Graphene superlubricity: A review

Superlubricity has drawn substantial attention worldwide while the energy crisis is challenging human beings. Hence, numerous endeavors are bestowed to design materials for superlubricity achievement at multiple scales. Developments in graphene-family materials, such as graphene, graphene oxide, and graphene quantum dots, initiated an epoch for atomically thin solid lubricants. Nevertheless, superlubricity achieved with graphene-family materials still needs fundamental understanding for being applied in engineering in the future. In this review, the fundamental mechanisms for superlubricity that are achieved with graphene-family materials are outlined in detail, and the problems concerning graphene superlubricity and future progress in superlubricity are proposed. This review concludes the fundamental mechanisms for graphene superlubricity and offers guidance for utilizing graphene-family materials in superlubricity systems.


Introduction
Human beings have a long history of understanding friction. Prehistoric people have recognized friction in two aspects: drilling wood for the fire and utilizing lubricants to reduce friction resistance in heavy transportation. However, as a very common physical phenomenon, friction is still not fully understood in the 21st century. Currently, approximately 1/3 of the global primary energy is consumed by friction, and approximately 80% of device failures are caused by wear, which is associated with friction. Nowadays, friction and wear consume 2%-7% of the gross domestic product (GDP) annually for industrialized countries [1,2]. These make minimizing friction as much as possible to save energy one of the main goals in both science and industrial communities [3,4]. With unremitting exploration, the understanding of the origin of friction has been deepened, and thus boosts the generation of innovative technology-superlubricity [5][6][7].
The concept of superlubricity was first proposed by Hirano and Shinjo [8] in 1990 to describe the lubricating state, in which the sliding friction and wear between tribopairs vanishes. This fantastic lubricating state was then defined as the sliding coefficient of friction (COF) is smaller than 0.01, and has become one of the most critical techniques to alleviate friction-induced energy losses and machine breakdown [9]. Over the past 30 years, significant achievement has been made in the superlubricity field, and it is classified into two main branches based on the characteristics of materials, including solid superlubricity and liquid superlubricity [10]. The typical materials for solid superlubricity include graphite and graphene materials [11][12][13][14], molybdenum disulfide (MoS 2 ) [15,16], diamond-like carbon (DLC) film [17,18], carbon nitride coatings [19], solid nanocomposites [20,21], and other solid materials [22]. The typical materials for liquid superlubricity include water [23], oil-based solutions [24][25][26], polyol solutions [27,28], acid solutions [29][30][31], salt solutions [32,33], and biological mucus [34][35][36]. Among these superlubricity materials, two-dimensional (2D) materials, such as www.Springer.com/journal/40544 | Friction the superlubricity mechanisms of graphene-family materials at multiple scales. Section 3 describes the superlubricity mechanisms of graphene as interface materials with the lubrication of liquids at the nano-scale and as additives in water-based lubricants at the macro-scale. Section 4 concludes this review and provides suggestions and perspectives for future studies on graphene-family materials in terms of superlubricity.

Graphene in solid superlubricity
Graphene is an atomic thin material with low surface energy, so that can decrease friction and adhesion on diverse substrates [47]. Taking its ultrathin thickness into account, graphene is a favorable lubricant for solid lubrication. Therefore, much research on the solid superlubricity behaviors of graphene-family materials has been conducted, and the corresponding solid superlubricity mechanisms have been revealed and concluded.

Incommensurate contact
It has been proved that one graphene nanoflake can be stable on an underlying graphene sheet with commensurate contact. This is because the atoms non-adiabatically jump beyond potential barriers between adjacent sites, resulting in energy dissipation throughout the friction course [48]. With incommensurate contact, the reduced energy barrier suppresses the energy dissipation, thereby resulting in superlow friction force between the graphene nanoflake and graphene sheet. The incommensurate contact has been considered the primary mechanism for the superlubricity achieved at graphene interlayers or graphene/gold interface [49] and graphene/WS 2 interface [50]. Dienwiebel et al. [51] studied the friction performance of graphite nanoflakes and graphite sheets in 2004. The results showed that the rotation angle of graphite nanoflakes concerning graphite sheets greatly influenced friction behavior, and the superlubricity was attributed to the incommensurate contact between graphite layers. Feng et al. [52] studied the friction performance between graphene nanoflakes and a graphene surface in 2013. During the transition from the initial commensurate state to the final commensurate state, both rotational and translational motions occurred. The commensurate contact state between graphene layers was first transferred to an incommensurate contact state, and then the graphene flake on top slipped quickly to achieve a commensurate contact state. The mean sliding distance of graphene flake during the transition from incommensurate to final commensurate state was shorter at a relatively higher temperature of 77 K (Fig. 2). This phenomenon implied that the transition from an incommensurate contact state to a commensurate contact state was probably induced by thermal fluctuation. Later, Liu et al. [13] proved that this thermally activated process still functioned to achieve superlubricity under ultrahigh vacuum conditions over the temperature range of 125-448 K. They performed nano-scale friction tests with a graphite-wrapped atomic force microscopy (AFM) probe sliding against a graphite substrate and achieved a record-low COF value of 4×10 −5 .
