Perspectives of friction mechanism of a-C:H film in vacuum concerning the onion-like carbon transformation at the sliding interface

Abstract

A-C:H films with low friction and good wear resistance have long been regarded as a potential space lubricating film. However, its superlubricity mechanism and failure process in vacuum still remains to be improved. To clarify its friction mechanism, here, we systematically investigated the tribological property of a typical a-C:H film under a high vacuum environment. The results show that the extremely low friction coefficient lasts 2700 cycles under a contact pressure of 930 MPa, and the entire friction process can be divided into three stages. The friction coefficient was first stable with a low value after a short period of running-in, then it underwent an evident fluctuation period and further decreased to an extremely low value (0.005) until it abruptly failed. The structural evolution of the a-C:H film on a sliding interface for different periods through the entire friction process was characterized and a dynamic friction mechanism was established. The self-mated (a-C:H/a-C:H) friction process and hydrogen passivation contributed to the decrease of the friction coefficient in the early stages of sliding. Then, the emission of hydrogen became evident under high local stress and more dangling bonds were exposed on the worn surface, which leads to the wild adhesion wear between the sliding surfaces. The alternated process between an old film and new film is in consonance with the fluctuation of friction coefficient. Afterwards, carbon onions with a closed spherical shell structure are spontaneously formed on the worn surface in the absence of the hydrogen passivation effect, which further reduce the friction coefficient to an extremely low value. This study provides guidance to the further design of a new generation of a-C:H films with a special structure that exhibits a better tribological performance in a vacuum environment.

---

1. Introduction

Along with the increasing demand of the humankind for space exploration, the development of high-performance solid lubricants as well as solid lubrication technology is applicable to the long-term operation of mechanical assemblies in a vacuum environment is becoming increasingly urgent. However, the space environment, involving ionizing radiation, atomic oxygen, weightlessness, and sharp temperature changes, is very complicated and brings about challenges to traditional solid lubrication films.¹ Hydrogenated diamond-like carbon (a-C:H) films with low friction and good wear resistance, excellent biocompatibility, high hardness and electrical resistivity, as well as high chemical inertness have been widely referenced as potential space lubricating materials.^2–4 Unfortunately, the lifetime of a-C:H films for preserving the superlow friction regime in a vacuum environment is still not very satisfactory.^5–7 To date, a large number of attempts have been successfully made to explore the friction and failure mechanism of a-C:H films in vacuum.^8–11 The hydrogen passivation and emission mechanisms were found to exert a great influence on the tribology performance of a-C:H films in vacuum. When hydrogen content in the film increases to a certain value (34–40%), the extremely low friction coefficient can be maintained for a longer time.⁹ In addition, adding a certain amount of hydrogen during friction tests also helps to prolong the life span of a-C:H films in vacuum environment. As a-C:H film is a three-dimensional covalent structure, constructed of sp² and sp³ hybridized carbon atoms, chemical bonds in the film will be broken by the repeated friction process and large amounts of dangling bonds will be exposed on the worn surface, which cause strong chemical interactions on rubbed surfaces that results in a high film friction coefficient and rapid failure. If more hydrogen atoms exist in the film or in the environment, the dangling bonds can be well passivated and a low friction coefficient can be maintained.¹² However, many studies also indicate that the wear life of a-C:H films in vacuum do not just depend on hydrogen content. For example, the NFC-6 amorphous carbon film (the content of hydrogen is about 40%) that was prepared by Erdemir et al.^13,14 exhibited a superlow friction coefficient (about 0.001) in a dry nitrogen atmosphere, whereas the lubrication ability of the film only remained for dozens of cycles in a vacuum environment. Vanhulsel et al.¹⁵ also reported that the wear life of a-C:H film with a hydrogen content of 51% is shorter than that with 49% hydrogen in a vacuum environment. In addition, Chen et al.^16,17 reported that the life of a-C:H film is closely related with internal stress. Fontaine et al.¹⁸ claimed that the viscoplasticity also significantly influences the wear life of the film. Liu² prepared a novel hydrogenated amorphous carbon film with a self-organized dual nanostructure, which exhibited a super-high elastic recovery of 95% and excellent tribological properties under vacuum. These phenomena indicate that the friction and wear process of a-C:H film in vacuum is a very complex process and it is greatly influenced by many factors. A more in-depth study needs to be conducted to investigate the friction behavior of a-C:H films in vacuum, which can provide guidance for developing a new generation of a-C:H films with a longer life in vacuum.

