To explore the potential of directly grown multi-layer graphene as an agent in reducing friction and wear of steel on steel tribo-pair, multi-layer graphene films were synthesized on GCr15 steel in a low-pressure chemical vapor deposition (LPCVD) setup using a gaseous mixture of acetylene and hydrogen onto a bearing steel substrate. An interlayer of electroplated nickel was deposited on steel to assist and accelerate the graphene deposition. The tribological performance was evaluated using a ball-on-disc tribometer with an average Hertzian pressure of 0.2, 0.28, 0.34, and 0.42 GPa over a stroke length of 5 mm against GCr15 steel ball and compared with bare steel and nickel-plated steel. The results indicate that the friction coefficient is dependent on the applied load and decrease with increasing load, and the minimum friction coefficient of ∼0.13 was obtained for an applied normal load of 1 N; however, the coating failed after 250 cycles. The decrease in friction coefficient has been attributed to the homogenization of the deposited multi-layer graphene along the sliding direction and transfer of graphene to counter-face ball leading to inhibition of metal-metal contact. The investigation suggests that this kind of coating has the potential of improving the tribological performance of metal-metal tribo-pairs.
Graphene is the hexagonal arrangement of sp2 carbon atoms forming a honeycomb shape  and is increasingly being explored as a solid lubricant to improve the tribological performance of mechanical components leading to energy saving and prolongment of their life [2–5]. The chemical vapor deposition (CVD) growth of graphene on transition metals has been reported to be the most preferred method to synthesize high-quality graphene over a large area [6,7]. Based on operating pressure, the different versions for CVD synthesis of graphene have been proposed, such as atmospheric-pressure chemical vapor deposition (APCVD), low-pressure chemical vapor deposition (LPCVD), and ultrahigh vacuum chemical vapor deposition (UHV-CVD) [8–10]. For a compromise between quality and cost, the LPCVD method has been reported as the most promising method of graphene deposition. LPCVD graphene coatings typically exhibit excellent uniformity, high purity, and good coverage over the substrate. At the same time, low pressure during CVD decreases the influence of residual gas (O, C, H) by avoiding unwanted gas-phase reactions to contaminate the coatings. Based on transition metal, the surface-growth and segregation-growth are two renowned CVD mechanisms to favor graphene growth . The former describes the physical deposition of carbon after dehydrogenation to form graphene, whereas the later one involves the synthesis of graphene by the dissolution of carbon atoms in the metal at high temperature followed by segregation during cooling. Nickel and copper have been reported to be the popular transition metals for graphene growth . Copper follows the surface-growth mechanism and fits best for monolayer graphene growth . However, nickel supports the segregation-growth mechanism and suits well for few to multi-layer growth of graphene [14,15]. Owing to higher carbon solubility and diffusivity, nickel, as a transition metal, allows a faster synthesis of graphene than copper . Methane (CH4), ethylene (C2H4), and acetylene (C2H2) are the most recognized gaseous carbon sources for graphene deposition. Out of these, acetylene offers a low pyrolysis temperature, shortens the reaction time, and promotes the graphene growth due to its higher surface adsorption energy .
