Effect of Counterbody on Friction and Wear Properties of Copper-MgP-Graphite Composites Prepared by Powder Metallurgy

: The purpose of this study was to investigate the inﬂuence of different counterbodies against Cu/magnesium phosphate treated graphite (Cu-MgPG) composite materials to ﬁnd the best material combination in terms of friction coefﬁcient and speciﬁc wear amount. A Cu matrix composite reinforced with 10 vol% magnesium phosphate treated graphite and pure Cu powder were prepared by powder metallurgy techniques under the same consolidation processing condition. The friction and wear properties of the composites were investigated at 10 N using a pin-on-disc tribometer on Al 2 O 3 , SiC and SUJ2 bearing steel counterbodies. The Cu-MgPG/Al 2 O 3 pair showed the lowest friction coefﬁcient, but the speciﬁc wear rate tended to increase slightly when compared with Cu/Al 2 O 3 pair. On the other hands, the Cu-MgPG/SUJ2 pair showed about the same speciﬁc wear rate as the Cu/SUJ2 pair, but the friction coefﬁcient was signiﬁcantly reduced. These phenomena are thought to be due to the fact that the added graphite acts as a solid lubricant during sliding and also suppresses the oxidation behavior of the sliding material.

Copper-graphite composites possess both copper and graphite properties, including excellent thermal and electrical conductivity, solid lubrication, low coefficient of thermal expansion and so on, which makes these composite widely used as a brush and bearing material in many applications [3,7]. In fact, components in this type of application are often exposed to high friction, temperature rise and environmental erosion [8]. The good friction properties of copper-graphite composites are often due to the fact that the graphite particles exposed on the surface of the composites act as lubricants during the friction process [9][10][11]. Considering that graphite is prone to oxidation in high temperature environments, the operating temperature range of ordinary copper-graphite composites is also limited.
In a previous work [12], it proposed a method to improve the oxidation resistance of graphite while maintaining the lubricating properties of graphite. In this study, the magnesium phosphate-treated graphite (MgPG) was used as a solid lubricating phase to mix with copper powder, and used powder metallurgy technique to make copper/magnesium phosphate-treated graphite composite materials (Cu-MgPG).
In the preparation of copper-graphite composite materials, many studies [9,[13][14][15][16] have investigated the effect of the content of graphite added on the properties of the final composites. While this work aims to study the friction properties of composites made of 10 vol% magnesium phosphate treated graphite and copper. Research [14] has shown that when the volume content of graphite is less than 50%, sintering temperature lies in 700 • C-950 • C to improve the metallurgy bonding between copper and copper in the material. At the same time, the raw material with treated graphite [12] used in this study was produced by sintered natural graphite with magnesium phosphate at 800 • C. Based on the above data, the sintering temperature of the composite material selected in this study was 800 • C.
Many scholars have studied the tribological properties of copper-graphite composite materials under different conditions, and the results show that the tribological behavior of copper-graphite composite materials is complex and dependent on external factors. These important external factors include applied load, sliding velocity, ambient temperature, humidity, and counterbody material. However, there are few studies on the effect of counterbody materials on the friction properties of copper-graphite composite materials.
In this study, powder metallurgy technique was used to prepare Cu-MgPG composite materials and a comparative sample of pure copper (Cu) materials prepared under the same conditions. Then we compared the physical properties, friction properties and surface microstructure of the pure copper materials and Cu-MgPG composites. In addition, in order to find the optimum tribopair combination of Cu and Cu-MgPG with the counterbody materials, the influence of the counterbody materials in terms of coefficient of friction and wear was investigated.

