- Email: [email protected]
Low friction mechanism of chlorine-doped amorphous carbon films sliding against an aluminium alloy
Tribology International 115 (2017) 573–579
Contents lists available at ScienceDirect
Tribology International journal homepage: www.elsevier.com/locate/triboint
Low friction mechanism of chlorine-doped amorphous carbon films sliding against an aluminium alloy Yuuki Tokuta a, *, Takashi Itoh b, Takahiko Shiozaki b, Masahiro Kawaguchi a, Shinya Sasaki c a b c
Tokyo Metropolitan Industrial Technology Research Institute, 2-5-10 Aomi, Koto-ku, Tokyo 135-0064, Japan Fujimetal Co. Ltd., 2-7-16 Hinode, Kawasaki-ku Kawasaki-shi, Kanagawa 210-0824, Japan Tokyo University of Science, 6-3-1 Niijuku, Karsushika-ku, Tokyo 125-8585, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords: Coating Self-lubricating Diamond-like
The effect of chlorine-doping on the tribological properties of amorphous carbon films was investigated. Chlorinedoped amorphous carbon films were deposited using a plasma-based ion implantation and deposition (PBII&D) method with vaporized tetrachloroethylene as a precursor. During sliding tests with aluminium alloy counter parts, chlorine-doped amorphous carbon films showed lower friction coefficients compared to hydrogenated amorphous carbon films. Both aluminium and a chlorine-based hydrate were observed on the wear track of the carbon films. This hydrate was a tribofilm formed via a tribochemical reaction during sliding between the chlorine-doped amorphous carbon films and aluminium alloy. The viscosity of this tribofilm was similar to polyalpha olefin and we propose that it acted as a lubricant and lowered the friction coefficient.
1. Introduction Amorphous carbon films are generally composed of sp2-and sp3-hybridized orbitals of carbon and hydrogen. In recent years, amorphous carbon films have attracted attention in research and industry due to their superior properties, such as optical properties, corrosion resistance, and biocompatibility [1–3]. In particular, amorphous carbon films have extraordinary wear and friction properties, and they have contributed to drastic improvement of the tribological properties of gears, cutting tools, and machine parts [4–6]. Since the seminal report regarding amorphous carbon films was published by Aisenberg and Chabot in 1971 [7], this material has attracted much attention from many research groups. As research on amorphous carbon films has progressed, it has been confirmed that various film properties depend on the structure. Grill proposed that the characteristics and structure of amorphous carbon films are determined by several factors, including the concentrations of hydrogen, sp3 bonding, and sp2 bonding in the films. In addition, amorphous carbon films can be classified into four types: amorphous carbon (a-C), hydrogenated amorphous carbon (a-C:H), tetrahedral amorphous carbon (taC), and hydrogenated tetrahedral amorphous carbon (ta-C:H) [8]. The aC:H and ta-C:H films contain up to approximately 50% hydrogen, a-C and ta-C films contain less than 1% hydrogen, and ta-C and ta-C:H consist of a high fraction of sp3 bonding. Hence, the term “amorphous carbon films”
* Corresponding author. E-mail address: [email protected] (Y. Tokuta). http://dx.doi.org/10.1016/j.triboint.2017.06.024 Received 27 March 2017; Received in revised form 31 May 2017; Accepted 16 June 2017 Available online 19 June 2017 0301-679X/© 2017 Elsevier Ltd. All rights reserved.
covers a wide range of carbon-based coatings and the investigation of the relationship between the structure and properties of the films is critical in their application. Erdemir reported that the hydrogen content of a-C:H films plays an important role in the friction behaviour [9]. a-C:H films contain a large amount of hydrogen, where the terminated hydrogen eliminates most of the dangling σ bonds and π-π interactions and contributes to superlow friction under dry conditions. The low friction of a-C:H films was attributed to the high chemical inertness of films as a result of the surface passivation of the dangling bonds. On the contrary, ta-C films are suitable for oil-lubricated application because of their unterminated dangling bonds inside the films which activate the wear surface. A tribochemical layer caused by friction has been observed to form on the surface of such films [10]. The friction behaviour of these films is strongly dependent on the sliding environment. Hence, it is clear that the optimisation of the film structure for a particular application is critical. Recent studies have demonstrated doping of amorphous carbon films with different elements (such as Si, F, B, N, Al Cr, Ti, and W) as a new technique for optimising the structure [11–17]. These efforts contributed to the understanding of mechanisms for improving many properties, for example thermal resistance, optical gap energy, surface energy, relaxation of internal stress, and activation of the tribochemical reaction. Si doping of amorphous carbon films has been shown to improve the high temperature behaviour in comparison to hydrogenated amorphous
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
carbon films and also achieve low friction under high humidity due to the formation of SiO2 particles through a tribochemical reaction [11,12]. In the case of F doping, the surface energy was shown to decrease and the optical band gap improve with increasing F content; the increase in the optical band gap energy induced by F doping was attributed to a modified film structure with a higher fraction of sp3 bonding compared to nondoped films [13,14]. Doping with metals, such as Al and Ti, has also been shown to contribute to a reduction in the surface energy, as observed for Si doping [15]. The dopant metal atoms inside the films form metal-O covalent bonds, localizing the electrons of the metal; thus, a decrease in dipoles contributes to reducing the surface energy. In addition, metal doping contributes to reducing internal stress and activating the chemical reaction between amorphous carbon films and lubricant additives [16,17]. The aim of this study is to investigate the effect of chlorine doping on the tribological properties of a-C:H films during sliding against an aluminium alloy under non-lubricated conditions. Our hypothesis is that the chlorine inside of the films will contribute to the formation of a chlorinated tribofilm at the sliding interface, in the same way as a reaction with chlorine from additives [18–20], where the chlorinated tribofilm can reduce friction coefficient. In our previous study, we observed a lowering of the friction due to chlorine-doped a-C:H films during sliding with aluminium alloy counter parts [21]. However, the exact mechanism behind this observed low friction was not clarified and hence, the major objective of this study is to reveal the low friction mechanism of such films. Non-doped a-C:H films and chlorine-doped a-C:H films were deposited using a plasma-based ion implantation and deposition (PBII&D) method using vaporized toluene (C6H5CH3) and tetrachloroethylene (C2Cl4). Various characterisation methods, including micro-laser Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), Elastic recoil detection analysis (ERDA), and nano-indentation hardness testing were employed to investigate the structure and mechanical properties. The chlorine-doped tribofilm formed through a tribochemical reaction was investigated using surface analysis techniques including Fourier transform infrared spectroscopy (FT-IR) and time-of-flight secondary ion mass spectroscopy (TOF-SIMS).
Fig. 1. Schematic diagram of the plasma-based ion implantation and deposition system used in this study.
Table 1 Deposition parameters of non-doped and chlorine-doped a-C:H films.