Nevertheless, the friction measurement for an ideal graphene interface stayed a challenge. Therefore, various approaches, such as growing a graphene film on a microsphere by chemical vapor deposition (CVD) and wrapping graphene onto an AFM probe, were Fig. 2 (a) Rotation model of graphene nanoflake, (b) interaction energy evolution between graphene nanoflake and graphene surface with rotation angle, and (c) average sliding distances of graphene nanoflake at various temperatures. Reproduced with permission from Ref. [52], © American Chemical Society 2013. employed to investigate the friction properties of graphene. Liu et al. [38] measured the friction of graphene interlayers and graphene/h-BN interface by utilizing a graphene-deposited SiO 2 microsphere. The superlubricity was achieved unrelated to the rotation angles of the two surfaces. In addition, the stable superlubricity was achieved under a local contact pressure of as high as 1 GPa and was as well insensible to the humidity. The stable superlubricity was ascribed to the general incommensurate contact between the contact surfaces, which were wrapped by graphene in random directions.
In addition to experimental investigation, the superlubricity of graphene-family materials was also studied through various simulation and calculation methods. Cahangirov et al. [53] simulated the superlubricity of multilayer graphen (MLG) coating on the Ni(111)/Ni(111) friction pair by using the first-principles calculations. MLG was placed between two parallel Ni surfaces, which shielded the strong attraction of the Ni surface and significantly reduced the adhesion and sliding friction. In addition, the combination of the single graphene layer and Ni surface weakened the coupling between graphene due to the chemical interaction between Ni-3d and graphene-π orbitals, and then realized the superlubricity characteristics of graphene on the Ni friction pair. Wang et al. [54] assembled van der Waals heterostructures with graphene and MoS 2 monolayers and achieved superlubricity. According to the density functional theory (DFT) calculation, the decrease of lateral force constant between layers was caused by the decrease of potential energy corrugation in the sliding process. The change of potential energy corrugation was determined by the fluctuation of interfacial charge density caused by the sliding. The superlubricity behavior of fluorographene/MoS 2 heterostructures was also studied by the DFT calculation [45]. It was found that the potential energy surface between fluorographene/MoS 2 heterostructures was very smooth, so there was almost no obstruction and energy loss in the sliding process, and then superlubricity was realized. This study showed that superlubricity depended on the formation of the Moiré pattern, which led to the elimination of local energy and the change of force. Leven et al. [55] used the resistivity imaging (RI) method to simulate the sliding physical characteristics of graphene sheets on the h-BN layer. The research showed that when the graphene sheet is large enough, the sliding friction would be very small regardless of the relative direction between two lattices to achieve robust superlubricity. Ansari et al. [56] used the RI model to study the superlubricity phenomenon of graphene and h-BN homo-and hetero-junctions. The results showed that 585-extended line defects acted as structural anisotropy enhancers in crystal, which significantly reduced the friction between two heterostructures or even homogeneous bilayers. Koren and Duerig [57] revealed the superlubricity phenomenon in a twisted bilayer graphene system with a 30° distortion through numerical simulation. The study showed that the geometric sequence of fundamental dodecagonal tiling elements in the quasicrystal structure had a superlubricity state in the lubrication process. The superlubricity behaviors of graphene interlayers and between graphene flakes and graphite were also validated by the molecular dynamic (MD) simulation [58]. It was found that the interlayer defect of vacancy had nearly no influence on the superlubricity behavior of graphene interlayers with incommensurate contact at a certain orientation [59]. Liu et al. [60] simulated the high-speed sliding of a graphene sheet on a graphite substrate and studied its superlubricity behavior. They showed that superlubricity was interrupted as the graphene flake rotated through continuous crystallographic alignments with graphite substrate. Ru et al. [61] studied the relationship between twist angle (θ) and COF values in bilayer graphene and MoS 2 /MoSe 2 heterostructures. They showed that by twisting a certain angle, incommensurate contact was obtained, and thus the superlubricity state. Recently, Bai et al. [62] proposed a misfit interval statistical method to quantitatively identify the geometric features of the Moiré pattern, in which the distribution of lattice mismatch was used as an indicator to present whether the interface was in the superlubricity state [63]. Besides, they showed that the θ and contact size played a key role in the generation of Moiré pattern and superlubricity achievement (Fig. 3).