Preparation of the Samples
The magnesium phosphate-treated graphite (MgPG) raw material used in this study was prepared according to the method of previous study [12]. The arithmetic average particle size of the magnesium phosphate-treated graphite (natural graphite powder, D50 = 93.16 µm; Mg(H 2 PO 4 ) 2 ·4H 2 O, MW:290.34, JUNSEIKAGAKU) used in this work was 131.70 µm.
The composites were fabricated by powder metallurgy technique. 10 vol% MgPG powder and copper powder (Nilaco, Ginza, Tokyo, Japan) were dry-mixed with a bench kneader machine (Irie Shokai Co., Ltd., Tokyo, Japan) for 3 h. After mixing, the powder mixture was first cold pressed at 30 MPa in 35 mm × 35 mm square mold, and then hotpressing in nitrogen flow (2 L/min) at 800 • C for 3 h with a uniaxial pressure of 40 kN. For comparison, parallel compacts made from pure copper powders were consolidated under the same conditions applied for the preparation of Cu-MgPG composites. All specimens were polished to get a mirror-effect and cleaned with acetone before every experiment. The polished surfaces were then observed by SEM (JEOL Ltd., JSM-7500F, Tokyo Japan). The typical SEM micrograph of the polished surface of Cu-MgPG composites and copper are shown in Figure 1. The magnesium phosphate-treated graphite particles dispersed in the copper matrix can be observed in Figure 1A.

Physical Properties
The densities of the sintered copper/MgP-graphite composite and copper block were measured by Archimedes' method. The theoretical density was calculated by dividing the bulk density of the sintered composite by the theoretical density calculated from the rule

Physical Properties
The densities of the sintered copper/MgP-graphite composite and copper block were measured by Archimedes' method. The theoretical density was calculated by dividing the bulk density of the sintered composite by the theoretical density calculated from the rule of mixtures using 8.95 g/cm 3 , and 2.25 g/cm 3 as densities for copper and magnesium phosphate, respectively. The hardness was investigated using a Micro Vickers hardness tester (HMV-G, SHIMADZU Corp., Kyoto, Japan) under a load of 0.5 kg with a dwell time of 10 s. Each sample was measured five times, and the data obtained were the average value. The physical properties of the sintered composites are shown in Table 1. The density of sintered samples reached 99% of the theoretical density, which demonstrated the efficiency of the powder metallurgy technology in producing high-density materials. From Table 1, it can also be noted that the addition of graphite increased the hardness of copper materials but decreased the apparent density of the sintered composites.

Friction Coefficient and Wear Test
The friction coefficient and wear sliding tests was carried out in a pin-on-disc tribometer (T-18-0162, NANOVEA Corp., Irvine, CA, USA) in ambient conditions (20 ± 0.5 • C and 45 ± 5% RH). The tribological properties of the prepared composite materials were investigated in a dry sliding test. Counterbodies used in the sliding test are 8 mm diameter polished balls (the average surface roughness is less than 0.01 µm) made from commercially available balls-SiC, Al 2 O 3 , SUJ2 bearing steel (Sato Tekkou Corp., Oita, Japan). The hardness values of SiC, Al 2 O 3 , and SUJ2 balls are 2400 HV, 1600 HV, and 770 HV, respectively, which are obtained from the supplier. Before test, both the composite materials and the counterbody were ultrasonically cleaned in an acetone bath for 10 min. Sliding friction and wear tests were performed with a circular sliding under a load of 10 N and the sliding velocity was 0.1 m/s, the wear track radius was 5 mm, and the total sliding distance was 2000 m. There are no other lubricants added in each sliding test. Figure 2 shows the schematic illustration of the sliding tester. Three tests were performed for each counterbody. The friction coefficients were continuously recorded and the wear volume on each sample disc and counterbody was calculated from the surface profile traces across the wear track using a surface profilometer (Mitutoyo Corp., 178-570-01, Kawasaki, Japan).
The worn volume of the disc (V d , mm 3 ) was calculated according to the following equation [17]: where S (mm 2 ) and r (mm) are the cross-sectional areas of the wear track, each at 90 • with respect to the previous one, and sliding radius, respectively. The worn volume of the ball (V b , mm 3 ) was calculated according to the following equations [17]: where h (mm) is the height of the removed material, R (mm) is original radius of the ball, and r ω (mm) is the radius of wear scar on the ball. The specific wear rate (WR, mm 3 /N·m) was further calculated according to the following equation [17]: where V (mm 3 ) is the wear volume of the disc (V d ) or ball (V b ), L (N) is the normal load, and S (m) is the total sliding distance.
The worn surfaces of sample disc after sliding test were studied by SEM and EDS to identify wear mechanisms.
with respect to the previous one, and sliding radius, respectively.
The worn volume of the ball ( , mm 3 ) was calculated according to the following equations [17]: (mm) is original radius of the ball, and (mm) is the radius of wear scar on the ball.
The specific wear rate ( , mm 3 /N·m) was further calculated according to the following equation [17]: is the wear volume of the disc ( ) or ball ( ), (N) is the normal load, and (m) is the total sliding distance.
The worn surfaces of sample disc after sliding test were studied by SEM and EDS to identify wear mechanisms.