Non-doped a-C:H films Chlorine-doped a-C:H films
2. Experimental setup
#1 #2 #3 #4 #5
Cl/(C þ Cl) ratio
Hydrogen content [atm %]
0.0
21.9
0.013 0.037 0.083 0.108 0.118
20.3 10.1 4.8 3.1 <3.0
deposition parameters, such as the applied voltage of 5 kV and RF power of 500 W were the same for all samples. In Table 1, the hydrogen content and the presence ratio of chlorine relative to the total number of carbon and chlorine Cl/(C þ Cl) of each film were measured using ERDA and XPS, respectively. ERDA was used determine the hydrogen content inside of each film at a depth of 50 nm from the film surface. XPS was used to measure the Cl/(C þ Cl), where the Cl/(C þ Cl) was estimated from the relative areas of the chlorine peaks (Cl 2p) and the carbon peaks (C 1s). The surface roughness parameter Ra of these amorphous carbon film were about 0.022–0.026 μm and film thickness was above 1.0 μm.
2.1. Deposition of amorphous carbon films using PBII&D A PBII&D system (PBII-R1000/Kurita Seisakusho) was employed to deposit the non-doped a-C:H films and chlorine-doped a-C:H films on an aluminium alloy (ISO-AlMg1SiCu) disk with a diameter of 30 mm. Nano-indentation hardness of substrate was 1.7 GPa. Fig. 1 shows a schematic diagram of the PBII&D system used in this study. In this method, the substrate itself is used as an RF antenna and a highfrequency power supply generates a plasma from the source gas around the substrate; then a negative charge is applied to the substrate. Using this process, plasma ions can be implanted into the substrate surface to form the desired film. This method has various advantages; for example, it is possible to form a uniform film on a complex surface at deposition temperatures of less than 473 K. Note that the films were deposited via a hybrid process of simultaneous chemical reaction and ion implantation, which was accelerated by the applied voltage. Prior to film deposition, all substrates were cleaned by Ar bombardment to remove contaminants and activate the surface. After Ar bombardment, Si interlayer was deposited using hexamethyldisiloxane (HMDSO) gas on the substrate surface. Table 1 shows the different chlorine and hydrogen contents of five types of samples prepared in this study. A non-doped a-C:H films were prepared using only toluene gas, while the chlorine-doped a-C:H films were deposited using two different gases, a mixture of toluene and tetrachloroethylene (C6H5CH3 þ C2Cl4; #1, #2, #3, and #4) or pure C2Cl4 (#5). When using the mixed gas, the flow ratio of C2Cl4 to C6H5CH3 was increased in order of #1 to #4; the other
2.2. Structure and mechanical properties Raman spectroscopy and XPS analyses were used to investigate the structure of the films. Specifically, the effect of chlorine-doping on the internal structure of the films was analysed by Raman spectroscopy and the effect on the sp3/sp2 ratio of the films was calculated from XPS data. The Raman spectrometer (In Via Reflex/Renishaw) was equipped with a YAG laser with a wavelength of 532 nm. The laser power, spot size, and spectral range were 0.25 W, 5.0 μm, and 800–2000 cm1, respectively. An XPS apparatus (Quantera/Ulvac-Phi) was used to analyse chlorine in the films and measurements were undertaken after a sputtering pretreatment using Arþ. The detection range of the carbon and chlorine peaks were 275–300 eV and 190–220 eV, respectively. The X-ray source, X-ray power, and target emission were AlKα, 10kV25W, and 3.0 mA. The sp3/(sp3 þ sp2) bonding ratio of carbon was investigated by analysing the 574
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
3. Results and discussion
XPS spectra using the method described by Riedo et al. [22]. In this method, sp2 bonding peak and sp3 bonding peak were defined at peak position of 284.5 eV and 285.3 eV respectively. The sp3/(sp3 þ sp2) were calculated from peak intensity ratio of these peaks. The peak fitting was achieved using Gaussian fitting software. A nano-indentation hardness tester (TI-950/Hysitron) with a Berkovich diamond tip was used to compare the film hardness and Young's modulus of the non-doped a-C:H films and chlorine-doped a-C:H films. The indentation speed was 5 nm/s and a maximum indentation depth of 100 nm was used to minimize the contribution from the substrate. An indenter calibration was achieved by Oliver-Parr method [23] The film hardness and Young's modulus were estimated by fitting the loaddisplacement curves obtained during indentation.
3.1. Structure and hardness of amorphous carbon films Fig. 2 shows Raman spectra of both of non-doped a-C:H (Cl/(C þ Cl) of 0.0) and chlorine-doped (Cl/(C þ Cl) of 0.013, 0.036, 0.083, 0.107, and 0.118) a-C:H films. In the Raman spectra of the amorphous carbon films, the D peak around 1350 cm1 (attributed to a disorder structure) and the G peak around 1550 cm1 (attributed to the graphite structure) were observed. The D peak is related to the breathing modes of sp2 atoms in rings, whereas the G peak is related to the bond stretching of all pairs of sp2 atoms in rings and chains [24–26]. The peak positions, the fullwidth-at-half-maximum (FWHM) of the peaks, and the intensity ratio of the D peak to the G peak (ID/IG) provide important information about the film structure. It can be seen in this figure that, while both the D and G peaks were observed for all films, the features of the Raman spectra were different. An increasing Cl/(C þ Cl) ratio resulted in a slight increase in the D peak intensity, indicating a structural change. In general, as the graphite-like structure formed inside the films, the D and G peaks tended to separate due to the increase in the intensity of the D peak. The evolution of a distinct D peak indicated the formation of a graphitic domain [27]. This result suggests that the a-C:H film structure was dependent on the precursor gas and a graphite-like structure was formed with increasing Cl/(C þ Cl) ratio. Fig. 3 shows the position of the G peak and its FWHM value as functions of the Cl/(C þ Cl) ratio of the films. It can be seen that, as Cl/(C þ Cl) increased, the G peak shifted toward a higher wavenumber and the FWHM(G) decreased. Both of these tendencies suggest an ordering of the film structure [28]. The change in ID/ IG is also important for understanding the film structure; Fig. 4 compares the ID/IG ratio measured using Raman spectroscopy with the sp3/ (sp2 þ sp3) ratio determined by XPS analysis. It can be seen that the ID/IG ratio was inversely proportional to Cl/(C þ Cl), attributed to the formation of the graphite-like structure (consistent with the shift in the G peak position and decrease in FWHM(G)) [29]. This was also confirmed by the decrease in the sp3/(sp2 þ sp3) ratio. These results suggested that a-C:H film deposition using chlorine-containing precursor gases contributed to the formation of a graphite-like structure with a high sp2 fraction. Fig. 5 shows the film hardness and Young's modulus as functions of Cl/(C þ Cl) ratio in the films. A reduction in the film hardness and Young's modulus was observed with increasing Cl/(C þ Cl) ratio. It is well known that the increased sp2 fraction decreases the film hardness and Young's modulus due to a loss of the diamond-like characteristics of the film [30]. Fig. 6 shows ID/IG ratio of Raman spectrum and sp3/ (sp2 þ sp3) bonding ratio calculated from XPS spectrum as function of films hardness. From this result, it was confirmed that the films hardness
2.3. Sliding tests and observation of wear tracks A reciprocating-type ball-on-disk sliding tester (Tribometer/CSM) was employed to perform sliding tests on each film. An aluminium alloy ball (ISO-AlMg1SiCu) with a diameter of 6 mm was used as the counter part material to slide against the disks with the deposited films. The nano-indentation hardness and surface roughness Ra of aluminium alloy used in this study were 1.8 GPa and 0.059 μm, respectively. The applied load, sliding speed, sliding distance, and total number of sliding cycles were 5 N, 20 mm/s, 10 mm, and 2000 cycles, respectively. The sliding tests were performed under non-lubricated conditions in air at a temperature of 20 ± 2 C and a relative humidity of 50 ± 5%. An optical microscope (VHX-500/Keyence) and a confocal laser-scanning microscope (OLS-4000/Olympus) were employed to observe the wear track and calculate the wear area to investigate the friction mechanism and wear properties. 2.4. Surface analysis after sliding tests FT-IR (FTIR-6600/Jasco) and TOF-SIMS (TRIFT/Ulvac-Phi) were used to investigate the presence of compounds along the wear track of the chlorine-doped a-C:H films. In the FT-IR analyses, the measurement range was set to 1000–4000 cm1 in microscopic mode. For TOF-SIMS measurements, we analysed in mass resolution mode using Gaþ ions as the ion source. Spot size and current density were 100 μm square and 1.0 nA in DC conversion. The negative fragment ions detected from the wear track were investigated to determine the type of tribofilm at the sliding interface.