Incommensurate contact-induced superlubricity is limited to being achieved at nano-and micro-scales. Achieving this kind of superlubricity in graphenewww.Springer.com/journal/40544 | Friction family materials at the macro-scale stays a challenge because of the dilemmas in sliding large contact areas of commensurate stacking fields. Both Li et al. [64] and Androulidakis et al. [65] achieved superlubricity through incommensurate contact at the macro-scale in 2020. A new principle was proposed to extend the superlubricity state based on countless micro-contact to macro-scale superlubricity (Fig. 4) [64]. The surface morphology of the steel substrate is carefully reconstructed, and the macro-surface is divided into countless micro-scale points to transform the macrosurface contact problem into micro-point contact. Thereafter, each micro-contact point was specially treated, such as pre-wear-in and coating nanocomposite of 2D materials, to achieve interlayer and incommensurate contact. Finally, the COF of the steel self-mated friction pair was greatly reduced to smaller than 0.01, and superlubricity at the macro-scale was achieved. This superlubricity has good universality to a wide range of materials, which is of significance for the industrialization of superlubricity. Androulidakis et al. [65] showed the existence of macro-scale superlubricity between graphene layers by measuring the shifts of Raman peaks. The random incommensurate stacking between graphene layers was responsible for the superlubricity achievement at the macro-scale.

Material transfer
Material transferring and forming a transfer film is also important for the achievement of superlubricity with graphene-family materials. The transfer film makes the contact between graphene surface and other material surfaces be changed into that between graphene-graphene contact. At the nano-scale, the contact between graphene nanoflakes facilitates the achievement of incommensurate contact and contributes to superlubricity achievement [66,67]. At the macro- | https://mc03.manuscriptcentral.com/friction scale, the transfer film of graphene-family materials can also greatly enhance the tribological performances of materials [68].
Li et al. [67] achieved superlubricity at SiO 2 /highly oriented pyrolytic graphite (HOPG) interface under an ambient condition at nano-scale. The superlubricity state could be tuned by changing the contact pressure. When the contact pressure exceeded 2.52 GPa, the measured COF was increased 10 times; while the contact pressure decreased to lower than 2.52 GPa, the superlubricity state was recovered. The failure of superlubricity under high contact pressure was ascribed to the delamination of top layers on HOPG, which demanded extra exfoliation energies. The superlubricity mechanism was attributed to the formation of a transfer film of graphene nanoflakes on silica while sliding against HOPG. The graphene transfer film made the sliding interface transfer into the graphene/HOPG interface (Fig. 5), and the stable superlubricity was due to the incommensurate contact between the graphene nanoflakes of the transfer film and HOPG substrate [66]. Later, Li et al. [69] prepared a gold nanocrystal-coated AFM probe, which was used to rub against a HOPG substrate. The formation of multiple gold-graphite heterogeneous interfaces in the contact area was responsible for the achievement of superlubricity. Tian et al. [70] achieved superlubricity by sliding an AFM probe against the WS 2 /graphene film. It was observed that graphene transfer would occur to the AFM probe during the friction process. Then, incommensurate contact between the transferred graphene probe and WS 2 /graphene film would be achieved, and superlubricity could be achieved. Most recently, Yu et al. [71] achieved superlubricity by the tribo-induced material transfer method. The COF achieved between graphene-transferred AFM probe and graphite substrate was as small as 10 −4 . This method could be applied to other 2D materials as well and provide an experimental platform to investigate the interlayer friction of 2D materials.
The macro-scale friction properties of graphene are more important for industrial applications, which have also been investigated. Macro-scale friction behavior of graphene layer fabricated by self-assembly process was studied by Li et al. [72]. They found the intermittent superlubricity phenomenon between graphene transfer film and HOPG substrate at the macro-scale. The COF as low as 0.001 was achieved by the formation of a graphene transfer film, while the counterpart steel ball slid against the HOPG substrate. The macro-scale COF could be attributed to the statistical frictional force of the transferred film sliding against the substrate in the existence of atomic steps at the nano-scale (Fig. 6) [72]. This resulted in random superlubricity with a short duration because the presence of atomic steps significantly decreased the possibility of superlubricity from a statistical viewpoint. This study provided evidence that the www.Springer.com/journal/40544 | Friction transferred graphene layer played a critical function in superlubricity achievement with graphene.

