the DSC and TGA curves of PEEK, PEEK/GOs, and PEEK/MoS
2s. As shown in
Figure 8a, it can be observed that the melting temperature of PEEK, PG03, PG05, PG07, PG09, and PG1 is about 348.9 °C, 348.9 °C, 343.9 °C, 343.9 °C, 343.9 °C, and 343.9 °C, respectively. The decrease of melting temperature of PEEK/GO is attributed to the addition of GO. When the GO content exceeds 0.5 wt.%, the melting temperature is constant, and is 5 °C lower than that of PEEK. This coincides with the results of Puértolas et al. [
3]. Interestingly, the melting temperatures of all PEEK/MoS
2s are almost the same, their values are all around 343.9 °C, and lower than that of PEEK in the
Figure 8b, which coincides with the results reported by Zalaznik et al. [
25]. It is worth noting that the MoS
2 content has no significant effect on the melting temperature. The reason is because the addition of MoS
2 or GO destroys the continuity of PEEK matrix, and the motion of the polymer chain is segmented during crystallization (white arrows in
Figure 7), which inhibits molecular stacking (blue arrows in
Figure 7) and results in the formation of less-perfect crystallites. As shown in
Figure 8c, the PEEK/GOs exhibit a similar thermal decomposition process with PEEK, which can be mainly ascribed to the decomposition of PEEK polymer molecular chains. The onset decomposition temperature of PEEK/GOs increases with the increasing of GO content. The onset decomposition temperature and weight residue of PEEK are approximately 578 °C and 56%, respectively. Compared to PEEK, for lower GO content, i.e., PG03 and PG05, the thermal decomposition temperature improves by 5 °C, while that of PG07, PG09, and PG1 improves by about 10 °C. In consequence, it can be seen that GO can effectively increase the thermal stability of composites, and PEEK/GOs have higher thermal stability with increasing GO content. It is observed that PEEK/MoS
2s exhibit a different thermal decomposition process to the PEEK (
Figure 8d), which can be ascribed to the barrier effect of MoS
2 on the diffusion of PEEK. The decomposition temperature of PM5 and PM10 is approximately 603 °C. Nevertheless, the decomposition temperature of composites reaches 608 °C when the MoS
2 content exceeds 15 wt.%. It is higher than PEEK by about 25–30 °C. It indicates that the presence of MoS
2 can effectively enhance the thermal stability of PEEK matrix. The thermal properties of composites are in relation with combined morphology between GO or/and MoS
2 and PEEK matrix (
Figure 7). However, the weight residue of PM5, PM10, PM15, PM20, and PM25 is 74%, 74%, 76%, 77%, and 79%, respectively. Compared with PEEK, the weight residue increases 32~41%. Similar observations were reported by Chen et al. [
12] and Gong et al. [
44]. Therefore, it can be seen that the MoS
2 can obviously promote the thermal stability of composites. Furthermore, it can be worth noting that the effect of MoS
2 on the thermal stability of composites is better than GO. This is mainly attributed to MoS
2‘s excellent high temperature resistance and the interaction between PEEK matrix, which results in the formation of a stable network structure within the composites, preventing the PEEK chain segments (white arrows in
Figure 7) from decomposition at high temperature [
7,
45]. In addition, the charring effect of MoS
2 at high temperature also contributes to the thermal stability of composites. It is well known that a large amount of heat will be generated during the process of friction and wear, especially under dry friction conditions [
46], and it will lead to excessive degradation of the polymer and poor tribological properties. Therefore, the excellent thermal stability of composites effectively prevents the high temperature harm, and improves their tribological performances in dry sliding friction and wear conditions.