Based on these previous studies, in the present research, the failure process of a typical a-C:H film with an extremely low friction coefficient in vacuum was investigated. In view of the fluctuation of the friction coefficient, the morphology and structure of the worn surface for different periods were characterized to establish a dynamic friction mechanism of a-C:H films throughout the entire friction process.

2. Experimental details

2.1 Preparation of a-C:H films

The a-C:H films were deposited on polished stainless steel substrates (1Cr₁₈Ni₉Ti; φ 24 mm × 8 mm, 220 HV) and silicon wafer substrates with a medium frequency unbalanced magnetron sputtering facility (SP-0806SI). The details about the deposition facility and deposition process are described elsewhere.¹⁷ First, a Si interlayer was deposited on the substrates in advance for about 280 nm to increase the adhesion strength between the a-C:H films and substrate. During the carbon film sputtering process, the deposition time was fixed at 3.5 h, and the a-C:H films were about 1.4–1.5 μm thick.

2.2 Characterization of a-C:H films

The fractured cross-section morphology of the as-prepared films was observed by a JSM-6701F field emission scanning electron microscope (FESEM). A JEOL-2010 high-resolution transmission electron microscope (TEM) was used to observe the microstructure of the as-prepared films. A Jobin-Yvon HR-800 Raman spectrometer with an Ar⁺ laser line of 532 nm was used to characterize the detailed bonding structures of the films as well as the worn surfaces. The nano hardness of the a-C:H films was evaluated using a Nanotest600 nanoindenter apparatus (Micro Materials Ltd., UK) with a Berkovich indenter at a load of 5 mN. The maximum indentation depth was less than 10% of the film thickness to minimize the substrate contribution.

2.3 The exploration of vacuum tribological behavior

Vacuum friction and wear tests were conducted with a rotational ball-on-disk tribometer that was sealed in a vacuum chamber. The details about this apparatus are described elsewhere.¹⁹ Friction tests of GCr15 steel ball (radius: 6 mm) against a-C:H films were conducted at a normal load of 5 N (corresponding maximum contact stress is about 930 MPa), a pressure of 1.0 × 10⁻⁴ Pa, and a sliding speed of 0.18 m s⁻¹. It was assumed that the tested a-C:H films failed when the value of the friction coefficient exceeded 0.15. Friction experiments were conducted at least three times under same experimental conditions. Then, the tribological performance of the film after different test durations was investigated. The test duration was selected as 300 cycles, 700 cycles, 1000 cycles and 2400 cycles. After the friction and wear tests, the morphology of the worn surfaces was observed by SEM and non-contact three dimensional surface profilometer (ADE Corporation, Massachusetts, USA). The structural information and phase compositions of the worn surfaces of both the film and corresponding steel counterface were characterized by TEM and Raman spectrometry.

3. Results and discussion

3.1 General properties of the as-prepared a-C:H film

Fig. 1 shows the FESEM and TEM morphologies of the as-prepared a-C:H film. As seen in Fig. 1a, the film shows a dense microstructure and well adheres to the Si wafer substrate. The total thickness of the film is about 1.716 μm, including 1.435 μm of a-C:H film and 0.281 μm of pure silicon interlayer. According to the TEM image and corresponding electron diffraction pattern in Fig. 1b, it can be found that the a-C:H film exhibits an amorphous structure. In addition, the as-deposited film possesses a relatively high hardness of 9.17 GPa and Young's modulus of 66.14 GPa.

Fig. 1 FESEM image of fractured cross-section (a) and TEM micrograph (b) of the as-prepared a-C:H film.

3.2 Friction and wear behavior of the as-prepared a-C:H film in a high vacuum environment

The friction coefficient curves of the a-C:H film under a vacuum environment are shown in Fig. 2. It clearly reveals that the friction test results are fairly reproducible. The variation of the friction coefficient with the sliding cycles can be approximately divided into three stages. Namely, at the initial stage, the friction coefficient sharply declines to a stable low value (0.01) after a short period of running-in and remains stable up to about 350 cycles. Then it undergoes an evident fluctuation period in the sliding cycles of 350–750. At the third stage (referring to the sliding cycles of 750–2600), the friction coefficient is further decreased to an extremely low value (0.005) and tends to be stabilized thereat, followed by an abrupt rise above the sliding cycles of 2700 cycles, at which failure of the a-C:H film starts. In addition, this similar variation of the friction coefficient has been reported in previous reports;^16,17,20,21 however, detailed research on this peculiar friction phenomenon has never been reported.