Structural changes
In recent years, researchers have focused on the influence of structural changes in graphene on tribological behaviors. Lee et al. [73] investigated the frictional behaviors of 2D materials, finding that the friction between the AFM tip and 2D materials is dependent on the number of atomic layers of those 2D materials. To figure out the inherent mechanisms of the thickness-dependent frictional behaviors of 2D materials, the experimental results were reproduced by using atomistic simulation, indicating that besides the true contact area, the quality of the atomic-scale contact (mainly commensurability and local pinning state for their case) is also important for the frictional behavior of 2D materials [74]. Zhang et al. [75] found that the friction of graphene sheets can be tuned reversibly with mechanical strain. The COF between silicon nitride probes and graphene decreased with an increased tensile strain of graphene. Superlubricity with a COF of nearly 0.001 can be achieved on strained graphene. The reduced COF can be attributed to the regulated contact quality through the in-plane strain of graphene. Wang et al. [76,77] investigated the friction between graphene flakes atop the strained graphene substrate. It was found that the frictional behaviors between graphene layers can be modulated by strain engineering of graphene. Robust superlubricity can be achieved through both uniaxial and biaxial stretching above a critical strain. The frictional behavior with strain engineering is irrelevant to the relative orientation mainly due to the complete lattice mismatch, which manifested as Moiré patterns between graphene layers. The evolution of the Moiré patterns plays a critical role in the lubrication behavior [55], where the friction force is significantly reduced attributed to the cancellation of lateral forces between the graphene layers within one supercell of the Moiré pattern. Most interestingly, it was found that the ratio between the area for the Moiré pattern and the graphene nanoflake is important for the lubrication performance. For the graphene flake orders larger than the area of the Moiré pattern, the interlayer friction can be reduced to the same order for an incommensurate state, which can be attributed to the Moiré pattern area dominating the frictional behavior instead of the rim area.
The structure evolution of graphene-family materials during friction is significant for superlubricity achievement [78]. This is derived from graphene sheets wrapping nanoparticles and forming nanoscrolls or similar nanostructures, which lead to contact region reduction and result in an incommensurate contact. Combined with the nano-ball bearing effect of nanoscrolls, which makes sliding friction transfer into rolling friction [79], thereby contributing to significant friction reduction to superlubricity level. Li et al. [80] achieved superlubricity by using graphite-like carbon and fullerene-like carbon as friction pairs in the N 2 environment. The existence of MLG nanoscrolls at the interface was verified by the Raman spectroscopy, which was also found to be aligned in the same direction. They believed that graphene nanoscrolls formed during the friction process acted as micro-ball bearings, which could bear partial pressure and reduce interlayer shear in the contact area. Combined with the incommensurate contact, the COF was greatly reduced to the superlubricity level. Macro-scale superlubricity can be also obtained through transitionmetal dichalcogenide nanoflake coating deposited on the amorphous carbon (a-C) substrate [81]. Load-driven graphitization of a-C wear products can be observed during the wear-in process, leading to the formation of van der Waals heterostructures and the realization of macro-scale superlubricity.
Graphene-family materials were as well combined with other materials, possessing various properties to achieve improved lubrication properties and condition adaptability. Berman et al. [82] demonstrated a new approach to achieving macro-scale superlubricity. In their study, the graphene was merged with nanodiamond or metal nanoparticles to form a friction pair with DLC film. A special nanoscroll or similar nanostructure was formed during the friction process and contributed to the macro-scale superlubricity achievement. Later, Jiang et al. [83] used MoWS 4 +graphene heterogeneous composite film to achieve superlubricity on steel/steel friction pair in dry Ar environment. They found that the wear debris had a scroll structure, and as a result, the steel/steel | https://mc03.manuscriptcentral.com/friction interface was transferred into the tribolayer-nanoscrolltribolayer interface, offering an incommensurate contact and resulting in macro-scale superlubricity. Zhang et al. [84] achieved superlubricity by using the graphene-coated microsphere as the lubricant on the graphene-coated ball and graphene-coated plate friction pairs under ambient conditions. The COF value of 0.006 was achieved in the air. The superlubricity was mainly attributed to two aspects. First, there were some exfoliated graphene flakes between the friction pairs, which provided very low shear strength. The second one was the swing and sliding of the microsphere in the contact area. The micro balls played a buffer function in the contact area, reducing the contact area of the graphene sheet, and the swing motion helped to distribute contact points to avoid the stress concentration of the graphene sheet on top. In these studies, nanoparticles played an important role in forming nanoscroll structure and consequently superlubricity achievement. Therefore, further revealing their mechanisms is of importance for the design and control of macro-scale superlubricity. Most recently, Li et al. [85] prepared a graphene/nano-SiO 2 particle composite coating on a monocrystalline silicon substrate and studied its superlubricity behavior. It was confirmed that nanoparticles could promote the formation of nanoscrolls as well as play the role of load-bearing. The nanoscroll structure is the key factor to achieving superlubricity rather than the counterpart. These findings were beneficial to the design and control of macro-scale superlubricity (Fig. 7).