The coating of graphene on steel has shown promising results in different applications because of its unique physical and mechanical properties. Pu et al.  and Stoot et al.  reported that the graphene-covered stainless steel exhibits outstanding corrosion resistance, and it can considerably increase the lifetime of future-generation bipolar plates for fuel cells. Currently, the tribological potential of graphene-based coatings has attracted much attention of scientists and engineers working in the area of tribology. Several studies have reported the improvement in friction and wear behavior of CVD grown graphene in a variety of experimental combinations. Different types of graphene coatings (such as monolayer, few-layer, multi-layer, and reduced graphene oxide coatings) have been reported to facilitate friction and wear reduction . Berman et al.  concluded that the SPG-coatings over steel reduced the friction and wear significantly by forming a conformal protecting layer on sliding interfaces. Restuccia and Righi  revealed that graphene can efficiently lubricate the steel-on-steel microscale contacts in dry as well as in the humid atmosphere and attributed the improvement in tribological properties to passivating effect and intense tribo-chemical action of graphene with native irons. Kim et al.  deposited the graphene oxide nano-sheet (GONS) coatings on silicon substrates via electrodynamic spraying process and reported a reduction in friction coefficient to a value of ∼0.1 attributing it to transfer of GONS coatings on to counter surface. Romani et al.  reported direct growth of graphene on API X80 steel using ethanol as a precursor at different partial pressures and recorded about three times lower friction coefficient for steel partially covered with graphene than bare steel. Hong et al.  deposited graphene on silicon wafer by electrophoretic deposition under different combinations of applied voltage and deposition time and reported over 80% reduction in friction coefficient. The graphene coatings with more compacted microstructure were reported to be more effective and durable as a solid lubricant.
According to previous research findings, the use of graphene as a solid lubricant is limited, partially by the continuous growth of graphene films over a large area, but more likely by weak adhesion of transferred graphene films to an arbitrary substrate. The present study is in continuation of our earlier published work  in which the multi-layer graphene was grown on nickel catalyzed steel substrate, and the optimum conditions of growth were determined. In the same work, the friction and wear behavior of graphene coating under unidirectional (rotary) sliding also reported by conducting tests at a fixed load of 0.5 N and constant sliding speed of 0.07 m/s. The current investigation is a detailed study on the friction and wear characteristics of multi-layer graphene, deposited on nickel-plated steel by chemical vapor deposition using the optimized growth conditions from our previous work , under reciprocating sliding motion and is aimed at exploring the tribological performance of this coating by carrying out friction tests at different normal loads of 0.1, 0.3, 0.5, and 1 N and a frequency of 1 Hz.
2 Materials and Methods
2.1 Synthesis and Characterization of Graphene Films.
The circular disc-shaped (ɸ 25 mm and thickness 7 mm) GCr15 steel samples (hardness ≥ 60 HRC) were prepared by grinding with different grades of emery papers (600#, 1200#, 2400#, and 4000#, respectively) followed by cloth polishing with diamond paste to obtain a smooth and bright surface (surface roughness Rq ≤ 20 nm). The specimens were cleaned in an ultrasonic bath with acetone and ethanol for about 15 min each to remove the surface impurities, dust, oil, and organic contaminants. A 3D profilometer (Zygo NexView, AMETEK Inc., CT) was used to confirm the surface finish. Prior to deposition, a 25 µm thick nickel layer was electroplated over the steel surface. The surface roughness (Rq) of the disc surface was ∼35 nm after nickel electroplating. Table 1 lists the results of surface composition analyzed by energy-dispersive spectroscopy (EDS) for steel samples before and after nickel electroplating. The graphene films were synthesized by a thermal CVD method with acetylene as a hydrocarbon source and hydrogen gas as a reduction medium. Figure 1(a) presents a schematic diagram of CVD setup consisting of a gas supply unit, a split tube furnace, and a vacuum system. The heating–soaking–cooling cycle followed to synthesize graphene is illustrated in Fig. 1(b). After placing the nickel-coated steel sample at the center of the split tube furnace, the quartz tube was pumped down to obtain a vacuum of order 1.2 × 10−2 Torr (base pressure, Pbase) using a rotary vacuum pump. The temperature was raised to 850 °C from room temperature over a time span of 30 min. The steel sample was annealed for the next 40 min in a hydrogen (20 sccm) rich environment. Hydrogen performs the cleaning of nickel surface and maintains sites for dehydrogenation of hydrocarbon, yielding diffusion, and carbon segregation . The hydrocarbon source (acetylene) was flown into the quartz tube with a flow rate of 6 sccm, and the total pressure (Pbase+ PH2 + PC2H2) of the tube was around 3.20 × 10−1 Torr for a period of 10 min to allow the desired reaction to take place. Acetylene gas got decomposed into carbon and hydrogen atoms at a higher temperature, and carbon atoms were absorbed by the sample surface during the reaction stage. The flow of acetylene was terminated after completing the reaction stage, while the flow of hydrogen gas was continued till the cooling of the furnace to room temperature. The synthesis of graphene occurred due to the segregation of carbon atoms over the surface during cooling.