Fig. 2 Friction coefficient curves of a-C:H film under a vacuum environment (inset is the schematic diagram of friction model).

3.3 SEM observation of the worn surfaces of the a-C:H film and steel ball counterface

To acquire more insight into the correlation between the friction behavior of the a-C:H film and the sliding cycles, we conducted friction tests at durations of 300 cycles in the initial stable period: 700 cycles corresponding to the evident fluctuation period, 1000 cycles at the superlow friction stage after the fluctuation period, as well as 2400 cycles that approaches to failure. The corresponding SEM images of the worn surfaces of the a-C:H film as well as the steel ball counterface after different friction durations are shown in Fig. 3 to 6.

Fig. 3 SEM images of the worn surfaces of a-C:H film (a) and steel ball counterface (b) after 300 cycles.

It can be seen that the wear scar of the a-C:H film after 300 cycles of sliding is shallow and smooth without any evident signs of damage (Fig. 3a), which indicates that the film underwent very slight wear in the early stage of friction. Moreover, a continuous and thin transfer layer seems to have formed on the surface of the counterface steel ball (Fig. 3b) in this case, which well corresponds to the low and stable friction coefficient therewith.

When the sliding cycles is extended to 700 cycles, evident signs of scuffing in association with a few wear debris are visible on the worn surface of the a-C:H film. Some sub-layer films seem to have been newly exposed on the wear scar owing to mild adhesion wear (Fig. 4a and b), whereas an increased amount of film material seem to be detached and transferred onto the worn surface of the counterface steel (Fig. 4c). These observations correspond well to the evident fluctuation of the friction coefficient in the sliding duration of 350–750 cycles.

Fig. 4 SEM images of the worn surfaces of a-C:H film (a, b) and steel ball counterface (c) after 700 cycles.

As the sliding process further increases to 1000 cycles, the corresponding worn surface of the a-C:H film shows signs of severe adhesion and scuffing as well as evident spalling in association with an increased amount of wear debris (Fig. 5a and b), whereas more film materials seem to have been transferred onto the worn surface of the steel counterface, but no compact and continuous transfer film is formed on the sliding surface (Fig. 5c). Such worn surface features seem to be contradictory to the very low and stable friction coefficient in the sliding cycles range of 750–1000, and the reasons accounting for this phenomenon will be further discussed in the following section.

Fig. 5 SEM images of the worn surfaces of a-C:H film (a and b) and steel ball counterface (c) after 1000 cycles.

When the sliding cycle reaches 2400 cycles, it seems that the worn surface of the a-C:H film is covered by a certain material (Fig. 6a and b). The spalling and scuffed grooves become more prominent after the worn film surface is ultrasonically cleaned (Fig. 6c). In the meantime, as seen in Fig. 6d, the wear track of the steel ball is notably enlarged and covered by a discontinuous transfer film. This is seemingly also contradictory to the quite low and stable friction coefficient in this period.

Fig. 6 SEM images of the worn surfaces of a-C:H film (a–c) and steel ball counterface (d) after 2400 cycles.

3.4 Three-dimensional morphology analysis of the worn surface of a-C:H film

The non-contact 3D surface morphologies of the a-C:H film after sliding against the steel counterface at different cycles are shown in Fig. 7. It can be seen that both the width and depth of the wear scar increase with extending sliding cycles. After 300 cycles of sliding, the wear scar of the a-C:H film is shallow and shows few signs of plough grooves with a wear depth of about 630 nm (Fig. 7a), whereas 700 cycles of sliding gives rise to a deeper (900 nm in depth) and wider wear scar, which is dominated by mild adhesion and scuffing (Fig. 7b). As the sliding cycles further increases to 1000 cycles and 2400 cycles, the wear scar of the a-C:H film is further widened and more severe adhesion and scuffing are observed (Fig. 7c and d).

Fig. 7 Three-dimensional surface morphologies of a-C:H film after sliding against a steel counterface at 300 cycles (a), 700 cycles (b), 1000 cycles (c), and 2400 cycles (d).