In addition to nanoparticles, graphene-family materials were as well merged with polymers to improve the tribological behaviors of coatings [86]. Macro-scale superlubricity has been achieved with a kind of polyethyleneimine/graphene oxide (GO) coating formulated in a dry N 2 atmosphere [87]. The analysis results of worn debris indicated that carbon nanoparticles were generated in a dry N 2 atmosphere or under vacuum conditions. These nanoparticles performed as rollers and led to a decrease in friction. These studies proved that by structure evolution of graphene-family materials, friction could be greatly reduced to superlubricity level.

Tribochemical reaction
Tribochemical reaction has well been documented as one of the main lubrication mechanisms. In the superlubricity domain, it contributed to friction reduction to the superlubricity level as well. Berman et al. [88] achieved macro-scale superlubricity with MoS 2 nanoflakes and nanodiamond particles deposited on Si/SiO 2 substrate against DLC ball under a N 2 Fig. 7 Superlubricity mechanism of graphene/nano-SiO 2 composite coating: (a) friction pair, (b) graphene/nano-SiO 2 particle composite coating at the interface, (c) role of nano-SiO 2 particles in promoting the formation of graphene nanoscrolls, and (d) solid-structured nano-SiO 2 particle-supported applied load. Reproduced with permission from Ref. [85], © The author(s) 2021.
www.Springer.com/journal/40544 | Friction atmosphere. During the friction process, nanodiamonds were transformed into onion-like carbon (OLC) by the tribocatalysis of MoS 2 . In Ref. [89], it was found that iron nanoparticles can also lead to the transformation of DLC to OLC. The in-situ formed OLCs with a diameter of 30-50 nm can separate the friction pairs and act as ball bearings, leading to the achievement of macro-scale superlubricity. Yin et al. [90] used graphene quantum dots to modify the properties of DLC film and achieved superlubricity in dry N 2 environment. They demonstrated that the superlubricity was achieved under a high speed of 15 cm/s and a large contact pressure of 1.9 GPa. They concluded that the formation of tribolayer consisted of a-C and silicalike SiO x was the key factor for achieving superlubricity.

Other mechanisms
The superlubricity behavior of graphene-based materials is of great complexity, especially at the macro-scale. Various new mechanisms have been proposed to explain the macro-scale superlubricity behaviors. Li et al. [91] investigated the tribological behaviors of the composite coating with ball-milling exfoliated graphene few-layer MoS 2 deposited on fullerene-like hydrogenated carbon (FL-C:H) and a-C:H substrates. Macro-scale superlubricity can be achieved at ambient conditions under 15 N with a composite coating deposited on FL-C:H substrate. It was found that the high elasticity of FL-C:H substrate is important for the maintenance of 2D material integrity and the achievement of heterointerfacial contacts, leading to macro-scale superlubricity under a high normal load. Besides the heterointerfacial contacts between different 2D nanoflakes, heterojunction can be also obtained through the structural changes at the sliding interface. Graphene-family materials have also been merged with other materials to promote friction behavior by synergy effect. Upadhyay et al. [92] utilized graphene and MoS 2 as fillers to formulate epoxy-based nanocomposite. The contents of graphene and MoS 2 ranged from 5-20 wt%. COF values of 0.017-0.028 and wear rates of (1.35-1.59)×10 −7 mm 3 /(N·m) were gained with the graphene-containing epoxy composites, and even smaller COF values of 0.0019-0.01 and wear rates of (1.00-1.48)×10 −7 mm 3 /(N·m) were gained with both graphene and MoS 2 -containing epoxy nanocomposite. The addition of graphene prevented MoS 2 from reacting with H 2 O molecules in the air to generate trioxides and sulfides of molybdenum, which can be attributed to the high energy barrier to the path of oxygen atom provided by graphene [92]. As a result, the bonding between the hydroxyl group and the composite surface was weakened, and the fillers could be easily sheared by the steel ball to generate tribolayer for greatly friction reduction [93].