Fig. 3. G peak position and FWHM(G) as functions of the Cl/(C þ Cl) ratio in the aC:H films.
Fig. 2. Raman spectra of the a-C:H films with different Cl/(C þ Cl) ratio. 575
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
Fig. 4. ID/IG ratio and sp3/(sp2 þ sp3) ratio as functions of the Cl/(C þ Cl) ratio in the aC:H films. Fig. 7. Average friction coefficient as a function of Cl/(C þ Cl) ratio.
increasing Cl/(C þ Cl) ratio. The friction coefficient of the non-doped aC:H film was about 0.107 and the chlorine-doped a-C:H films with Cl/ (C þ Cl) ratio of 0.013, 0.037, 0.083, 0.108, and 0.118 showed values of 0.070, 0.066, 0.060, 0.059, and 0.042, respectively. Of particular note, even when the Cl/(C þ Cl) ratio was very small (0.013 for #1), the friction coefficient reduced by approximately 40% compared with the non-doped a-C:H film. The small error bars on the data shows repeatability in five tests. These results suggest that chlorine doping can reduce the friction coefficient of a-C:H films. Fig. 8 shows the wear areas measured for the films and counter parts after sliding tests. There was a threshold in the correlation between the Cl/(C þ Cl) and the wear area; when the Cl/(C þ Cl) was below 0.083, the wear areas of the disk and ball were less than that of the non-doped aC:H film; however when the Cl/(C þ Cl) was 0.108, the wear properties deteriorated and the wear area increased significantly. In the range of Cl/ (C þ Cl) of 0.013–0.083, the chlorine-doped a-C:H films showed improved wear resistance and low aggressiveness against the counter parts, despite the lower hardness compared to that of the non-doped aC:H film. These results suggest that the non-doped a-C:H and chlorinedoped a-C:H films have different friction mechanisms. Fig. 9 shows a surface image of the wear track on the aluminium alloy ball observed after the sliding test against chlorine-doped a-C:H film (#1), where the presence of a liquid substance adhered to the wear track was observed. This liquid material is thought to be a tribofilm, which contributed to the low friction, high wear resistance, and low aggressiveness against the counter parts.
Fig. 5. Film hardness and Young's modulus as functions of the Cl/(C þ Cl) ratio in the aC:H films.
Fig. 6. ID/IG ratio and sp3/(sp2 þ sp3) ratio as functions of the film hardness.
indicated lower value with less sp3 bonding. Hence, the tendencies of the results of the nano-indentation hardness tests were in a good agreement with the results of Raman spectroscopy and XPS. 3.2. Friction behaviour of amorphous carbon films Fig. 8. Wear areas of the ball and disk as functions of Cl/(C þ Cl) ratio.
Fig. 7 shows that the average friction coefficient decreased with 576
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
Fig. 10. FT-IR measurements of the wear track of the aluminium alloy ball sliding on (a) chlorine-doped a-C:H films and (b) non-doped a-C:H films.
Fig. 9. Optical microscopy image of the wear track of sample #1 after the sliding test where the liquid regions are indicated.
3.3. Surface analysis of wear tracks Fig. 10 show FT-IR spectra of the wear track of the aluminium alloy ball after the sliding tests against non-doped a-C:H films and chlorinedoped a-C:H films (#1). The wear track of the ball in friction with chlorine-doped a-C:H film (Fig. 10(a)) showed two peaks, at 3400 cm1 and 1560 cm1, while no significant peaks were observed in the case of the non-doped a-C:H film (Fig. 10(b)). One possibility is that these characteristic peaks indicate the presence of water molecules or hydroxy. In the FT-IR analysis of gaseous water, the symmetric OH stretching vibration (v1) and asymmetric OH stretching vibration (v3) are centred at 3656 and 3755 cm1, respectively, while the bending mode of σ(H2O) is observed around 1594 cm1. In the case of the liquid water, the band consists of broadened v1 and v2 peaks due to the formation of the hydrogen bond in the liquid, and these peaks cannot be distinguished. These deformations in the stretching mode result in the peak position shifting approximately 200–400 cm1 [31]. These results suggest that hydrate and/or water adsorption occurred at the wear track during sliding through a tribochemical reaction. Fig. 11 shows the negative fragment ion peaks detected from the wear track of chlorine-doped a-C:H film (#1) by TOF-SIMS analysis. These results show information about the detected negative ions, obtained by subtracting the spectrum recorded outside of the wear track from that obtained along the wear track. It can be seen that several kinds of particular fragment ion peaks were observed, such as H2OCl, H2O2Al, HAlCl, HOAlCl, and H2OAlCl, which suggest the presence of a chlorinealuminium based hydrate.