Figure 2.
Schematic illustration of the pin on disc tester.

Evaluation Method of Worn Surface
The surface morphology and elemental distribution of the samples were characterized using a scanning electron microscope (SEM, JSM-7500F, JEOL Corp., Tokyo, Japan) equipped with energy dispersive X-ray detector (EDS, JEOL Corp., Tokyo, Japan). Worn surfaces of the samples were also investigated by an X-ray photoelectron spectroscope (XPS) to understand wear-induced surface chemistry modification.

Friction Coefficient and Wear Rate
Typical evolution of the friction coefficients of copper block sliding against SiC, Al2O3, and SUJ2 counterbodies at 10N are shown in Figure 3. All of the friction coefficient diagrams contain two stages such as Figure 3, named the (1) run-in and (2) steady state stage. In the run-in stage, the friction coefficient shows a large and irregular increase and

Evaluation Method of Worn Surface
The surface morphology and elemental distribution of the samples were characterized using a scanning electron microscope (SEM, JSM-7500F, JEOL Corp., Tokyo, Japan) equipped with energy dispersive X-ray detector (EDS, JEOL Corp., Tokyo, Japan). Worn surfaces of the samples were also investigated by an X-ray photoelectron spectroscope (XPS) to understand wear-induced surface chemistry modification.