3.5 Raman spectrometric analysis of the worn surface of a-C:H film and counterface steel ball

Fig. 8 shows Raman spectra of the as-prepared film and the worn surfaces of the a-C:H film after sliding at different cycles. All of the worn surfaces of the a-C:H film in different durations show a sharp G peak at about 1550 cm⁻¹ and a D peak at about 1380 cm⁻¹ (Fig. 8), which is consistent with the typical structural feature of diamond-like carbon coatings.²² Interestingly, along with the increase of the sliding cycles, the G peaks of the worn surfaces of the a-C:H film tend to gradually shift towards the higher frequency region, whereas the intensity of the D peaks tends to rise therewith (Table 1). This indicates that the extended sliding cycles helps to enhance the structural change of the film on the frictional interface,^23,24 which is related to the variation of the friction coefficient at the extended sliding cycles of 750 cycles and above.

Fig. 8 Raman spectra of the as-prepared a-C:H film and the worn surfaces of a-C:H film after different sliding cycles.

Table 1 The fitting results of the Raman spectra in Fig. 8

Investigated spot	D peak position (cm^−1)	G peak position (cm^−1)	ID/IG
Original film	1364.4	1531.9	1.75
After 300 cycles	1368.3	1536.1	1.8
After 700 cycles	1377.8	1545.6	1.92
After 1000 cycles	1392.1	1555.6	2.02
After 2400 cycles	1406.0	1557.1	2.43

---

3.6 TEM analyses of the worn surfaces of a-C:H film at sliding cycles of 700, 1000 and 2400

As reported elsewhere, the extension of the sliding cycles causes severe adhesion wear on the sliding surface, thereby finally leading to the failure of the a-C:H film.^25,26 In the present research, severe adhesion wear, scuffing and even the film peeling off are also observed when the sliding cycles are extended to 1000 cycles and 2400 cycles, but the friction coefficient is lowered to a very low level (about 0.005) and stabilized there after an evident fluctuation in the sliding cycle range of 350–750. To reveal the possible reasons leading to such an anomalous change in the friction coefficient of the a-C:H film, we conducted TEM analyses of the worn surfaces of the a-C:H film after sliding cycles of 700, 1000 and 2400. TEM images of the worn surface after sliding for 700 cycles are shown in Fig. 9a and b. Some regular particles at the nanoscale are clearly visible on the worn surface after 700 cycles (Fig. 9a). Moreover, the magnified TEM image in Fig. 9b further reveals that these particles mainly show an amorphous structure, which is very similar to the original film. However, some lamellar structures were formed at the edge of these particles and the distance between the planes is approximately 0.34 nm, which closely matched the lattice parameter d₀₀₂ of bulk graphite.²⁷ After sliding for 1000 cycles, it is worth noting that some highly ordered spherical nanoparticles with a size of 25–35 nm emerge on the frictional interface of the a-C:H film (Fig. 9c). The magnified image in Fig. 9d further indicates that the ordered spherical nanoparticles consist of onion-like carbon, which is comprised of concentric graphitic shells²⁸ and the distance between the graphite shells is approximately 0.34 nm. When the sliding cycles reach 2400, similar spherical nanoparticles are also left on the wear scar (Fig. 9e). Nevertheless, as shown in Fig. 9f, the integrity of the onion-like carbon structure after 2400 cycles is destroyed to some extent and particularly its outer layers are peeled off evidently, which is possibly damaged by the repeated shear force during the adequately extended sliding cycles. Therefore, we suppose that the graphite nanosheets consisting of onion-like carbon, generated via increased sp² bonding during the extended sliding cycles, accounts for the extremely low friction coefficient of the a-C:H film during the extended sliding cycles of 1000 and above (up to about 2400 cycles).

Fig. 9 TEM images of the worn surfaces of a-C:H film after sliding against the steel ball counterface after 700 cycles (a and b), 1000 cycles (c and d) and 2400 cycles (e and f).

3.7 Friction mechanism of a-C:H film in a vacuum environment

According to the abovementioned analyses, the dynamic friction mechanism of the a-C:H film in different friction periods under a vacuum environment can be schematically illustrated, as shown in Fig. 10. As shown in Fig. 10a, at the first stage (within 300 cycles), the initial contact mainly occurs between a-C:H film and steel ball. During this process, repetitious sliding between the two contact surfaces leads to the transfer of the film material onto the steel ball counterface, thereby generating a continuous ultrathin transfer film thereon in the early stage of sliding. Along with the formation of the continuous transfer film, the original film-steel contact is transformed to a-C:H film/a-C:H film contact. As a result, the friction coefficient is significantly reduced owing to the strong hydrogen-passivation between the sliding interfaces.^29,30

Fig. 10 Schematic diagram illustrating the friction and wear process of a-C:H film in a vacuum environment.

