Tribological Properties of a Self-Repairing Material: Functionalized Graphene/montmorillonite Nanosheets Lubricant Additives


 The functionalized graphene/montmorillonite (FG/MTT) nanosheets were synthesized through chemically bonding by a simple, green method, which has remarkable dispersion stability in oil and its lubricating performance was evaluated by a four-ball tribometer. The test results show that FG/MTT has a preeminent lubricant property when the concentration is 0.4 mg/ml. Compared with the bare oil sample, its average friction coefficient (FC) and wear scar diameter (WSD) decrease by 50.4 % and 13.2 %, respectively. The synergistic effect between FG and MTT was further explored by comparing the lubricant mechanism of the different additives. After synthetically analyzing worn surface by means of scanning electron microscopy and X-ray photoelectron spectroscopy, the lubrication mechanism of the FG/MTT nanocomposite as oil additive is discussed and postulated: The FG/MTT with weak interlayer adhesion is filled between the friction pairs to avoid contact and clinging of some asperities, and the sliding between the layers plays a role in lubrication. Furthermore, FG/MTT will react with the surface of the friction pair to form a repair layer composed of Fe2O3, SiC, SiO2, and aluminosilicate, mending the grinding surface and promoting the hardness after friction.


Introduction
Lubricating oil is widely used in various elds as a lubricant that reduces mechanical loss and improves work e ciency. It consists of base oil and additives. The properties of additives have an important in uence on the lubricating performance of lubricating oil. To meet the requirements for lubricating performance in different elds, many lubricating oils containing diverse additives have been produced.
Common lubricant additives include the following categories: soft metal nanomaterials (nano copper, nickel, etc.), borate materials (lanthanum borate), carbon nanomaterials (carbon nanospheres), and so on. [1,2] However, as the performance requirements become multifarious, it is di cult for a single additive to ful ll the requirements. Therefore, the research on composite lubricant additives is becoming crucial. Simultaneously, in addition to ensuring that additives have good lubricating properties, the dispersibility of additives in the base oil is also one of the indicators for evaluating lubricants. Lubricating oil additives with good dispersion stability have better lubricating properties.
Graphene Oxide (GO) is a new type of layered carbon nanomaterial with macroscopic dimensions, which has a nanometer-level thickness, but can be stretched inde nitely on the same plane. Multilayer graphene is prone to sliding between layers, so it has good lubricating properties. Mao et al. [3] analyzed the lubricating properties of three different morphologies of reduced graphene oxide, and the result showed that the morphology of the graphene sheets directly affects its lubrication performance. The lubricity of graphene is not only affected by structural properties such as surface defects and stacking methods but also related the presence of some hydrophilic groups such as hydroxyl and carboxyl groups on the surface, which may cause agglomeration and sedimentation when dispersed in oil. To ameliorate dispersion, domestic and foreign studies have also conducted further investigations on this issue. In addition to being a single-component additive, graphene is compounded with other additives and added to the base oil to further improve the performance of the lubricant. Meng et al. [4] successfully loaded nano-gold particles on multi-graphene oxide to prepare nano-Sc-Au/GO composites with supercritical carbon dioxide, which has good tribological properties.
As a natural material with abundant reserves and low cost, silicate minerals have been widely developed and used in various elds. In recent years, the research on silicate materials used as lubricant additives has also exploited rapidly. Wang et al. [5] added serpentine nanoparticles as lubricant additives to base oil, and the friction coe cients and wear spot diameters were reduced by 22.8 % and 34.2 %, respectively.
Montmorillonite (MTT) is a layered mineral that is composed of a nanometer-thick negative silicate layer on the surface and is accumulated together by the electrostatic action between layers. Its main component is (Al, Mg) 2 (SiO 10 )(OH) 2 ·nH 2 O. The crystal cell in its crystal structure comprises two layers of silicon-oxygen tetrahedron sandwiched with a layer of aluminum oxygen octahedron. Some studies have shown that montmorillonite as a lubricating additive has the excellent self-repairing ability for the surface of friction pair. Still, the repairing effect only will be vaild after a long time of friction and wear test [6,7].