Fig. 11. TOF-SIMS analysis of the wear track of chlorine-doped a-C:H film #1.
water molecules from the atmosphere. From this standpoint, XPS analysis was carried out on an aluminium chloride hexahydrate reagent and the wear track of the aluminium ball after a sliding test with a chlorine-doped a-C:H film. Fig. 13 shows the Cl 2p peaks from the XPS analysis for the (a) surface of a chlorine-doped aC:H film, (b) wear track of the aluminium alloy ball after sliding tests with chlorine-doped a-C:H film, and (c) aluminium chloride hexahydrate reagent. Obvious differences in the Cl 2p peak position between the chlorine-doped a-C:H film surface and the other samples were observed. Of particular note, the peak position and FWHM of the Cl 2p peak detected from the wear track of the aluminium alloy ball were similar to that of the aluminium chloride hexahydrate reagent. Hence, it is possible that the composition of the tribofilm formed through tribochemical reaction by the friction between chlorine-doped a-C:H films and the aluminium alloy was aluminium chloride hydrate (such as hexahydrate). As aluminium chloride hydrate is deliquescent, it is possible that the liquid seen on the wear track of the aluminium alloy after the sliding test was the liquefied form of this material. In order to investigate the effect of the mechanical behaviour of aluminium chloride hexahydrate on the friction, we measured the viscosity of the liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) by using vibration type viscometer (SV-A/ A&D). In this measurement, Titanium blade was used as a vibration tips, and it vibrate at 30 Hz. Calibration was performed using pure water, and the resistance occurring in the vibration tip was detected as a voltage
3.4. Low friction mechanism of chlorine-doped a-C:H films To summarize the sliding tests and surface analyses, the coupling between the chlorine-doped a-C:H films and the aluminium alloy had the function of improving the tribological properties, and the results suggested that this was due to the formation of a hydrated tribofilm via a tribochemical reaction. We propose that aluminium chloride hexahydrate (AlCl3(H2O)6) is a candidate hydrate containing both Cl and Al. This compound contains six water molecules which are coordination bonded with aluminium in the centre and three chlorine ions around the coordinate. This substance has deliquescent behaviour; it adsorbs water molecules and becomes a liquid with exposure to the atmosphere. Fig. 12(a) shows a photograph of aluminium chloride hexahydrate dehydrated using silica gel, and Fig. 12(b) shows the liquefied material after exposing this reagent to air for 1 h (relative humidity of 50 ± 5%). Hence, it is clear that the liquid state was easily formed by adsorption of 577
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
value and the viscosity was calculated. The viscosity of pure water, polyalpha olefin 4 (PAO4) of general base oil, and liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) were compared. The liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) was prepared by exposure to an environment with humidity of 50 ± 5% for 72 h. Fig. 14 shows the results of the viscosity measurements, where pure water, PAO4, and liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) showed values of 0.94, 26.56, and 28.28 mPa s, respectively. The viscosity of liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) was much higher than that of pure water and nearly the same as PAO4. It is possible to consider that the liquefied aluminium chloride hydrate with high viscosity played a role of lubricant in reducing friction coefficient. To summarize these considerations and experimental results, aluminium chloride hydrate is thought to be a plausible candidate for the tribofilm formed via a tribochemical reaction between chlorine-doped aC:H films and the aluminium alloy. The aluminium chloride hydrate was liquefied due to water adsorption from the atmosphere, forming a tribofilm at the sliding interface which acted as a lubricant. It is possible that this tribofilm prevented contact between the solid surfaces and reduced the friction at the sliding interface. This friction process is proposed as one of the mechanisms for the observed decrease in the friction coefficient with Cl addition. 4. Conclusion This study investigated two different types of a-C:H films (chlorinedoped and non-doped) deposited onto an aluminium alloy (ISO-AlMg1SiCu) via the PBII&D process. The friction properties of these films were studied using a reciprocating ball-on-disk sliding tester. Structural analyses using Raman spectroscopy and XPS, and surface analysis using FTIR and TOF-SIMS were performed to investigate the low friction mechanism of the chlorine-doped a-C:H films. Our findings are summarized as follows. (1) The chlorine-doped a-C:H films formed graphite-like structures, unlike the non-doped a-C:H films. In addition, the film hardness and Young's modulus decreased with increasing Cl/(C þ Cl) ratio in the films. (2) The chlorine-doped a-C:H films showed lower friction coefficients compared with the non-doped a-C:H films. The friction coefficients of chlorine-doped a-C:H films decreased with increasing Cl/(C þ Cl) ratio. For Cl/(C þ Cl) ratio up to 0.083, the wear areas of the disk and ball were reduced compared with the non-doped aC:H films, whereas for Cl/(C þ Cl) ratio 0.108, the wear properties deteriorated.
Fig. 12. Photographs showing the results of exposing aluminium chloride hexahydrate to atmospheric air. (a) Dehydrated reagent. (b) Reagent exposed to the atmosphere for 1 h.
Fig. 13. Cl 2p peaks detected from the (a) chlorine-doped a-C:H film surface and (b) wear track of aluminium alloy ball after a sliding test, and the (c) aluminium chloride hexahydrate reagent.
Fig. 14. Comparison of the viscosity of pure water, PAO4, and liquefied aluminium chloride hexahydrate. 578
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
(3) Based on the surface observations and FT-IR analysis of the wear track of the aluminium alloy after sliding tests, the presence of a hydrate material was confirmed. TOF-SIMS and XPS analyses suggested that this material was aluminium chloride hydrate. (4) The viscosity of liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) prepared by exposure to a humid environment was nearly the same value as that of PAO4.