The synthesized graphene samples were characterized by optical microscopy, scanning electron microscopy (SEM), Raman spectroscopy, and X-ray photoelectron spectroscopy (XPS). The morphology and structural properties of as-deposited graphene were studied through optical microscopy (VHX-600K, Keyence, Japan) and field emission gun scanning electron microscopy (FEG-SEM). Raman spectroscopy was performed using a high-resolution Raman spectrometer (LabRAM HR Evolution, Horiba Jobin Yvon, Japan) with a laser excitation wavelength of 532 nm, laser power of 13 mW, scan range of 800–3000 cm−1, and scanning time of 0.1 s per spot. The sample was focused using a 50× objective lens, and the size of the focal spot of the laser was 1.25 µm in diameter.
The XPS analyses were performed using Quantera SXM (ULVAC-PHI, Kanagawa, Japan) with a monochromatic Al Kα X-rays source of energy 1486.6 eV. The applied power and beam diameter of X-rays were 23.56 W and 100.0 µm, respectively. The angle of incidence was set at 45 deg for X-rays. To examine the elemental distribution versus depth, etching was performed by a sputtering argon gun. Peak fitting allowing to decompose the XPS spectra in different components assigned to different surface species was performed using the CasaXPS processing software with a Shirley background. A high-resolution transmission electron microscope (HRTEM, Tecnai G2 20 TWIN, FEI, CT) was utilized to measure the thickness and to reveal the structure of as-deposited graphene.
2.2 Tribological Behavior of Graphene Films.
The tribological behavior of synthesized graphene films was investigated by using a ball-on-disc type of tribometer (UMT-5, Bruker, Karlsruhe, Germany) against the GCr15 steel ball of 6 mm diameter. The disc was set to reciprocate against the stationary counterpart (ball) with a frequency of 1 Hz (sliding speed of 0.01 m/s) over a stroke length of 5 mm. The applied normal loads of 0.1, 0.3, 0.5, and 1 N were used for friction tests, corresponding to the average Hertzian contact pressures of 0.2, 0.28, 0.34, and 0.42 GPa, respectively. To ensure the repeatability of friction and wear results, the sliding tests were repeated at least three times for every condition. The morphologies of the worn tracks on the disc and wear scar on the counterpart (ball) were systemically analyzed by Raman spectroscopy, optical microscopy, scanning electron microscopy (SEM) equipped with an energy dispersive spectroscopy (EDS), and transmission electron microscopy.
3 Results and Discussion
Figure 2 depicts the micrograph of the multi-layer graphene-coated steel surface, as examined under SEM. Graphene could be observed to be present in the form of patches on nickel-plated steel. The growth of graphene on steel without a specific catalyst is quite challenging , and hence, we opted electroplating of a nickel layer over the surface, which accelerates the synthesis process. Graphene deposition follows the mechanism of precipitation and growth. According to this, the carbon from acetylene gets dissolved into nickel to form a Ni–C solid solution at the higher temperature during the reaction stage. At the time of cooling, the solubility of carbon in nickel decreases, and the carbon atoms precipitate out from the Ni–C solid solution during cooling and coalesce to form graphene film .