As a natural silicate mineral, even after surface modi cation, it is still arduous to stably disperse in the oil for a long time. Hence, it has a crucial impact on the tribological properties.
Herein, functionalized graphene oxide (FGO) and its composites with silicates were synthesized, and their tribological properties and abrasion self-repairing performance were systematically investigated. Through simple hydrothermal compounding, the MTT sheets were successfully intercalated and reduced in the surface of FGO to obtain FG/MTT materials that can be stably dispersed in lubricating oil. The prepared nanocomposite of FG/MTT was used as lubricant additive in 15w40 engine oil, and its lubricating performances were evaluated by a four-ball tribometer. After the synergistic effect of the two materials, the material, as a lubricant additive, not only has excellent antifriction, antiwear, and self-healing ability but also can form a self-repairing layer which improves the surface hardness of the friction pair.

Materials
Montmorillonite was purchased from Zhejiang Fenghong Chemical Co., Ltd (Zhejiang, China). Silane coupling agent (KH550, chemically pure) was purchased from Shanghai Alading Biochemical Technology Co., Ltd. Ethanol (analytically pure) was purchased from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). The GO was synthesized according to a modi ed Hummers method in our laboratory as previous report [8].

Synthesize of the Functionalized graphene oxide
The GO was modi ed with a silane coupling agent (KH550) to get the functionalized graphene oxide (FGO). Firstly, 0.025 wt % GO suspension was prepared in 160 g water under ultrasonication, then 6 mL KH 550 was added slowly in the GO suspension. The mixture was kept stirring for 60 min at room temperature, and was transferred into the ask and stirred for 2 h at 80 ℃ to get the FGO. After centrifugation, the solid was kept in the Falcon centrifuge tube and washed in deionized water three times.
2.3 Synthesize of the functionalized graphene/montmorillonite 1.6 g montmorillonite was ground by wet grinding of 25 ml ethanol in mortar for 30 min and was dispersed into 50 ml water by ultrasonication for 2 h. The montmorillonite suspension was added slowly into the FGO aqueous suspension and stirred for 2 h at room temperature. The mixture was transferred into hydrothermal reactor kettle for 2 h at 120 ℃ to get the functionalized graphene/montmorillonite (FG/MMT). The solid was kept in the Falcon centrifuge tube and washed in ethanol three times by centrifuge to get the wet sample of FG/MMT.

Characterization of FG/MMT
Morphology analyses of samples were performed on a transmission electron microscope (TEM, JEOL JEM-2100) and the eld emission scanning electron microscopy (SEM, Quant 250 FEG) with energydispersive spectrum (EDS). For the TEM observation, the samples were rstly dispersed in ethanol by ultrasonic treatment and then dropped on carbon copper grids. Fourier transform infrared spectroscopy (FTIR) was recorded on Nicolet IS10 (ThermoFisher) with Ever G10 optical source. The powder X-ray diffraction pattern (XRD) analyses were carried out on a Bruker D8 Advance diffractometer with a Cu Kα radiation (λ = 1.5406 Å), whose scanning angle ranged from 5° to 80° of 2θ. Additionally, X-ray photoelectron spectroscopy (XPS) measurements were performed on Thermo Escalab 250 (U.K.) at the monochromatic Al Kα (150 W, 500 μm and 1486.6 eV) radiation.

Tribological Test
The tribological properties were performed on an MRS-10G four-ball tribometer (Jinan Yongce Industrial Equipment Co., Ltd., China.) under 197 N with a rotary velocity of 600 rpm at room temperature for 3600 s. The friction pair consists of four identical balls, one rotating on the top while the other three xed beneath. The balls were composed of GCr15 bearing steel with a diameter of 12.7 mm (Shanghai Steel Ball Plant CO., Ltd., China.). All test-section components were cleaned ultrasonically in petroleum ether for 3 min and dried in the air before tests. The friction coe cient was recorded automatically by a strain sensor, and the wear scar diameter was measured by a 15J optical microscopy (Shanghai Optical Instruments Sixth Factory Co., Ltd., China). Each scar was surveyed at least ve times to guarantee the standard deviations less than 5%.