[14] Ahn HJ, Kim JB, Hwang BH, Baik HK, Park JS, Kang D. Structural and chemical characterization of fluorinated amorphous carbon films (a-C: F) as a liquid crystal alignment layer. Diam Relat Mater 2008;17:2019–24. [15] Zhang P, Tay BK, Yu GQ, Lau SP, Fu YQ. Surface energy of metal containing amorphous carbon films deposited by filtered cathodic vacuum arc. Diam Relat Mater 2004;13:459–64. [16] Sheeja D, Tay BK, Sze JY, Yu LJ, Lau SP. A comparative study between pure and Alcontaining amorphous carbon films prepared by FCVA technique together with high substrate pulse biasing. Diam Relat Mater 2003;12:2032–6. [17] Yang L, Neville A, Brown A, Ransom P, Morina A. Friction reduction mechanisms in boundary lubricated W-doped DLC coatings. Tribol Int 2014;70:26–33. [18] Davey W. The extreme pressure lubricating properties of some chlorinated compounds as assessed by the four-ball machine. J Inst Petrol 1945;31:73–88. [19] Gao F, Kotvis PV, Tysoe WT. The surface and tribological chemistry of chlorine- and sulfur- containing lubricant additives. Tribol Int 2004;37:87–92. [20] Furlong O, Gao F, Kotvis P, Tyose T. Understanding the tribological chemistry of chlorine-, sulfur and phosphorus-containing additives. Tribol Int 2008;40:699–708. [21] Tokuta Y, Itoh T, Shiozaki T, Kawaguchi M, Sasaki S. Effect of chlorine doping on tribological properties of amorphous carbon films deposited by plasma-based ion implantation and deposition. Tribol Int 2017;113:377–82. [22] Riedo E, Comin F, Chevrier J, Schmithusen F, Decossas S, Sancrotti M. Structural properties and surface morphology of laser-deposited amorphous carbon and carbon nitride films. Surf Coating Tech 2000;125:124–8. [23] Oliver WC, Pharr GM. An improved technique for determining hardness and elasticmodulus using load and displacement sensing indentation experiments. J Mater Res 1992;7:1564–83. [24] Ferrari AC, Robertson J. Resonant Raman spectroscopy of disordered, amorphous, and diamond-like carbon. Phys Rev B 2001;64:075414–513. [25] Ferrari AC, Robertson J. Raman spectroscopy of amorphous, nanostructured, diamond–like carbon, and nanodiamond. Philos Trans Roy Soc A 2004;362: 2477–512. [26] Nakamizo M, Kammereck R, Walker PL. Laser Raman studies on carbons. Carbon 1974;12:259–67. [27] Ferrari AC, Robertson J. Interpretation of Raman spectra of disordered and amorphous carbon. Phys Rev B 2000;61:14095–107. [28] Nakazawa H, Kawabata T, Kudo M, Mashita M. Structural changes of diamond-like carbon films due to atomic hydrogen exposure during annealing. Appl Surf Sci 2007;253:4188–96. [29] Kalish R, Lifshitz Y, Nugent K, Prawer S. Thermal stability and relaxation in diamond-like-carbon. A Raman study of films with different sp3 fraction (ta-C to aC). Appl Phys 1999;74:2936–8. [30] Li H, Xu T, Wang C, Chen J, Zhou H, Liu H. Annealing effect on the structure, mechanical and tribological properties of hydrogenated diamond-like carbon films. Thin Solid Film 2006;515:2153–60. [31] Hind A, Al A, Grassian VH. FT-IR study of water adsorption on aluminium oxide surface. Langmuir 2003;19:341–7.
Acknowledgment Funding: This work was supported by JSPS KAKENHI [grant number JP15K17964]. References [1] Grill A. Electrical and optical properties of diamond-like carbon. Thin Solid Film 1999;355:189–93. [2] Trucio-Ortega D, Rodil SE, Muhl S. Corrosion behavior of amorphous carbon deposited in 0.89% NaCl by electrochemical impedance spectroscopy. Diam Relat Mater 2009;18:1360–8. [3] Grill A. Diamond-like carbon coatings as biocompatible materials-an over view. Diam Relat Mater 2003;12:166–70. [4] Kano M. Diamond-like carbon coating applied to automotive engine components. Tribol Online 2014;9:135–42. [5] Kalin M, Vizintin J. The tribological performance of DLC-coated gears lubricated with biodegradable various pinion/gear material combinations. Wear 2005;259: 1270–80. [6] Dai M, Zhou K, Yuan Z, Ding Q, Fu Z. The cutting performance of diamond and DLC coated cutting tools. Diam Relat Mater 2000;9:1753–7. [7] Aisenberg S, Chabot R. Ion-beam deposition of thin films of diamond-like carbon. Appl Phys 1971;42:2953–8. [8] Grill A. Diamond-like carbon: state of the art. Diam Relat Mater 1999;8:428–34. [9] Erdemir A. Genesis of super low friction and wear in diamond-like carbon films. Tribol Int 2004;37:1005–12. [10] Kano M. Super low friction of DLC applied to engine cam follower lubricated with ester-containing oil. Tribol Int 2006;39:1682–5. [11] Camargo SS, Neto ALB, Santos RA, Feire Jr FL, Finger F. Improved high-temperature stability of Si incorporated a-C: H films. Diam Relat Mater 1998;7:1155–62. [12] Donnet C. Recent progress on the tribology of doped diamond-like and carbon alloy coatings: a review. Surf Coating Tech 1998;100–101:180–6. [13] Hakovirta M, He XM, Nastasi M. Optical properties of fluorinated diamond-like carbon films produced by pulsed glow discharge plasma immersion ion processing. Appl Phys 2000;88:1456–9.
579
Contents lists available at ScienceDirect
Tribology International journal homepage: www.elsevier.com/locate/triboint
Low friction mechanism of chlorine-doped amorphous carbon films sliding against an aluminium alloy Yuuki Tokuta a, *, Takashi Itoh b, Takahiko Shiozaki b, Masahiro Kawaguchi a, Shinya Sasaki c a b c
Tokyo Metropolitan Industrial Technology Research Institute, 2-5-10 Aomi, Koto-ku, Tokyo 135-0064, Japan Fujimetal Co. Ltd., 2-7-16 Hinode, Kawasaki-ku Kawasaki-shi, Kanagawa 210-0824, Japan Tokyo University of Science, 6-3-1 Niijuku, Karsushika-ku, Tokyo 125-8585, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords: Coating Self-lubricating Diamond-like
The effect of chlorine-doping on the tribological properties of amorphous carbon films was investigated. Chlorinedoped amorphous carbon films were deposited using a plasma-based ion implantation and deposition (PBII&D) method with vaporized tetrachloroethylene as a precursor. During sliding tests with aluminium alloy counter parts, chlorine-doped amorphous carbon films showed lower friction coefficients compared to hydrogenated amorphous carbon films. Both aluminium and a chlorine-based hydrate were observed on the wear track of the carbon films. This hydrate was a tribofilm formed via a tribochemical reaction during sliding between the chlorine-doped amorphous carbon films and aluminium alloy. The viscosity of this tribofilm was similar to polyalpha olefin and we propose that it acted as a lubricant and lowered the friction coefficient.
1. Introduction Amorphous carbon films are generally composed of sp2-and sp3-hybridized orbitals of carbon and hydrogen. In recent years, amorphous carbon films have attracted attention in research and industry due to their superior properties, such as optical properties, corrosion resistance, and biocompatibility [1–3]. In particular, amorphous carbon films have extraordinary wear and friction properties, and they have contributed to drastic improvement of the tribological properties of gears, cutting tools, and machine parts [4–6]. Since the seminal report regarding amorphous carbon films was published by Aisenberg and Chabot in 1971 [7], this material has attracted much attention from many research groups. As research on amorphous carbon films has progressed, it has been confirmed that various film properties depend on the structure. Grill proposed that the characteristics and structure of amorphous carbon films are determined by several factors, including the concentrations of hydrogen, sp3 bonding, and sp2 bonding in the films. In addition, amorphous carbon films can be classified into four types: amorphous carbon (a-C), hydrogenated amorphous carbon (a-C:H), tetrahedral amorphous carbon (taC), and hydrogenated tetrahedral amorphous carbon (ta-C:H) [8]. The aC:H and ta-C:H films contain up to approximately 50% hydrogen, a-C and ta-C films contain less than 1% hydrogen, and ta-C and ta-C:H consist of a high fraction of sp3 bonding. Hence, the term “amorphous carbon films”
* Corresponding author. E-mail address: [email protected] (Y. Tokuta). http://dx.doi.org/10.1016/j.triboint.2017.06.024 Received 27 March 2017; Received in revised form 31 May 2017; Accepted 16 June 2017 Available online 19 June 2017 0301-679X/© 2017 Elsevier Ltd. All rights reserved.