Friction Coefficient and Wear Rate
Typical evolution of the friction coefficients of copper block sliding against SiC, Al 2 O 3 , and SUJ2 counterbodies at 10 N are shown in Figure 3. All of the friction coefficient diagrams contain two stages such as Figure 3, named the (1) run-in and (2) steady state stage. In the run-in stage, the friction coefficient shows a large and irregular increase and decrease for a period of time; in the steady-state stage, the friction coefficient oscillates continuously within a certain range. The formation of the run-in state stage may be related to the increase of the contact area, the work hardening effect of wear and the accumulation of debris at the pin-disk interface, and the change of the wear mechanism from two-body wear to three-body wear [18]. The reported friction coefficient in this study is the average of these data in the steady stage.
An initial run-in period followed by a steady-state period was observed in all of the tribopairs, while in the run-in stage, the friction coefficient of copper block fluctuates within certain range in a fluctuating manner against SiC, SuJ2, and Al 2 O 3 . The friction coefficient of copper block changed in the range of 0.65-0.85 against SiC, Al 2 O 3 , and SUJ2 counterbodies. When sliding against different counterbodies, the Cu/Al 2 O 3 tribopair has the lowest coefficient of friction. Cu/SiC tribopair showed the highest friction coefficient in the steady-state period.
decrease for a period of time; in the steady-state stage, the friction coefficient oscillates continuously within a certain range. The formation of the run-in state stage may be related to the increase of the contact area, the work hardening effect of wear and the accumulation of debris at the pin-disk interface, and the change of the wear mechanism from two-body wear to three-body wear [18]. The reported friction coefficient in this study is the average of these data in the steady stage. An initial run-in period followed by a steady-state period was observed in all of the tribopairs, while in the run-in stage, the friction coefficient of copper block fluctuates within certain range in a fluctuating manner against SiC, SuJ2, and Al2O3. The friction coefficient of copper block changed in the range of 0.65-0.85 against SiC, Al2O3, and SUJ2 counterbodies. When sliding against different counterbodies, the Cu/Al2O3 tribopair has the lowest coefficient of friction. Cu/SiC tribopair showed the highest friction coefficient in the steady-state period.
Typical evolution of the friction coefficients of Cu-MgPG composites sliding against SiC, Al2O3, and SUJ2 counterbodies at 10N are shown in Figure 4. The average friction coefficients of all of the tribopairs in steady-state period are summarized in Figure 5.   An initial run-in period followed by a steady-state period was observed in all of the tribopairs, while in the run-in stage, the friction coefficient of copper block fluctuates within certain range in a fluctuating manner against SiC, SuJ2, and Al2O3. The friction coefficient of copper block changed in the range of 0.65-0.85 against SiC, Al2O3, and SUJ2 counterbodies. When sliding against different counterbodies, the Cu/Al2O3 tribopair has the lowest coefficient of friction. Cu/SiC tribopair showed the highest friction coefficient in the steady-state period.
Typical evolution of the friction coefficients of Cu-MgPG composites sliding against SiC, Al2O3, and SUJ2 counterbodies at 10N are shown in Figure 4. The average friction coefficients of all of the tribopairs in steady-state period are summarized in Figure 5.  The friction coefficient detected with the tribopairs of Cu-MgPG composite material is significantly lower than that of the pure copper material. The friction coefficient of Cu-MgPG composites changed in the range of 0.15-0.35 against SiC, Al2O3, and SUJ2 counterbodies. Similar to the results of copper tribopair, Cu-MgPG/Al2O3 tribopair has the lowest coefficient of friction. However, the Cu-MgPG/SUJ2 tribopair showed the highest friction coefficient in the steady-state period. From Figure 5, it can be noted that the addition of magnesium phosphate treated graphite decreased the friction coefficient of the sintered composites against each counterbody.
The specific wear rate ( ) of the sintered disc is displayed in Figure 6, and the re- The friction coefficient detected with the tribopairs of Cu-MgPG composite material is significantly lower than that of the pure copper material. The friction coefficient of Cu-MgPG composites changed in the range of 0.15-0.35 against SiC, Al 2 O 3 , and SUJ2 counterbodies. Similar to the results of copper tribopair, Cu-MgPG/Al 2 O 3 tribopair has the lowest coefficient of friction. However, the Cu-MgPG/SUJ2 tribopair showed the highest friction coefficient in the steady-state period.
From Figure 5, it can be noted that the addition of magnesium phosphate treated graphite decreased the friction coefficient of the sintered composites against each counterbody.
The specific wear rate (WR) of the sintered disc is displayed in Figure 6, and the results showed WR of sintered copper discs changed in the range of 3.13 × 10 −6 -6.27 × 10 −6 mm 3 /N·m against Al 2 O 3 , SiC, and SUJ2 counterbodies at a load of 10 N. From Figure 5, it can be noted that the addition of magnesium phosphate treated graphite decreased the friction coefficient of the sintered composites against each counterbody.
The specific wear rate ( ) of the sintered disc is displayed in Figure 6, and the results showed WR of sintered copper discs changed in the range of 3.13 × 10 −6 -6.27 × 10 −6 mm 3 /N·m against Al2O3, SiC, and SUJ2 counterbodies at a load of 10 N. For the sintered copper disc, the lowest wear rate disappeared when sliding against Al2O3 counterbody, which is similar to the relationship of the friction coefficient. However, the highest WR disappeared when sliding against SUJ2 counterbody that is different For the sintered copper disc, the lowest wear rate disappeared when sliding against Al 2 O 3 counterbody, which is similar to the relationship of the friction coefficient. However, the highest WR disappeared when sliding against SUJ2 counterbody that is different with the result of the friction coefficient. In addition, the WR results with sintered Cu-MgPG disc also shows that the lowest wear rate and highest wear rate disappeared when sliding against Al 2 O 3 and SUJ2 counterbody, respectively. Considering the data in Figures 5 and 6, the specific wear rate of the sintered disc against SiC, Al 2 O 3 , SUJ2 counterbodies didn't show obvious dependence on their friction coefficients. At the same time, even though the average friction coefficient of the sintered Cu-MgPG composite materials decreased significantly with sintered copper materials, the specific wear rate increased instead when sliding against SiC and Al 2 O 3 ball.
The average specific wear rates of SiC, Al 2 O 3 , SUJ2 counterbodies with two sintered discs are summarized in Figure 7.
It can be found from Figure 7 that Al 2 O 3 ball exhibited lower specific wear rate when sliding against two kinds of sintered disc compared with other counterbodies. The specific wear rate of SiC ball was the highest when sliding against sintered Cu-MgPG composite materials which is corresponding to the result shown in Figure 6. From the comparison between Figures 6 and 7, it is clear that the specific wear rates between the sintered disc and counterbodies were consistent. 5 and 6, the specific wear rate of the sintered disc against SiC, Al2O3, SUJ2 counterbodies didn't show obvious dependence on their friction coefficients. At the same time, even though the average friction coefficient of the sintered Cu-MgPG composite materials decreased significantly with sintered copper materials, the specific wear rate increased instead when sliding against SiC and Al2O3 ball.
The average specific wear rates of SiC, Al2O3, SUJ2 counterbodies with two sintered discs are summarized in Figure 7. It can be found from Figure 7 that Al2O3 ball exhibited lower specific wear rate when sliding against two kinds of sintered disc compared with other counterbodies. The specific wear rate of SiC ball was the highest when sliding against sintered Cu-MgPG composite materials which is corresponding to the result shown in Figure 6. From the comparison between Figures 6 and 7, it is clear that the specific wear rates between the sintered disc and counterbodies were consistent.      Figure  8D-F displayed the worn surface of the sintered Cu-MgPG sample. It can be seen that after the sliding experiment, the worn surface left on the sintered Cu-MgPG sample is different from sintered copper sample. Clear grooves appear on the wear surface of the graphite-added sintered samples. From the analysis of the results of EDS (Figure 9), it can be inferred that the formation of these grooves is related to the oxides formed on the sur-   Figure 8D-F displayed the worn surface of the sintered Cu-MgPG sample. It can be seen that after the sliding experiment, the worn surface left on the sintered Cu-MgPG sample is different from sintered copper sample. Clear grooves appear on the wear surface of the graphite-added sintered samples. From the analysis of the results of EDS (Figure 9), it can be inferred that the formation of these grooves is related to the oxides formed on the surface.