Analysis of Worn Steel Surfaces
The surface morphology and composition of the steel balls after the tribological test were analyzed with SEM (Quant 250 FEG) with energy-dispersive spectrum (EDS) and XPS (Thermo Escalab 250). The FTIR spectra of the samples in Figure 3(a) show that the typical peaks of GO appear at 3300, 1720, 1600, and 1050 cm −1 , corresponding to the presence of hydroxyl, carboxyl and epoxy groups, respectively.

Results And Discussion
Compared with GO, the disappearance of the peak at 1720 cm -1 and the new peaks at 1560 cm -1 (N-H stretching vibration) and 1460 cm -1 (C-N stretch of amide) of FGO indicate the formation of C-N-C bands in the FGO sample, intensely supporting the presence of the amide groups on the FGO, which means that the amino group on the silane coupling agent reacts with the carbonyl group on GO to form an amide bond [9,10]. The disappearance of the bands near 1150 cm -1 (symmetrical stretching vibration peak of the epoxy group) demonstrates the reaction of amino groups of KH550 with the epoxy group on GO. In the case of FGO, the bands near 1100 cm -1 correspond to the stretching vibration peaks of Si-O of Si-OH.
The asymmetric peaks at 2920 and 2850cm -1 are assigned to the C-H stretching vibrations of the alkyl group (-CH 2 , -CH 3 ), derived from the alkyl chain of the silane coupling agent KH550. It is proved that the KH550 molecules successfully bonded to the graphene oxide surface to form the functional graphene oxide. As the infrared spectrum of MMT exhibits, the peak at 3610 cm -1 corresponds to the stretching vibration peaks of -OH groups, and the peak near 980 cm -1 is assigned to the stretching vibration peaks of Si-O of Si-O-Si, which could serve as a hydrogen-bonding sites for the functional groups of FGO [11]. Compared with the infrared spectra of single FGO and MMT, the FG/MMT composite shows the blueshifts of asymmetric peaks at 2990 and 2900 cm -1 and the blue-shift of the stretching vibration peak of Si-O bond at 1010 cm -1 , revealing that the FG/MMT composites were successfully synthesized [12,13]. The TEM images of composites shown in Fig.1 also demonstrate this.
The XRD patterns of the GO, FGO, MMT, and FG/MMT are shown in Fig.3(b). There is a strong (001) diffraction peak at 2θ = 11.45°in the XRD pattern of GO, indicating that the layer spacing is 0.77nm. In the XRD pattern of FGO, the strong peak at 2θ = 11.45 ° almost disappears, and a relatively broad peak appears at 2θ =23.52 °, which corresponds to the characteristic peak of graphene (002) [14,15]. It shows that the carbon structure of FGO becomes more disordered due to the graft of silane coupling agent on it. The layer spacing of (001) can be calculated to be 1.296 nm based on the (001) diffraction peak of MMT at 2θ = 7.117 ° according to Bragg's law [16]: The typical diffraction peak of MMT also can be found in FG/MMT composite, and the (001) diffraction peak at 2θ = 5.79 ° indicates the layer spacing is 1.58 nm according to Eq. (1). It means that the MMT is successfully intercalated into FGO layers [13].
After graphene oxide was treated with a silane coupling agent, the functionalized graphene oxide (FGO) was produced. The grafting of KH550 with GO makes its crystal plane spacing increase. After MMT is introduced into FGO during the hydrothermal reaction, the MMT sheets are inserted between the graphene oxide sheets to make the crystal plane spacing larger. The enhancement and leftward shift of the diffraction peak of the MMT (001) crystal plane in the composite products also prove that the montmorillonite lamellae are intercalated successfully between graphene lamellae. From these results, it can be concluded that the FG/MMT composite gets a larger interlayer spacing after intercalation.