As one of the thinnest solid lubricants, one-atom-thick graphene can present outstanding friction behavior. Meanwhile, graphene has been utilized to produce coatings or utilized as additives to improve the performances of polymers, metal, and other materials. Even though the research on solid superlubricity related to graphene-family materials still stays in the beginning period, the study results prove that graphene-family materials are promising lubricants in superlubricity domains (Table 1). It is expected that even better friction behavior would be achieved with graphene-family materials in the near future.

Graphene in liquid superlubricity
Graphene has been proved to perform outstandingly not only in solid superlubricity but also in liquid superlubricity, in which graphene is commonly used as tribopair materials or lubricating additives at multiple scales. The main mechanism for liquid superlubricity at the nano-scale is weak interactions between graphene and liquid molecules, while that at the macro-scale is complicated. The mainstream mechanisms for generalized liquid superlubricity at the macro-scale include the hydration effect, double electric layer effect, and fluid effect [94]. Hydration layers are generally formed by charges enclosed by water molecules. The hydration layers can support a part of the load without being extruded and maintain the fluidity features of the liquid while being compressed [95]. The typical examples that achieve liquid superlubricity through the hydration effect include hydrated ions, surfactants, phosphatidylcholine liposomes or bilayers, and polymer brushes [96][97][98][99]. In terms of the double electric layer, it is sensitive to the surface charge distribution, and minor charge Friction 11(11): 1953-1973 (2023) | https://mc03.manuscriptcentral.com/friction reverse would appear for highly concentrated and confined salt solutions, which depends on the bulk concentration and surface charge density [100]. The charge reverse phenomenon at high concentrations has proved to be beneficial to superlubricity achievement [32]. Liquid superlubricity can also be achieved when a fluid film is formed via the hydrodynamic effect to separate the contact surfaces and provide a low shear resistance. The typical examples that achieve liquid superlubricity through fluid effect are water and acid-based solutions [101,102].
Different from the liquid superlubricity achieved with pure liquids at the macro-scale, the liquid superlubricity mechanisms involved in graphene-family materials rely on the combination of tribochemical reactions, adsorption effect, and fluid effect in general, all of which account for the improved lubrication performances and superlubricity achievement.

Weak interactions between graphene and liquid molecules
Graphene was proved to achieve a superlubricity state at the nano-and micro-scales when incommensurate contact with special liquid molecules was developed [103]. Li et al. [104] achieved superlubricity at graphene/ hydrophobic self-assembled fluoroalkyl monolayer (SAFM) interface when lubricated with water. The tests were carried out on the AFM, and a superlow COF of 0.0003 was achieved when the contact pressure was below 14.5 MPa (Fig. 8). The MD simulation result presented that a several-nanometer-thick water layer was intercalated between graphene and SAFMs. Because of the low contact pressure (14.5 MPa), the intercalated water layer would not be squeezed out. Thanks to the weak interaction between graphene and water, sliding between graphene and water molecules would easily occur and resulted in superlubricity achievement. They also reported a superlow COF value of 0.001 when the zwitterions in a lipid bilayer were sliding against graphene under water lubrication conditions [105]. They proved that superlubricity could be achieved by water intercalation between graphene and other surfaces, and provided a new method to reduce the friction between graphene and other surfaces. Most recently, Zhang et al. [106] achieved superlubricity on the HOPG surface with the lubrication of ionic liquids. The research showed that cations at negative voltages and anions at positive voltages could contribute to superlubricity achievement by tuning the properties of the boundary layer.

Combined mechanisms for liquid superlubricity at macro-scale
The anti-wear mechanism of graphene-family materials generally accounts for the formation of graphene adsorption films and tribochemical layers, which protect friction pairs from direct contact and result in wear resistance [107]. In the case of friction, GO as additives have proved to be beneficial for friction reduction to a COF value of 0.02 [108]. Nevertheless, the COFs of graphene-family materials that acted as additives in liquid lubrication were generally in the range of 0.02-0.1 [109][110][111][112], which was much larger than the superlubricity level.