Since Raman spectroscopy has been proven as one of the most widely used techniques for nondestructive characterization of carbon-based material, we selected it for the initial characterization of synthesized graphene. Raman spectrum presented in Fig. 3 clearly indicates the presence of three characteristics peaks identified as D, G, and 2D peaks centered around ∼1350 cm−1, ∼1580 cm−1, and ∼2700 cm−1, respectively [28,29]. The presence of a single and sharp 2D peak in the Raman spectrum shown in Fig. 3 clearly indicates that the deposited film is of graphene rather than graphite, which has two components in 2D peak, as reported earlier . As evidenced in the figure, the G band is higher than 2D band, and I2D/IG approaches a value of ∼0.64, which suggests that the steel surface was covered by multi-layer graphene. There is one more weak peak at ∼2460 cm−1 (G*). The value of the ratio of ID/IG is 0.07, which indicates the growth of the high quality of multi-layer graphene due to very low-intensity D-peak. The surface roughness of multi-layer graphene-coated steel as measured by a noncontact 3D optical profilometer is found to be ∼56 nm (Rq).
The transmission electron microscopy was also used to characterize the synthesized multi-layer graphene, and TEM images are presented in Fig. 4. FIB in situ lift-out technique was used to prepare the cross-sectional samples for TEM. Metallic layers of Cr and Pt were deposited on the surface to prevent the damage of multi-layer graphene before the FIB process. The cross-sectional TEM of the synthesized multi-layer graphene shown in Fig. 4(a) indicates that the thickness of multi-layer graphene is about 21 ± 2 nm. Another high-resolution TEM image in Fig. 4(b) is presented to identify the layered structure of graphene films.
Elemental composition and binding energy of multi-layer graphene were determined by X-ray photoelectron spectroscopy. Figure 5(a) shows a broad survey XPS spectrum of as-deposited multi-layer graphene with an intense peak at 284.6 eV corresponding to C1s (∼92 at%). Two other low-intensity peaks were also observed at 532 eV and 854 eV representing O1s (∼7.4 at%) and Ni2p (∼0.6 at%), respectively. The O1s peak refers to molecular oxygen (O2) that originated from the air. Further, a de-convolution of the C1s spectrum of multi-layer graphene into a main peak at 284.6 eV and three other peaks centered at 285.9, 287.7, and 289 eV, respectively, was performed and presented in Fig. 5(b). Peak fitting was done using Voigt approximation (Gaussian–Lorentzian). The peak at 284.6 eV has been assigned to C=C and stands for sp2-hybridized graphite-like carbon atoms. The peak at 285.9 eV represents the presence of the sp3-hybridized amorphous carbon atoms. The remaining two peaks at 287.7 eV and 289 eV may be attributed to C=O (ketone, aldehyde) and O–C=O (ester, acid, carboxylic) groups, respectively [30,31]. The binding energy of each peak and the quantitative evaluation of the fitted curve are presented in Table 2. Overall, the results confirm that the multi-layer graphene films were synthesized successfully using low-pressure CVD with few-oxygen containing functional groups on the surface of alloy steel.
The other aspect of this work is to study the tribological performance of multi-layer graphene-coated bearing steel under reciprocating sliding motion. For this, a set of experiments with different sliding pairs, namely, steel versus steel (as a reference pair), steel versus nickel-coated steel and steel versus multi-layer graphene-coated steel, were performed in the air environment under a normal load of 0.5 N. The friction coefficient was measured continuously during the sliding test for 600 cycles. The results for the coefficient of friction are plotted against wear cycles in Fig. 6. One can observe that the coefficient of friction for bare steel was initially low and then increased with wear cycles after the removal of the oxide layer presented on the bare steel surface. Severe fluctuations recorded throughout the friction test may be attributed to the accumulation of wear debris in between the contact surfaces, which might have initiated three-body abrasion. The fluctuation was abated to some extent by a nickel-plating over the steel surface, as obvious from the figure. The friction coefficient was also reduced slightly and oscillated at around 0.59. For the multi-layer graphene-coated steel, the friction coefficient was reduced significantly and followed a very smooth curve all over the friction test. The friction of multi-layer graphene-coated steel is about five times lower than that measured for bare steel, which may be attributed to the lubricious effect of graphene .