The Raman patterns of the GO, FGO, and FG/MMT are shown in Fig.3(c). The D band at around 1340 cm -1 shows the vibrations of disordered sp 3 carbon atoms; the G band at 1585 cm -1 re ects the vibrations of sp 2 carbon atoms in the hexagonal lattice of graphene. The ratio of the intensity of the D and G band (I D /I G ) is in direct proportion to the disordered degree of the graphene lattice [17]. The I D /I G of FGO (1.081) is higher than the I D /I G of GO (I D /I G = 0.986), indicating that the FGO is more disordered with more defects in the surface. The I D /I G of FG/MTT (0.920) is lower than that of FGO and GO, indicating that the surface defect of FG/MTT decreases due to the reduction of FGO and the MTT chemically bonded with KH550 to form FG/MTT composite during the hydrothermal process.

Friction and Wear Performance
The images of FG/MMT composite dispersed in the 15w40 engine oil after 30 and 120 days are presented in Fig.4. The dispersion stability of additives in lubricating oil is a considerable standard to evaluate the performance of additives [18,19], which affects the tribological properties of oil samples. After 30 days of sedimentation experiment, it is obvious that the FG/MMT composite can be welldispersed without sediments at 0.4 and 1.0 mg/ml. While after 120 days, the upper color of the two concentrations FG/MTT oil sample is similar, which is deeper than the 15w40 oil sample. However, there is a little sediment on the bottom of the bottle for the oil sample at 1.0 mg/ml, but can be re-mixed uniformly by a slight shake. It means the composite materials have excellent dispersion stability in oil. Moreover, the dispersion stability of the composite is more remarkable at a lower concentration of 0.4 mg/ml.
According to the conclusion of the sedimentation experiments, the oil samples of FG/MMT at 0.4mg/ml with good dispersion were prepared. For comparison, the bare oil and oil samples of MMT, GO, and FGO were also prepared. The tribological tests were carried out under the same conditions, and the results are shown in Fig.5. It presents the variation of friction coe cient (FC) of different oil samples overtime in the tribological test. It can be seen from Figure 5a that the FC values of several oil samples except for FG/MMT basically increase rst and then decrease gradually. This may be explained as follows: During the process of friction, some large particles are squeezed by the asperities and broken into tiny particles by the action of mechanical motion, resulting in the increased FC. With the increase of time, the small particles could uniformly ll the friction surface to form a lubricant layer, thereby decreasing the FC.
The FC curve of the bare 15w40 oil sample reached its peak at 890s, about 0.129, and dropped to 0.120 at the end of 3600 s. For the oil samples containing GO, FGO, MTT, and FG / MTT, the friction coe cient declines to 0.090, 0.095, 0.105, and 0.051 at 3600 s, respectively, indicating that each oil sample has a certain lubricating effect with antiwear and antifriction properties for each additive, due to the special two-dimensional layered structure of the above additives. The layers could make relative slides easily with each other, which has the effect of reducing friction. However, when only MTT or GO is added, the oscillation of FC is relatively immutable, but the lubrication performance of the base oil is slightly elevated.
Compared to bare oil sample, GO oil sample has the same initial FC and the analogous ascent trend as bare oil in the rst 600 s but shows the fast decrease rate of FC during the long-term test. It means GO shows some antiwear and antifriction properties as reported [20,21]. GO and FGO oil samples demonstrate a similar uctuation tendency of FC over time. Still, FGO presents a lower FC over the time range of the experiment, which indicates FGO oil exhibits more advantageous antiwear and antifriction properties than GO oil. It might be due to the better dispersion in the presence of KH550 modi cation. In the case of MMT oil sample, its long-term tribological property surpasses that of bare oil but worse than the GO and FGO oil samples.