covers a wide range of carbon-based coatings and the investigation of the relationship between the structure and properties of the films is critical in their application. Erdemir reported that the hydrogen content of a-C:H films plays an important role in the friction behaviour [9]. a-C:H films contain a large amount of hydrogen, where the terminated hydrogen eliminates most of the dangling σ bonds and π-π interactions and contributes to superlow friction under dry conditions. The low friction of a-C:H films was attributed to the high chemical inertness of films as a result of the surface passivation of the dangling bonds. On the contrary, ta-C films are suitable for oil-lubricated application because of their unterminated dangling bonds inside the films which activate the wear surface. A tribochemical layer caused by friction has been observed to form on the surface of such films [10]. The friction behaviour of these films is strongly dependent on the sliding environment. Hence, it is clear that the optimisation of the film structure for a particular application is critical. Recent studies have demonstrated doping of amorphous carbon films with different elements (such as Si, F, B, N, Al Cr, Ti, and W) as a new technique for optimising the structure [11–17]. These efforts contributed to the understanding of mechanisms for improving many properties, for example thermal resistance, optical gap energy, surface energy, relaxation of internal stress, and activation of the tribochemical reaction. Si doping of amorphous carbon films has been shown to improve the high temperature behaviour in comparison to hydrogenated amorphous
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
carbon films and also achieve low friction under high humidity due to the formation of SiO2 particles through a tribochemical reaction [11,12]. In the case of F doping, the surface energy was shown to decrease and the optical band gap improve with increasing F content; the increase in the optical band gap energy induced by F doping was attributed to a modified film structure with a higher fraction of sp3 bonding compared to nondoped films [13,14]. Doping with metals, such as Al and Ti, has also been shown to contribute to a reduction in the surface energy, as observed for Si doping [15]. The dopant metal atoms inside the films form metal-O covalent bonds, localizing the electrons of the metal; thus, a decrease in dipoles contributes to reducing the surface energy. In addition, metal doping contributes to reducing internal stress and activating the chemical reaction between amorphous carbon films and lubricant additives [16,17]. The aim of this study is to investigate the effect of chlorine doping on the tribological properties of a-C:H films during sliding against an aluminium alloy under non-lubricated conditions. Our hypothesis is that the chlorine inside of the films will contribute to the formation of a chlorinated tribofilm at the sliding interface, in the same way as a reaction with chlorine from additives [18–20], where the chlorinated tribofilm can reduce friction coefficient. In our previous study, we observed a lowering of the friction due to chlorine-doped a-C:H films during sliding with aluminium alloy counter parts [21]. However, the exact mechanism behind this observed low friction was not clarified and hence, the major objective of this study is to reveal the low friction mechanism of such films. Non-doped a-C:H films and chlorine-doped a-C:H films were deposited using a plasma-based ion implantation and deposition (PBII&D) method using vaporized toluene (C6H5CH3) and tetrachloroethylene (C2Cl4). Various characterisation methods, including micro-laser Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), Elastic recoil detection analysis (ERDA), and nano-indentation hardness testing were employed to investigate the structure and mechanical properties. The chlorine-doped tribofilm formed through a tribochemical reaction was investigated using surface analysis techniques including Fourier transform infrared spectroscopy (FT-IR) and time-of-flight secondary ion mass spectroscopy (TOF-SIMS).
Fig. 1. Schematic diagram of the plasma-based ion implantation and deposition system used in this study.
Table 1 Deposition parameters of non-doped and chlorine-doped a-C:H films.
Non-doped a-C:H films Chlorine-doped a-C:H films
2. Experimental setup
#1 #2 #3 #4 #5
Cl/(C þ Cl) ratio
Hydrogen content [atm %]
0.0
21.9
0.013 0.037 0.083 0.108 0.118
20.3 10.1 4.8 3.1 <3.0
deposition parameters, such as the applied voltage of 5 kV and RF power of 500 W were the same for all samples. In Table 1, the hydrogen content and the presence ratio of chlorine relative to the total number of carbon and chlorine Cl/(C þ Cl) of each film were measured using ERDA and XPS, respectively. ERDA was used determine the hydrogen content inside of each film at a depth of 50 nm from the film surface. XPS was used to measure the Cl/(C þ Cl), where the Cl/(C þ Cl) was estimated from the relative areas of the chlorine peaks (Cl 2p) and the carbon peaks (C 1s). The surface roughness parameter Ra of these amorphous carbon film were about 0.022–0.026 μm and film thickness was above 1.0 μm.
2.1. Deposition of amorphous carbon films using PBII&D A PBII&D system (PBII-R1000/Kurita Seisakusho) was employed to deposit the non-doped a-C:H films and chlorine-doped a-C:H films on an aluminium alloy (ISO-AlMg1SiCu) disk with a diameter of 30 mm. Nano-indentation hardness of substrate was 1.7 GPa. Fig. 1 shows a schematic diagram of the PBII&D system used in this study. In this method, the substrate itself is used as an RF antenna and a highfrequency power supply generates a plasma from the source gas around the substrate; then a negative charge is applied to the substrate. Using this process, plasma ions can be implanted into the substrate surface to form the desired film. This method has various advantages; for example, it is possible to form a uniform film on a complex surface at deposition temperatures of less than 473 K. Note that the films were deposited via a hybrid process of simultaneous chemical reaction and ion implantation, which was accelerated by the applied voltage. Prior to film deposition, all substrates were cleaned by Ar bombardment to remove contaminants and activate the surface. After Ar bombardment, Si interlayer was deposited using hexamethyldisiloxane (HMDSO) gas on the substrate surface. Table 1 shows the different chlorine and hydrogen contents of five types of samples prepared in this study. A non-doped a-C:H films were prepared using only toluene gas, while the chlorine-doped a-C:H films were deposited using two different gases, a mixture of toluene and tetrachloroethylene (C6H5CH3 þ C2Cl4; #1, #2, #3, and #4) or pure C2Cl4 (#5). When using the mixed gas, the flow ratio of C2Cl4 to C6H5CH3 was increased in order of #1 to #4; the other
2.2. Structure and mechanical properties Raman spectroscopy and XPS analyses were used to investigate the structure of the films. Specifically, the effect of chlorine-doping on the internal structure of the films was analysed by Raman spectroscopy and the effect on the sp3/sp2 ratio of the films was calculated from XPS data. The Raman spectrometer (In Via Reflex/Renishaw) was equipped with a YAG laser with a wavelength of 532 nm. The laser power, spot size, and spectral range were 0.25 W, 5.0 μm, and 800–2000 cm1, respectively. An XPS apparatus (Quantera/Ulvac-Phi) was used to analyse chlorine in the films and measurements were undertaken after a sputtering pretreatment using Arþ. The detection range of the carbon and chlorine peaks were 275–300 eV and 190–220 eV, respectively. The X-ray source, X-ray power, and target emission were AlKα, 10kV25W, and 3.0 mA. The sp3/(sp3 þ sp2) bonding ratio of carbon was investigated by analysing the 574
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
3. Results and discussion
XPS spectra using the method described by Riedo et al. [22]. In this method, sp2 bonding peak and sp3 bonding peak were defined at peak position of 284.5 eV and 285.3 eV respectively. The sp3/(sp3 þ sp2) were calculated from peak intensity ratio of these peaks. The peak fitting was achieved using Gaussian fitting software. A nano-indentation hardness tester (TI-950/Hysitron) with a Berkovich diamond tip was used to compare the film hardness and Young's modulus of the non-doped a-C:H films and chlorine-doped a-C:H films. The indentation speed was 5 nm/s and a maximum indentation depth of 100 nm was used to minimize the contribution from the substrate. An indenter calibration was achieved by Oliver-Parr method [23] The film hardness and Young's modulus were estimated by fitting the loaddisplacement curves obtained during indentation.