To examine the durability of multi-layer graphene coatings under different loads, the tribological behavior of multi-layer graphene-coated steel was also investigated for different normal loads of 0.1, 0.3, 0.5, and 1 N. Figure 7 presents the variation of friction coefficient recorded throughout the friction tests for varying loads. The friction coefficient is found to decrease with increasing load. Under a normal load of 0.1 N, the average value of the friction coefficient was recorded as ∼0.19. As the load increased to 0.3 N and 0.5 N, the friction coefficient reduced to ∼0.17 and ∼0.15, respectively. The minimum average friction coefficient of 0.13 was recorded for 1 N normal load, but the graphene film could only sustain this value up to 250 cycles, and the friction coefficient was found to shoot up sharply to a value of 0.9 beyond that as indicated in Fig. 7(a) before attaining a value of ∼0.6 for the remaining test. The sudden increase in friction coefficient after 250 cycles strongly suggested the damage of graphene coatings. The average values of friction coefficient throughout the friction test for bare steel, nickel-plated steel, and multi-layer graphene-coated steel was also plotted as a function of normal load, as presented in Fig. 7(b). The average values of the friction coefficient for graphene-coated steel before and after coating failure are also presented in Fig. 7(b).
The worn surfaces of disc, as well as counterpart ball, were analyzed using optical microscopy, SEM, and EDS to explore the mechanisms of wear for different tribo-pairs. Figures 8(a)–8(c) show the worn surfaces of steel, nickel-plated steel, and multi-layer graphene-coated steel after sliding under a load of 0.5 N, whereas Fig. 8(d) presents the surface of multi-layer graphene-coated steel worn under a load of 1 N. The width of wear track for multi-layer graphene-coated steel is significantly less than other. The multi-layer graphene coating over steel provides the protection to the underlying substrate by avoiding the direct metal-metal contact and utilizes the property of low interlayer shearing strength of graphene in affecting a reduction in friction. It can be noted that when the applied normal load is 1 N, the graphene coating was damaged significantly, and the material beneath is exposed to the counterpart, which causes a sudden increase in friction.
The SEM micrographs of worn-out as-received steel, nickel-plated steel, and multi-layer graphene-coated steel specimens at a load of 0.5 N are presented in Fig. 9 along with the corresponding energy-dispersive spectrums of the whole micrographs. The figure also contains the elemental analysis of the entire surface as an inset in the EDS spectrum. Figure 9(a) shows fine scratches along the sliding direction, which appear to be covered by a layer of oxide probably of iron, as evidenced by the EDS spectrum along with some wear debris particles. It also presents some places from where the layer might have been detached, leading to the creation of voids and micro-cracks. Figure 9(b) clearly suggests a substantial difference in the morphology of nickel-plated steel with more obvious adhesion on the worn surface. The surface is covered by a relatively thick and well compacted glazed layer of nickel oxide, as indicated by the EDS spectrum. The glazed layer has been shown to be effective in reducing both friction and wear . However, this layer appears to have been delaminated at few places leading to exposure of the underlying substrate pointing toward adhesion and delamination wear. For multi-layer graphene-coated steel, the graphene, as a solid lubricant, might have reduced the plastic deformation as well as adhesion wear, and the abrasive wear appears to have dominated as evidenced from relatively deep tracks in Fig. 9(c). The elemental analysis shown in Fig. 9(c) suggests a significant increase in carbon content for multi-layer graphene-coated steel, which appears in the form of a compact layer over the surface, which gives protection to the underlying metal and inhibits the direct metal–metal contact leading to a reduction in friction as shown in Fig. 7. This increase in carbon may have come from the remaining fragments of multi-layer graphene after the friction tests, which was further supported by Raman spectroscopy and transmission electron microscopy.