The Morphology Analysis of Worn Surface
In addition, SEM images of Figure 8 D, E showed that there was a partial detachment of graphite form the inner region of the pockets. For the samples with graphite addition, graphite particles agglomerate in regions between Cu matrix grains, stored in pockets with internal voids [19]. At the same time, it indicates that the reduced density and low theoretical density (as shown in Table 1) of the graphite-containing samples is associated with the internal voids in pockets of graphite present in the composites. Meanwhile, in Figure 8F, graphite particles staying on the worn surface are observed, which are presumably exposed from the graphite-storing voids during the friction process.
Elemental analysis of all samples after sliding experiments showed the presence of a large amount of oxygen on the wear track. However, material shedding, plastic deformation and numerous grooves can be clearly seen on the wear track of the graphite added sintered samples. There are only a few grooves on the wear track surface of the sintered copper sample. This proves that different oxides are formed on the surface of the two samples. No obvious graphite film formation was observed in the EDS mapping in Figure 9D,E, but observed on the wear track formed by the Cu-MgPG/SuJ2 tribopair shown in Figure 9F, there are some linear tracks formed by the graphite element.
According to the results in Figure 5, it can be known that the lowest friction coefficient of sintered disc appeared when sliding against Al 2 O 3 counterbody, while the hardness of the Al 2 O 3 ball is in the middle of the other two counterbodies. It was proved that the hardness of the counterbody is not the key to determining the friction coefficient. Combined with the EDS results, it is assumed that the reduction in the average friction coefficient of Cu/Al 2 O 3 and Cu-MgPG/Al 2 O 3 tribopairs should be related to the oxides produced on the worn surface during the friction process.
XPS results in Figures 10 and 11 were also conducted on the sintered copper sample and sintered Cu-MgPG sample surfaces after sliding against different counterbodies to confirm the oxidation in sliding. Results from XPS analysis for Al 2p , O 1s , C 1s core levels obtained on the worn surface of sintered samples sliding against Al 2 O 3 counterbody are provided in Figure 10. copper sample. This proves that different oxides are formed on the surface of the two samples. No obvious graphite film formation was observed in the EDS mapping in Figure  9D,E, but observed on the wear track formed by the Cu-MgPG/SuJ2 tribopair shown in Figure 9F, there are some linear tracks formed by the graphite element.
According to the results in Figure 5, it can be known that the lowest friction coefficient of sintered disc appeared when sliding against Al2O3 counterbody, while the hardness of the Al2O3 ball is in the middle of the other two counterbodies. It was proved that the hardness of the counterbody is not the key to determining the friction coefficient. Combined with the EDS results, it is assumed that the reduction in the average friction coefficient of Cu/Al2O3 and Cu-MgPG/Al2O3 tribopairs should be related to the oxides produced on the worn surface during the friction process.
XPS results in Figures 10 and 11 were also conducted on the sintered copper sample and sintered Cu-MgPG sample surfaces after sliding against different counterbodies to confirm the oxidation in sliding. Results from XPS analysis for Al2p, O1s, C1s core levels obtained on the worn surface of sintered samples sliding against Al2O3 counterbody are provided in Figure 10.
As shown in Figure 10A,D, the Al2p peak(P1 + P2 + P3) fitted to the original peak is composed of three peaks P1, P2, P3 (as revealed by Gaussian curve fitting), each of which are assigned to different bonds containing Al. The Al2p peak at 75 eV corresponds to the presence of Al2O3 [20]. Meanwhile, there are also two unknown peaks appeared in copper sample and Cu-MgPG sample. The fitted O1s peak was deconvoluted into three parts using Gaussian curve fitting. The two main O1s peak at 530.35 eV and 531.4 eV appeared in Figure 10B correspond to the presence of CuO [21] and Al2O3 [22] or Al2CuO4 [23] respectively. While as shown in Figure 10E, the O1s peak at 530.2 eV corresponds to the presence of CuO2 [24], O1s peak at 532.0 eV belongs to Al2O3 [25]. The C element in the EDS results ( Figure 9A) indicates that there is also carbon in the wear surface of the pure copper sample. After the analysis of C1s by XPS, the fitted C1s peak was deconvoluted into three separate peaks using Gaussian curve fitting. C1s peak at 284.8 eV as shown in Figure 10C, and 284.9 eV as shown in Figure 10F corresponds to the presence of diamond structure [26,27]. It can be speculated that there are some abrasive particles (diamond) remaining after the copper sample is polished. In addition, the C1s peak at 285.7 eV in Figure 10F corresponds to the presence of graphite [28]. This proves that during the friction process, there are graphite particles exposed from the graphite-storing voids and stay on the friction contact surface.  SUJ2 counterbody remaining on the disc surface. The Mn2p peak at 642.2 eV corresponds to the presence of MnO2 [29] and the peak at 650.0 eV corresponds to the presence of Mn [30]. However, the peak representing the Mn element is not significant on the Cu-MgPG sample. The Si2p peak at 103.1 eV corresponds to the presence of SiO2 [31]. The peak representing the Si element is not significant on the Cu/MgPG sample. Combined with the results of the Specific wear rate shown in Figures 6 and 7, it can be speculated that the sliding test with Cu-MgPG sample against SUJ2 counterbody can reduce the average friction coefficient while reducing wear.