In 2018, Ge et al. [113] achieved superlubricity through the synergy effect of GO and ethylene glycol (EG) at various interfaces, including Si 3 N 4 /SiO 2 , Si 3 N 4 /Si 3 N 4 , and Si 3 N 4 /sapphire interfaces at the macro-scale (Fig. 9). The COF decreased to smaller | https://mc03.manuscriptcentral.com/friction than 0.01 after a wear-in period of approximately 600 s; afterward, the COF further decreased to 0.0037 and was stabilized for 2 h. Besides, the results showed that the wear volume after lubrication with GO-containing EG was only 5% of that of EG. The superlubricity and wear-resistant capacities were ascribed to the GO adsorption layer formed at the interface, which kept the surfaces from severe wear. The superlow shear strength of GO interlayers resulted in superlubricity achievement and wear-resistant behaviors. Additionally, the formation of a hydrated network at the GO/EG interface resulted in a low shear strength and contributed to superlubricity achievement. Moreover, they proved that the GO adsorption layer and tribochemical layer alone could not lead to the achievement of superlubricity, and severe wear would occur quickly. The existence of fluid film could protect the GO adsorption layer and tribochemical layer in addition to forming a hydrated network at the GO/EG interface. Therefore, GO should be utilized in combination with fluid film to efficiently utilize the synergic effect of them, which in turn results in the achievement of the liquid superlubricity at the macro-scale. Recently, it was found that the graphene oxide quantum dots (GOQDs) as nanoadditives can significantly reduce the wear of friction pairs and simultaneously shorten the wear-in process before achieving superlubricity [114]. With the lubrication of EG solution with GOQDs, superlubricity with a COF of 0.0068 can be achieved with an only 6 s wear-in period. It was found that the adsorbed GOQDs in the tribolayer play an important role in the enhanced tribological performance, which can be attributed to the interlayer shearing and rolling effect for boundary lubrication and the enhanced load-carrying capability provided by GOQDs.
The introduction of graphene-family materials into liquid lubricant usually results in lower friction and enhanced wear-resistant. However, the choice of graphene-family materials is largely limited by the dispersity and stability of graphene-family materials in the lubricants. Aiming to the problem, a novel strategy using hydrophobic graphene coatings and an aqueous solution of glycerol was proposed by Liu et al. [115] to achieve macro-scale superlubricity with a COF of 0.004 and suppressed wear of friction pairs. It was found that the in-situ formed tribolayer containing graphene nanoflakes is critical for achieving superlubricity, which can provide boundary lubrication through the low interlayer shearing strength of graphene.
Nevertheless, the research into graphene-family additives in liquid superlubricity is still in the preliminary period. Ge et al. [116] made a comparison between graphene-family additives and traditional additives to review the feasibilities of graphene-family materials in terms of liquid superlubricity achievement. According to the data of liquid superlubricity research over the years (Table 2), they found that the contact pressures during the superlubricity period were rarely exceeding 300 MPa (without coatings). This may hinder the application of liquid superlubricity. To solve this issue, they combined GO with ionic liquid and achieved superlubricity at the Si 3 N 4 /sapphire interface under a high contact pressure of 600 MPa (Fig. 10) [116]. An adsorption layer of GO was observed on the worn surfaces directly, which may exhibit load-carrying capacity and keep the solid asperities from direct conflict. Moreover, a boundary layer formed by ionic liquid performed as outstanding wear-resistant material and resulted in low wear. The superlow shear strength and extreme pressure property of GO interlayers played a critical part in achieving superlubricity under high contact pressures. These results evidently showed that graphene-family materials possessed better superlubricity properties than conventional materials and proved graphenefamily materials the promising materials in the domain of liquid superlubricity.