As for FG/MMT oil sample, it shows the lowest initial FC value; moreover, the FC continues to decrease from the beginning and end to the lowest value of 0.051 at 3600s compared to the rest samples. Therefore, the FG/MTT oil sample shows the best property compared to GO, FGO, MMT, and bare oil samples. Moreover, the excellent tribological performance of FG/MTT re ects the different mechanism from the test, which will be investigated further. Wherein the average friction coe cient of FG/MTT is 50.4 % lower than that of 15w40 base oil, which proclaims the intense enhancement of the lubricating properties. It was con rmed that the prepared FG / MTT as the additive material is excellent in friction reduction.
The wear scar diameter (WSD) is also one of the common indexes to evaluate the lubrication performance [22]. The WSD of the steel ball after the four-ball tribometer test was measured, and the average WSD of the steel balls in different oil samples was calculated. The result is shown in Fig.5(b). When no additives are added, the average WSD of the steel ball tested in the 15w40 oil sample is 0.333mm, and after adding different additives, the WSD becomes smaller, indicating that the addition of different additives promotes the 15w40 oil lubrication properties. Moreover, the average WSD of FG/MTT oil sample is at least 0.289 mm, which is 13.2 % lower than that of base oil sample, proving FC/MTT as an additive with the eximious ability to reduce friction.
The FC and WSD for the FG/MMT oil sample with different concentration (0, 0.2, 0.4, 0.8, 1.0 and 1.6 mg/mL) tested at 197 N and 600 rpm for 3600s are shown in Fig.5(c). The change range of the FC can be seen from the gure. The uctuation range of the 15w40 engine oil is the largest, and that of the FG/MMT oil sample is relatively smaller, which reveals that the composite material has preeminent antifriction ability. When the concentration of FG/MTT is 0.2 mg/mL, the additive can effectively reduce the FC, but the anti-friction and anti-wear ability is not as good as that of the 0.4 mg/mL sample. The reason is the de ciency of material in the oil sample, and it cannot be evenly covered on the wear surface during the friction and wear process so that there are vacancies in some places. The curves of FC and WSD with concentration are concave, and both of them reach the lowest value at 0.4 mg/ml. The FC and WSD reduce 50.4 % and 13.2 % compared to the base oil, respectively. The FC and WSD increase gradually with concentration rising to 0.8~1.6 mg/mL, which declares that the higher concentration of the composite hurts improving the lubricating performance. It would be resulted in the agglomerate of the composite at a high concentration.
FG/MTT composite not only has excellent lubrication performance of antifriction and antiwear but also contains oil-friendly functional groups to promote dispersion. Through the bonding and intercalation of two functional materials, the agglomeration between the composite materials can be effectively reduced, and the composite materials can be more stable and evenly dispersed in the lubricating oil for a long time. Fig.6 shows the SEM images of the worn surface of friction pair in different oil samples after tested for 1 h. With comparison, the WSD of the 15w40 oil samples (Fig.6a and b) is the largest, and the wear surface is rough with a small amount of wear traces. In contrast, the wear scars formed in the FGO oil sample ( Fig.6c and d), MTT oil sample ( Fig.6e and f), FG/MTT oil sample ( Fig.6g and h) become smaller in sequence. Although the WSD of the MTT oil sample is reduced compared with that of the 15w40 oil sample, there are more apperant scratches on the friction surface of MTT oil sample. This is probably because the particles of MTT material are relatively large and stiff, which damage the friction surface in the initial wear process. Both the wear sufaces in FGO and MTT oil samples still show scratches. However, it is worthy to note that the surface of wear scar in FG/MTT oil sample becomes much smoother, even no evident scratches observed in Fig.6 (g and h). It indicates that FG/MTT exhibits good antiwear property and self-repairing function, which can also be proved by further lubrication properties. As carefully observed, the grooves and valleys of this worn surface seem to be lled with the uncharacterized materials ( Fig.6g and h).