Conclusions
In this study, a Cu-MgPG material was prepared, and the experimental results proved that the Cu-MgPG material in this study has self-lubricating properties. In addition, under fixed experimental conditions (fixed load 10N and sliding speed 0.1 m/s), the optimum tribopair combination of Cu and Cu-MgPG with the counterbody material, the influence of the counterbody materials in terms of coefficient of friction and wear was examined, and the following findings were obtained.
(1) The friction coefficient detected with the tribopairs of Cu-MgPG composite material is significantly lower than that of the pure copper material since the added graphite acts as a solid lubricant.
(2) The Cu-MgPG/SUJ2 pair was found to decrease both the friction coefficient and specific wear amount compared to the Cu/SUJ2 pair. From the results of XPS and EDS analysis of the sliding surfaces, it is considered that the added MgPG acted as a solid lubricant and suppressed the oxidation behavior of the material. On the other hand, the Cu-MgPG/Al2O3 pair was found to have the lowest coefficient of friction and specific wear of all of the pairs.
In general, the results of this study have demonstrated the application potential of Cu/MgPG materials under specific circumstances. However, in the case of practical application, there are changes in loads and sliding speeds, so the materials prepared in this study cannot be immediately put into the application of brush or bearing materials. In future research, it is necessary to continue studying the effects of sliding loads and sliding speeds on the coefficient of friction and wear with different tribopairs.  As shown in Figure 10A,D, the Al 2p peak(P 1 + P 2 + P 3 ) fitted to the original peak is composed of three peaks P1, P2, P3 (as revealed by Gaussian curve fitting), each of which are assigned to different bonds containing Al. The Al 2p peak at 75 eV corresponds to the presence of Al 2 O 3 [20]. Meanwhile, there are also two unknown peaks appeared in copper sample and Cu-MgPG sample. The fitted O 1s peak was deconvoluted into three parts using Gaussian curve fitting. The two main O 1s peak at 530.35 eV and 531.4 eV appeared in Figure 10B correspond to the presence of CuO [21] and Al 2 O 3 [22] or Al 2 CuO 4 [23] respectively. While as shown in Figure 10E, the O 1s peak at 530.2 eV corresponds to the presence of CuO 2 [24], O 1s peak at 532.0 eV belongs to Al 2 O 3 [25]. The C element in the EDS results ( Figure 9A) indicates that there is also carbon in the wear surface of the pure copper sample. After the analysis of C 1s by XPS, the fitted C 1s peak was deconvoluted into three separate peaks using Gaussian curve fitting. C 1s peak at 284.8 eV as shown in Figure 10C, and 284.9 eV as shown in Figure 10F corresponds to the presence of diamond structure [26,27]. It can be speculated that there are some abrasive particles (diamond) remaining after the copper sample is polished. In addition, the C 1s peak at 285.7 eV in Figure 10F corresponds to the presence of graphite [28]. This proves that during the friction process, there are graphite particles exposed from the graphite-storing voids and stay on the friction contact surface.
XPS analysis of worn surfaces of copper and Cu-MgPG composite materials against SuJ2 counterbody and SiC counterbody is shown in Figure 11. Compared with the analysis results of EDS (as shown in Figure 9C), the presence of Mn element was detected on the wear surface with XPS analysis ( Figure 11A) after the copper sample sliding against SUJ2 counterbody. It can be speculated that the Mn element comes from the debris of the SUJ2 counterbody remaining on the disc surface. The Mn 2p peak at 642.2 eV corresponds to the presence of MnO 2 [29] and the peak at 650.0 eV corresponds to the presence of Mn [30]. However, the peak representing the Mn element is not significant on the Cu-MgPG sample. The Si 2p peak at 103.1 eV corresponds to the presence of SiO 2 [31]. The peak representing the Si element is not significant on the Cu/MgPG sample.
Combined with the results of the Specific wear rate shown in Figures 6 and 7, it can be speculated that the sliding test with Cu-MgPG sample against SUJ2 counterbody can reduce the average friction coefficient while reducing wear.