3.1. Structure and hardness of amorphous carbon films Fig. 2 shows Raman spectra of both of non-doped a-C:H (Cl/(C þ Cl) of 0.0) and chlorine-doped (Cl/(C þ Cl) of 0.013, 0.036, 0.083, 0.107, and 0.118) a-C:H films. In the Raman spectra of the amorphous carbon films, the D peak around 1350 cm1 (attributed to a disorder structure) and the G peak around 1550 cm1 (attributed to the graphite structure) were observed. The D peak is related to the breathing modes of sp2 atoms in rings, whereas the G peak is related to the bond stretching of all pairs of sp2 atoms in rings and chains [24–26]. The peak positions, the fullwidth-at-half-maximum (FWHM) of the peaks, and the intensity ratio of the D peak to the G peak (ID/IG) provide important information about the film structure. It can be seen in this figure that, while both the D and G peaks were observed for all films, the features of the Raman spectra were different. An increasing Cl/(C þ Cl) ratio resulted in a slight increase in the D peak intensity, indicating a structural change. In general, as the graphite-like structure formed inside the films, the D and G peaks tended to separate due to the increase in the intensity of the D peak. The evolution of a distinct D peak indicated the formation of a graphitic domain [27]. This result suggests that the a-C:H film structure was dependent on the precursor gas and a graphite-like structure was formed with increasing Cl/(C þ Cl) ratio. Fig. 3 shows the position of the G peak and its FWHM value as functions of the Cl/(C þ Cl) ratio of the films. It can be seen that, as Cl/(C þ Cl) increased, the G peak shifted toward a higher wavenumber and the FWHM(G) decreased. Both of these tendencies suggest an ordering of the film structure [28]. The change in ID/ IG is also important for understanding the film structure; Fig. 4 compares the ID/IG ratio measured using Raman spectroscopy with the sp3/ (sp2 þ sp3) ratio determined by XPS analysis. It can be seen that the ID/IG ratio was inversely proportional to Cl/(C þ Cl), attributed to the formation of the graphite-like structure (consistent with the shift in the G peak position and decrease in FWHM(G)) [29]. This was also confirmed by the decrease in the sp3/(sp2 þ sp3) ratio. These results suggested that a-C:H film deposition using chlorine-containing precursor gases contributed to the formation of a graphite-like structure with a high sp2 fraction. Fig. 5 shows the film hardness and Young's modulus as functions of Cl/(C þ Cl) ratio in the films. A reduction in the film hardness and Young's modulus was observed with increasing Cl/(C þ Cl) ratio. It is well known that the increased sp2 fraction decreases the film hardness and Young's modulus due to a loss of the diamond-like characteristics of the film [30]. Fig. 6 shows ID/IG ratio of Raman spectrum and sp3/ (sp2 þ sp3) bonding ratio calculated from XPS spectrum as function of films hardness. From this result, it was confirmed that the films hardness
2.3. Sliding tests and observation of wear tracks A reciprocating-type ball-on-disk sliding tester (Tribometer/CSM) was employed to perform sliding tests on each film. An aluminium alloy ball (ISO-AlMg1SiCu) with a diameter of 6 mm was used as the counter part material to slide against the disks with the deposited films. The nano-indentation hardness and surface roughness Ra of aluminium alloy used in this study were 1.8 GPa and 0.059 μm, respectively. The applied load, sliding speed, sliding distance, and total number of sliding cycles were 5 N, 20 mm/s, 10 mm, and 2000 cycles, respectively. The sliding tests were performed under non-lubricated conditions in air at a temperature of 20 ± 2 C and a relative humidity of 50 ± 5%. An optical microscope (VHX-500/Keyence) and a confocal laser-scanning microscope (OLS-4000/Olympus) were employed to observe the wear track and calculate the wear area to investigate the friction mechanism and wear properties. 2.4. Surface analysis after sliding tests FT-IR (FTIR-6600/Jasco) and TOF-SIMS (TRIFT/Ulvac-Phi) were used to investigate the presence of compounds along the wear track of the chlorine-doped a-C:H films. In the FT-IR analyses, the measurement range was set to 1000–4000 cm1 in microscopic mode. For TOF-SIMS measurements, we analysed in mass resolution mode using Gaþ ions as the ion source. Spot size and current density were 100 μm square and 1.0 nA in DC conversion. The negative fragment ions detected from the wear track were investigated to determine the type of tribofilm at the sliding interface.
Fig. 3. G peak position and FWHM(G) as functions of the Cl/(C þ Cl) ratio in the aC:H films.