It is known that graphene-family material could present different properties depending on its functional groups [117,118]. Nevertheless, which functional groups within GO contributing to friction reduction, wear-resistant, and superlubricity are still required to be revealed. Ge et al. [119] studied the effect of functional groups on the superlubricity achievement of GO-containing liquid lubricant (Fig. 11). They found that the -NH 2 group was better than -OH and -COOH in terms of superlubricity achievement because the larger adhesive force between functional groups and contact surfaces. The larger adhesive force made the adsorption layer of GO-NH 2 more robust,

Conclusions and perspective
This work reviews the superlubricity performances of graphene materials in detail. The lamellar structure and excellent physical property make graphene a promising material for achieving superlubricity, and the corresponding superlubricity mechanisms are illustrated in Fig. 12. Graphene-family materials have drawn much attention as favorable lubricants thanks to their exclusive physical, chemical, and mechanical properties. Numerous advances have been accomplished with the improvement of graphene production technologies and trial technologies. Graphene-based superlubricity systems have already shown prospects for potential engineering applications. Recently, Peng et al. [120] found that a superlubricity and wear-free sliding condition for over 100 km can be achieved by graphite/DLC hetero-junctions under ambient conditions. The achievement of the wear-free superlubricity condition can be attributed to the weak van der Waals interaction between friction pairs, atomic-smooth interface, high in-plane strength of graphene, and the low out-of-plane stiffness of the graphite flake. Besides sliding friction, rolling-sliding friction can be also commonly observed in mechanical parts like a rolling bearing. A superlubricity state with a COF of around 0.003 was achieved by Mutyala et al. [121] through DLC-coated pairs with the lubrication of 2D MoS 2 nanoflakes combined with nanodiamond particles in an oil-free and dry nitrogen environment. It was found that the formation of tribofilms containing diamonds with graphene-like outer layers formed by rubbing is critical for the achievement of the superlubricity state. The proposed superlubricity system led to a reduction of COF by more than 20 times compared to oil-lubricated conditions; meanwhile, the wear of friction pairs was hard to be observed. Those studies broadened the application of superlubricity systems for engineering applications in the future. Nevertheless, several issues for graphenefamily materials in lubrication applications still exist. Solid superlubricity involved with graphene-family materials is limited at nano-and micro-scales, while the friction usually grows larger at macro-scale. In addition, the weak adhesion capacity of graphene on to substrate still impedes the engineering purposes of graphene as coating materials. Besides, environmental factors influence the lubrication behaviors of graphenefamily materials in general. There are three strategies to overcome these drawbacks of graphene-based materials as solid lubricants generally. Firstly, enhanced lubrication and wear-resistance performance can be achieved by the functionalization of graphene, which has been verified by both experiments [68,122]  www.Springer.com/journal/40544 | Friction and simulation [123]. Secondly, structural regulation of graphene-based materials is also an effective strategy to achieve better tribological performance. Researchers found that the tribological behaviors of graphene-based materials can be improved through strain engineering [65,[75][76][77] or enhanced interfacial bonding between graphene-based materials and substrates [124]. At last, superlubricity can be more easily achieved through the synergetic effect with graphene-based materials and other nanomaterials. Previous literature suggested that the macro-scale superlubricity can be achieved by the in-situ formed nanoscrolls with graphene-based materials and nanoparticles [82,84,85] or the heterostructures containing graphene and other 2D materials [64,81,91]. Better environmental adaptivity can be also achieved through the synergetic effect between graphene-based materials and other nanomaterials [125,126]. In terms of liquid superlubricity, the major problems are related to the unsteadiness of graphene-family materials in oil-based lubricants and the high cost of graphene-family material production. For now, the addition of stabilizers, surface functionalization, and other techniques are widely employed to enhance the stability of graphene-family materials in oil-based lubricants. In addition, it was also found that the tribological behaviors and the stability of graphenebased nanoadditives can be further improved by specified functionalization [111,119,127,128]. Nevertheless, there is still room for enhancement that needs to be resolved for applications of graphenecontaining oils. Besides, the cost of graphene production is usually expensive, and thus new techniques Fig. 12 Outline of the mechanisms of graphene-family materials in superlubricity. In solid superlubricity, the major mechanisms include incommensurate contact, material transfer, and structure evolution. In liquid superlubricity, the main mechanisms include weak interactions between graphene and liquid molecules and the graphene adsorption layer. Reproduced with permission from Ref. [52], Friction 11(11): 1953-1973 (2023) | https://mc03.manuscriptcentral.com/friction are needed to economically achieve the large-scale production of graphene-family materials and graphenecontaining lubricants for engineering purposes. In addition, graphene-family materials can only achieve liquid superlubricity as the additives in water-based lubricants. The COFs of graphene-family additives in oil-based lubricants usually lie in the range of 0.02-0.1, and achieving superlubricity in oil with graphene-family additives is still a challenge. Graphenefamily additives are difficult to disperse in oils, which returns to the first problem.
It is considered that these concerns would be resolved soon, and then graphene-family materials would be used as lubricants in many domains, including thermal power plants, automobiles, spacecraft, and machining industries. Because of the ability to provide improved lubrication performances, graphenefamily materials significantly facilitate the decrease in energy consumption during the friction process.