The SEM images of the counterparts (balls) after the friction test at a load of 0.5 N are presented in Figs. 10(a)–10(c), whereas Fig. 10(d) shows the worn surface of the ball after testing at 1 N. It could be observed that wear scar is reduced significantly for multi-layer graphene-coated steel for friction test at a normal load of 0.5 N in comparison with both bare and nickel-coated pair with no obvious loss of ball material. However, when the load reaches to a value of 1 N, an increase in loss of ball material was observed due to metal–metal contact brought about by the failure of the coating after 250 cycles. Table 3 gives the chemical composition of different areas of the wear scar as obtained through EDS. The area named ball represents the chemical composition of the original ball. An increase in carbon percentage for the steel ball sliding against multi-layer graphene-coated steel with minimum oxidation could be observed, indicating the transfer of carbon (originated from multi-layer graphene) to the ball surface during sliding.
The surface protection of the disc by graphene layers and the transfer of the layer on the ball during sliding is confirmed by Raman spectroscopy (Fig. 11) of the wear track as well as of wear scar. It can be observed that there is a sharp D-peak for wear scar as well as for wear track, which indicates the increase in defects of synthesized multi-layer graphene films. This increase in defects points toward the damaging and disordering of graphene during sliding . In brief, the Raman spectrum confirms the presence of the graphene layer with a slight increase in defects.
Figure 12(b) displays the wear volume of steel ball against steel, nickel-plated steel, and multi-layer graphene-coated steel, and it is evident from Fig. 12(b) that wear of steel ball sliding against steel and nickel-plated steel becomes more and more severe with an increase in applied load. On the other hand, the wear of steel ball sliding against multi-layer graphene-coated steel decreases comparatively. The increase in normal load from 0.1 N to 0.5 N led to a decrease in wear scar diameter and, subsequently, the wear volume for multi-layer graphene-coated steel. This reduction in wear volume may be attributed to the homogenization and smoothening of graphene coatings during the sliding with the increase in load . However, for a load of 1 N, the wear of the steel ball increased due to failure of multi-layer graphene coating after ∼250 cycles, and for the remaining cycles, the severity of wear increased due to a direct metal to metal contact. Undoubtedly, it can be clearly observed that the wear for multi-layer graphene-coated steel up to the load of 0.5 N, is much lower than that of bare steel and nickel-plated steel, alike to the trends observed for the coefficient of friction.
Further, TEM was used to characterize the wear track and wear scar. Cross-sectional TEM samples were prepared using a similar procedure, as mentioned before for TEM of as-deposited multi-layer graphene. TEM images of wear scar shown in Figs. 13(a) and 13(b) reveal the presence of a transferred tribo-layer, which consists of fragments of graphene and amorphous carbon. The presence of a tribo-layer confirmed that the surface was well protected. The analysis of wear track reveals that the thickness of multi-layer graphene was reduced to 10 ± 2 nm and consists mainly of the fragments of graphene, as evident from Figs. 13(c) and 13(d).
The present study was conducted for the evaluation of the tribological potential of directly grown multi-layer graphene coating on GCr15 steel in rubbing against GCr15 steel ball under reciprocating motion under different loads. The multi-layer graphene was successfully grown via the CVD method on GCr15 steel by incorporating an interlayer of electroplated nickel. Multi-layer graphene films on steel enhance the tribological performance significantly, and a transfer layer consisting of fragments of graphene sheets and amorphous carbon was observed over the counter surface (ball). The results indicate that the friction coefficient is dependent on the applied load and decreases with increasing load. The minimum friction coefficient of ∼0.13 was recorded for an applied normal load of 1 N; however, the coating failed after 250 cycles. The decrease in friction coefficient has been attributed to the homogenization of the deposited multi-layer graphene along the sliding direction and transfer of graphene films to counter-face ball leading to inhibition of metal–metal contact. The results indicate that this kind of coating has excellent potential for improving the tribological performance of metal–metal tribo-pairs. However, more studies need to be conducted to evaluate the tribological behavior of these coatings under different speeds and environments.
The authors are thankful to the National Natural Science Foundation of China (Grant No. 51975314) for financial support.
Data Availability Statement
The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request. The authors attest that all data for this study are included in the paper. Data provided by a third party listed in Acknowledgments. No data, models, or code were generated or used for this paper.