During the sliding process, all the additives including FGO, MTT, and the synthesized FG/MTT composites are easily deposited on the contact pair surfaces and thus form a protective lm, which can smooth the surfaces and reduce friction and wear effectively. Compared to FGO and MTT additive,the self-repairing performance of FG/MTT is the best Fig.7(a-d) are the EDS diagrams of the selected areas in Fig.6(b, d, f, h) of the wear scars in the FGO, MTT, and FG/MTT oil samples, respectively. Compared with the bare oil sample, the FGO sample exhibits the increase in the contents of carbon and oxygen and the presence of silicon from the anchored KH550 on GO, which means a friction lm formed from FGO additive. The wear surface of the MTT sample is rich in Page 9/22 C, O, Na, Si, and other elements. The Na element must come from the MTT additive, replying MTT additive also has a self-repairing effect. The content of C, Na, and Si elements on the worn surface of FG/MTT oil sample is signi cantly higher than those for FGO and MTT oil samples, which proves the presence of a self-repairing layer and its repair capacity is far greater than that of a single material. Fig.8 shows the micro indentation hardness test results of the steel ball surfaces and wear spots for bare oil, FGO, MTT and, FG/MTT oil samples. Compared to the hardness results in the bare 15w40 oil, It can be clearly seen that MTT additive almost hardly improves the hardness of the worn surface, although it can reduce the FC value. However, both FGO and GO additives can make the worn surface harder than the bare steel ball surfaces with 10.04 % and 6.91 % higher value of HV0.5. Most importantly, for FG/MTT composite oil sample, the hardness of the worn surface is 22.4 % higher than that of the steel ball surface. It re ects that under the action of FG/MTT composite additive, the worn scar in the process of friction and wear test becomes harder due to a repairing "coating" lm generated from FG/MTT, whose hardness is much higher than that of the original steel ball material. Moreover, this repairing lm should be different from the physical mixture of FGO and MTT due to the synergistic effect between two components of FG/MTT, which is prepared by chemical reaction.
To further reveal the friction-reducing and antiwear mechanism of the FG/MTT dispersed oil, the valence states of several typical elements on the wear scar of the steel ball are examined by XPS measurements. Fig.9 shows the full spectrum from the binding energy of 0 to 1200 eV and the curve-tted XPS spectra of C1s, O1s, Si2p, Fe2p, Na1s, and Cr2p on the wear scar surface. The C1s spectrum shows the peak at 285.2eV, which can be attributed to the carbon atoms in the functional group: C-C [23,24]. It proves that FGO is attached to the wear scar surface, which can also be seen from the silane bond in FGO corresponding to the peak at 101.2 and 102.1eV in the Si2p spectrum. Existing studies have shown that silicate minerals will generate SiO 2 after friction [25] In the FG/MTT wear scar XPS spectrum, the Si2p peak at 103.8eV and the O1s peak at 532.8eV also reveal that the FG/MTT tribochemical reaction occurred during the friction process to generate SiO 2 . The peak of O1s at 531.9 eV indicated the formation of metal carbonates [26], which is accorded with the emergency of the C1s peak at 288.2eV [27].  [19,28,29]. The formation of the layer containing Fe 2 O 3 could be bene cial to the improvement of the performance [28]. The Si2p peak at 102.9eV and 100.7eV are ascribed to chemical bonds of aluminosilicate and SiC, respectively [27]. These may be formed owing to the ash temperature and pressure during the friction process.
The chemical bond in FG/MTT is broken by mechanical force, releasing active groups (O-, Si-, -Mg-OH [30]) and chemically reacting with the substances on the surface of the friction pair to form the aluminosilicate, SiC, Fe 2 O 3, and SiO 2 . SiO 2 and SiC have good mechanical properties and creep resistance, lling the defects of the grinding surface to improve the surface hardness [31]. The peaks at 1071.6 eV of Na 1s originate from the FG/MTT composite, which testi es the destruction of the chemical bonds between the layers of MTT during the friction process. The peaks at 586.8 and 576.9 eV of Cr2p originated from Cr 2 O 3 , a product of the tribology reaction, which also protects the friction surface [32].