As the friction process increases to 700 cycles (Fig. 10b), the local hydrogen bonding of the a-C:H film is partly destroyed under the long-time repeated shear force,³¹ thereby exposing the active bonds at the contact interface where mild adhesion wear occurs and then the local weld zone is sheared off.²⁵ In addition, as we all know, a-C:H films are very brittle and possess high internal stress; therefore, the wear debris are easily produced during the adhesive process. At the same time, the friction coefficient began to be a little higher and unstable. As a result, the adhesion process as well as the alternate appearance of a worn film and fresh film leads to the evident fluctuation of the friction coefficient therewith.³²

Following the fluctuation process, more severe adhesion wear occurs and then local weld zone is sheared off at the further extended sliding cycles (Fig. 10c), and the friction coefficient is even lowered to an extremely low level (0.005), which is due to the formation of a well-ordered onion-like carbon structure on the sliding surface (Fig. 10d). The formation of such a special carbon structure on the frictional interface involves complex factors such as mechanical excitation, frictional heating, and even polymerization and oxidation.³² In brief, the constant breaking of the hydrogen bonding on the worn surface leads to the generation of two-dimensional graphite nanosheets, which can further transform into an onion-like carbon structure.³³ Bollmann et al. reported that the rolling up of graphite nanosheets at the tribological interface can cause them to act like a roller bearing that not only decreases surface energy but also forms a superlubricity friction interface.³⁴ Thus, we can reasonably infer that it is the onion-like carbon structure on the sliding surface that accounts for the significantly reduced friction coefficient during the extended sliding cycles. On the one hand, the onion-like carbon nanoparticles can serve as spacers to prevent direct rough contact between the rubbed surfaces. On the other hand, the onion-like carbon structures can act as nano-scale ball bearings to transform the sliding friction to rolling friction to some extent, thereby contributing to the reduced friction. In addition, there are no dangling bonds at the surface of the onion-like carbon structure that are subjected to ordering under the extended sliding cycles of 1000 cycles, and hence, weak intermolecular bonding with the counterface material might also account for the reduced friction coefficient to some extent.³⁵

However, when the sliding cycles is further extended to a high enough level, the onion-like carbon structure will be gradually exfoliated and destroyed in the contact area under the repeated shear force. Namely, the outer layers of the onion-like carbon structure will be peeled off to generate many graphite fragments. As a result, the direct contact between the counterface steel ball and the graphite fragments leads to the rapid deterioration of the wear condition in association with the abrupt rise of the friction coefficient and ultimate failure of the a-C:H film.

4. Conclusions

The friction and wear behavior of an as-prepared a-C:H film on a stainless steel substrate were evaluated in a vacuum environment, and the correlation between the tribological behavior of the a-C:H film and the sliding cycles was systematically investigated. The dynamic friction mechanism of the a-C:H film throughout the entire friction process is successfully established. It is worth noting that carbon onions with a closed spherical shell structure were spontaneously formed on the worn surface in the absence of the hydrogen passivation effect, which further reduced the friction coefficient of the film to an extremely low value. This study provides guidance for the further designing of a new generation of a-C:H films with a special structure that exhibits better tribological performance in a vacuum environment.

Acknowledgements

The authors are grateful to the Ministry of Science and Technology of China (“973” project, grant no. 2013CB632300), the National Natural Science Foundation of China (Grant no. 51275509 and 51405474) and the Chinese Academy of Sciences for financial support.