Conclusions
In this study, a Cu-MgPG material was prepared, and the experimental results proved that the Cu-MgPG material in this study has self-lubricating properties. In addition, under fixed experimental conditions (fixed load 10 N and sliding speed 0.1 m/s), the optimum tribopair combination of Cu and Cu-MgPG with the counterbody material, the influence of the counterbody materials in terms of coefficient of friction and wear was examined, and the following findings were obtained.
(1) The friction coefficient detected with the tribopairs of Cu-MgPG composite material is significantly lower than that of the pure copper material since the added graphite acts as a solid lubricant.
(2) The Cu-MgPG/SUJ2 pair was found to decrease both the friction coefficient and specific wear amount compared to the Cu/SUJ2 pair. From the results of XPS and EDS analysis of the sliding surfaces, it is considered that the added MgPG acted as a solid lubricant and suppressed the oxidation behavior of the material. On the other hand, the Cu-MgPG/Al 2 O 3 pair was found to have the lowest coefficient of friction and specific wear of all of the pairs.
In general, the results of this study have demonstrated the application potential of Cu/MgPG materials under specific circumstances. However, in the case of practical application, there are changes in loads and sliding speeds, so the materials prepared in this study cannot be immediately put into the application of brush or bearing materials. In future research, it is necessary to continue studying the effects of sliding loads and sliding speeds on the coefficient of friction and wear with different tribopairs.

Data Availability Statement:
The data presented in this study are available in this article.