Lubricating Mechanism Analysis
According to the experimental results, the lubrication mechanism of FG/MTT can be explained as follows Fig.10(a) and (b) are microscopic schematic diagrams of the surface of the friction pair, During the friction process, the FG/MTT nanosheets are uniformly distributed on the surface of the friction pair under the action of mechanical force to form a "lubricating layer", which avoids some asperities of the friction pair directly contact with each other and reduces the wear of the workpiece. FG/MTT is a twodimensional lamellar structure, and its interlayer bonding force is weak, which makes the layer slide easily, thereby playing a lubricating effect and reducing the coe cient of friction, and form a selfrepairing layer as the SEM and EDS images of the worn scars exhibit.  Figure 10(e). The Mg/Al atoms in MTT can be replaced by other metal atoms [33], and the collapse of asperities will release energy by friction, resulting in ash temperature, which make the Mg/Al atoms in the MTT exchange with Fe atoms in the friction pair, and aluminosilicate and SiO 2 are produced in the tribochemical reaction [25].
XPS energy spectrum analysis shows the existence of SiC and Fig.10(d) illustrates the generation principle. The ash temperature engendered by the energy released from the disintegration of the asperity provides the reaction conditions. The reaction equation is as follows: The formula of FC in the friction energy model proposed by Heilmann and Rigney is as follows [34]: f g, W, τ max and τ s mean furrow friction coe cient, load, ultimate shear force, and average shear force of grinding surface, respectively. Since the real contact area (A r ) is more di cult to determine, analogous to the de nition of hardness H as load/contact area, using 1/H instead of A r /W. Eq. (3) can be transformed into the following form: According to the Eq. (4), it can be seen that the FC is inversely proportional to the hardness, and the hardness test shows that the hardness elevates after adding FG/MTT, the FC will also decrease accordingly, thus providing lubrication In general, FG/MTT not only avoids the contact of some asperities, but its special two-dimensional structure also plays an eminent role in declining friction and ameliorates the surface hardness of the friction pair, thereby reducing the friction coe cient. As a result of the frictional ash temperature and high pressure, complex tribological reactions may among MMT, FGO, and metal elements on the surface of the friction pair, to generate SiC, SiO 2 , aluminosilicate, and Fe 2 O 3 . The above substances could cover the metal surface to form a repair layer to protect and repair the surface.

Conclusion
In this work, the FG/MTT lubricant additive was successfully prepared, After that, the structure and morphology is characterized and the tribological properties is tested. According to the characterization results of XRD, TEM, FT-IR, and Raman spectroscopy, and infrared absorption spectroscopy, it can be proved that FG/MTT was successfully prepared. By analyzing the consequences of the tribological test and the settlement experiment, the following conclusions are drawn: FG/MTT can be evenly dispersed in oil. After comparing the test results of GO, MTT, FGO, and FG/MTT four-ball friction tester, it can be concluded that the lubrication performance of the above four substances has been improved. The performance of FG/MTT is the best one, whose the test nal FC is 57.5 % lower than that of 15W40 simple oil. After exploring the lubrication performance of different concentrations of FG/MTT, it is certi cated that the performance is best when the concentration is 0.4 mg/ml, and the average friction coe cient and wear scar diameter are reduced by 50.9 % and 19.1 %, respectively.
The hardness test shows that after adding GO, FGO,and FG/MTT, the hardness of the grinding surface has been increased by 6.9 %, 10.0 %, and 22.4 % respectively, which improves the bearing capacity of the grinding surface.
To analyze the lubrication mechanism of FG/MTT, the grinding surface after the friction experiment was analyzed by EDS and XPS, and the lubrication repair mechanism was explained as follows FG/MTT is a two-dimensional layered composite material. During the friction process, its micronanosheets ll between the asperities on the contact surface of the friction pair to form a lubricating layer and avoid direct contact of some asperities.
The ash temperature generated during the friction process makes the FG/MTT and the friction pair carry out a tribological reaction. Owing to its special structure, FG/MTT will release the active group in the interlayer domain during the interlayer sliding process, which may react with the surface of the friction pair to form a repair layer composed of the SiC, SiO 2, aluminosilicate, and Fe 2 O 3, declining the wear of the grinding surface. Figure 1 The synthesis process of FGO and FG/MTT