References

1. M. Tagawa, M. Muromoto, S. Hachiue, K. Yokota, N. Ohmae, K. Matsumoto and M. Suzuki, Tribol. Lett., 2005, 18, 437.
2. X. Liu, J. Yang, J. Hao, J. Zheng, Q. Gong and W. Liu, Adv. Mater., 2012, 24, 4614.
3. S. Bai, H. Murabayashi, Y. Kobayashi, Y. Higuchi, N. Ozawa, K. Adachi, J. Michel Martin and M. Kubo, RSC Adv., 2014, 4, 33739.
4. K. Enke, H. Dimigen and H. Hübsch, Appl. Phys. Lett., 1980, 36, 291.
5. M. Kalin and J. Vižintin, Wear, 2005, 259, 1270.
6. A. Erdemir, O. L. Eryilmaz, I. B. Nilufer and G. R. Fenske, Surf. Coat. Technol., 2000, 133, 448.
7. K. Miyoshi, B. Pohlchuck, K. W. Street, J. S. Zabinski, J. H. Sanders and A. Voevodin, Wear, 1999, 65, 225.
8. K. Vercammen, J. Meneve, E. Dekempeneer, J. Smeets, E. W. Roberts and M. J. Eiden, Surf. Coat. Technol., 1999, 120–121, 612.
9. C. Donnet, J. Fontaine, A. Grill and T. L. Mogne, Tribol. Lett., 2000, 9, 137.
10. L. Cui, Z. Lu and L. Wang, ACS Appl. Mater. Interfaces, 2013, 5, 5889.
11. D. W. Kim and K. W. Kim, Wear, 2013, 297, 722.
12. J. Fontaine, Proc. Inst. Mech. Eng., Part J, 2008, 222, 1015.
13. O. L. Eryilmaz and A. Erdemir, Surf. Coat. Technol., 2007, 201, 7401.
14. F. Gao, A. Erdemir and W. T. Tysoe, Tribol. Lett., 2005, 20, 221.
15. A. Vanhulsel, F. Velasco, R. Jacobs, L. Eersels, D. Havermans, E. W. Roberts, I. Sherrington, M. J. Anderson and L. Gaillard, Tribol. Int., 2007, 40, 1186.
16. Y. Wu, H. X. Li, L. Ji, Y. Ye, J. Chen and H. Zhou, Tribol. Int., 2014, 71, 82.
17. H. Song, L. Ji, H. Li, X. Liu, H. Zhou, L. Liu and J. Chen, J. Phys. D: Appl. Phys., 2014, 47, 235301.
18. J. Fontaine, Proc. Inst. Mech. Eng., Part J, 2013, 227, 729.
19. X. Fan and L. Wang, ACS Appl. Mater. Interfaces, 2014, 6, 14660.
20. A. Erdemir and C. Donnet, J. Phys. D: Appl. Phys., 2006, 39, R311.
21. A. Erdemir, J. Fontaine and C. Donnet, in Tribology of Diamondlike Carbon Films: Fundamentals and Applications, ed. C. Donnet and A. Erdemir, Springer, New York, 2008, p. 246.
22. A. Grill, Diamond Relat. Mater., 1999, 8, 428.
23. T. Ma, L. Wang, Y. Hu, X. Li and H. Wang, Sci. Rep., 2014, 4, 3662.
24. T. Ma, Y. Hu, L. Xu, L. Wang and H. Wang, Chem. Phys. Lett., 2011, 514, 325.
25. J. Fontaine, T. Le Mogne, J. L. Loubet and M. Belin, Thin Solid Films, 2005, 482, 99.
26. S. Miyake, S. Takahashi, I. Watanabe and H. Yoshihara, ASLE Trans., 1987, 30, 121.
27. H. S. Myung, Y. S. Park, B. Hong, J. G. Han, Y. H. Kim, J. Y. Lee and L. R. Shaginyan, Thin Solid Films, 2006, 494, 123.
28. W. Choi, I. Lahiri, R. Seelaboyina and Y. Kang, Crit. Rev. Solid State Mater. Sci., 2010, 35, 52.
29. I. Sugimoto and S. Miyake, Appl. Phys. Lett., 1990, 56, 1868.
30. M. Kalin and J. Vižintin, Tribol. Int., 2006, 39, 1060.
31. A. R. Nyaiesh and W. B. Nowak, J. Vac. Sci. Technol., 1983, 308, A1.
32. K. Y. Eun, K. R. Lee, E. S. Yoon and H. Kong, Surf. Coat. Technol., 1996, 86–87, 569.
33. C. Meunier, P. Alers and L. Marot, Surf. Coat. Technol., 2005, 200, 1976.
34. W. Bollmann and J. Spreadborough, Nature, 1960, 186(2), 29.
35. A. Htsushi, M. Igarashi and T. Kaito, Tribol. Int., 2004, 37, 899.

---