Fig. 2. Raman spectra of the a-C:H films with different Cl/(C þ Cl) ratio. 575
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
Fig. 4. ID/IG ratio and sp3/(sp2 þ sp3) ratio as functions of the Cl/(C þ Cl) ratio in the aC:H films. Fig. 7. Average friction coefficient as a function of Cl/(C þ Cl) ratio.
increasing Cl/(C þ Cl) ratio. The friction coefficient of the non-doped aC:H film was about 0.107 and the chlorine-doped a-C:H films with Cl/ (C þ Cl) ratio of 0.013, 0.037, 0.083, 0.108, and 0.118 showed values of 0.070, 0.066, 0.060, 0.059, and 0.042, respectively. Of particular note, even when the Cl/(C þ Cl) ratio was very small (0.013 for #1), the friction coefficient reduced by approximately 40% compared with the non-doped a-C:H film. The small error bars on the data shows repeatability in five tests. These results suggest that chlorine doping can reduce the friction coefficient of a-C:H films. Fig. 8 shows the wear areas measured for the films and counter parts after sliding tests. There was a threshold in the correlation between the Cl/(C þ Cl) and the wear area; when the Cl/(C þ Cl) was below 0.083, the wear areas of the disk and ball were less than that of the non-doped aC:H film; however when the Cl/(C þ Cl) was 0.108, the wear properties deteriorated and the wear area increased significantly. In the range of Cl/ (C þ Cl) of 0.013–0.083, the chlorine-doped a-C:H films showed improved wear resistance and low aggressiveness against the counter parts, despite the lower hardness compared to that of the non-doped aC:H film. These results suggest that the non-doped a-C:H and chlorinedoped a-C:H films have different friction mechanisms. Fig. 9 shows a surface image of the wear track on the aluminium alloy ball observed after the sliding test against chlorine-doped a-C:H film (#1), where the presence of a liquid substance adhered to the wear track was observed. This liquid material is thought to be a tribofilm, which contributed to the low friction, high wear resistance, and low aggressiveness against the counter parts.
Fig. 5. Film hardness and Young's modulus as functions of the Cl/(C þ Cl) ratio in the aC:H films.
Fig. 6. ID/IG ratio and sp3/(sp2 þ sp3) ratio as functions of the film hardness.
indicated lower value with less sp3 bonding. Hence, the tendencies of the results of the nano-indentation hardness tests were in a good agreement with the results of Raman spectroscopy and XPS. 3.2. Friction behaviour of amorphous carbon films Fig. 8. Wear areas of the ball and disk as functions of Cl/(C þ Cl) ratio.
Fig. 7 shows that the average friction coefficient decreased with 576
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
Fig. 10. FT-IR measurements of the wear track of the aluminium alloy ball sliding on (a) chlorine-doped a-C:H films and (b) non-doped a-C:H films.
Fig. 9. Optical microscopy image of the wear track of sample #1 after the sliding test where the liquid regions are indicated.
3.3. Surface analysis of wear tracks Fig. 10 show FT-IR spectra of the wear track of the aluminium alloy ball after the sliding tests against non-doped a-C:H films and chlorinedoped a-C:H films (#1). The wear track of the ball in friction with chlorine-doped a-C:H film (Fig. 10(a)) showed two peaks, at 3400 cm1 and 1560 cm1, while no significant peaks were observed in the case of the non-doped a-C:H film (Fig. 10(b)). One possibility is that these characteristic peaks indicate the presence of water molecules or hydroxy. In the FT-IR analysis of gaseous water, the symmetric OH stretching vibration (v1) and asymmetric OH stretching vibration (v3) are centred at 3656 and 3755 cm1, respectively, while the bending mode of σ(H2O) is observed around 1594 cm1. In the case of the liquid water, the band consists of broadened v1 and v2 peaks due to the formation of the hydrogen bond in the liquid, and these peaks cannot be distinguished. These deformations in the stretching mode result in the peak position shifting approximately 200–400 cm1 [31]. These results suggest that hydrate and/or water adsorption occurred at the wear track during sliding through a tribochemical reaction. Fig. 11 shows the negative fragment ion peaks detected from the wear track of chlorine-doped a-C:H film (#1) by TOF-SIMS analysis. These results show information about the detected negative ions, obtained by subtracting the spectrum recorded outside of the wear track from that obtained along the wear track. It can be seen that several kinds of particular fragment ion peaks were observed, such as H2OCl, H2O2Al, HAlCl, HOAlCl, and H2OAlCl, which suggest the presence of a chlorinealuminium based hydrate.
Fig. 11. TOF-SIMS analysis of the wear track of chlorine-doped a-C:H film #1.
water molecules from the atmosphere. From this standpoint, XPS analysis was carried out on an aluminium chloride hexahydrate reagent and the wear track of the aluminium ball after a sliding test with a chlorine-doped a-C:H film. Fig. 13 shows the Cl 2p peaks from the XPS analysis for the (a) surface of a chlorine-doped aC:H film, (b) wear track of the aluminium alloy ball after sliding tests with chlorine-doped a-C:H film, and (c) aluminium chloride hexahydrate reagent. Obvious differences in the Cl 2p peak position between the chlorine-doped a-C:H film surface and the other samples were observed. Of particular note, the peak position and FWHM of the Cl 2p peak detected from the wear track of the aluminium alloy ball were similar to that of the aluminium chloride hexahydrate reagent. Hence, it is possible that the composition of the tribofilm formed through tribochemical reaction by the friction between chlorine-doped a-C:H films and the aluminium alloy was aluminium chloride hydrate (such as hexahydrate). As aluminium chloride hydrate is deliquescent, it is possible that the liquid seen on the wear track of the aluminium alloy after the sliding test was the liquefied form of this material. In order to investigate the effect of the mechanical behaviour of aluminium chloride hexahydrate on the friction, we measured the viscosity of the liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) by using vibration type viscometer (SV-A/ A&D). In this measurement, Titanium blade was used as a vibration tips, and it vibrate at 30 Hz. Calibration was performed using pure water, and the resistance occurring in the vibration tip was detected as a voltage
3.4. Low friction mechanism of chlorine-doped a-C:H films To summarize the sliding tests and surface analyses, the coupling between the chlorine-doped a-C:H films and the aluminium alloy had the function of improving the tribological properties, and the results suggested that this was due to the formation of a hydrated tribofilm via a tribochemical reaction. We propose that aluminium chloride hexahydrate (AlCl3(H2O)6) is a candidate hydrate containing both Cl and Al. This compound contains six water molecules which are coordination bonded with aluminium in the centre and three chlorine ions around the coordinate. This substance has deliquescent behaviour; it adsorbs water molecules and becomes a liquid with exposure to the atmosphere. Fig. 12(a) shows a photograph of aluminium chloride hexahydrate dehydrated using silica gel, and Fig. 12(b) shows the liquefied material after exposing this reagent to air for 1 h (relative humidity of 50 ± 5%). Hence, it is clear that the liquid state was easily formed by adsorption of 577
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
value and the viscosity was calculated. The viscosity of pure water, polyalpha olefin 4 (PAO4) of general base oil, and liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) were compared. The liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) was prepared by exposure to an environment with humidity of 50 ± 5% for 72 h. Fig. 14 shows the results of the viscosity measurements, where pure water, PAO4, and liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) showed values of 0.94, 26.56, and 28.28 mPa s, respectively. The viscosity of liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) was much higher than that of pure water and nearly the same as PAO4. It is possible to consider that the liquefied aluminium chloride hydrate with high viscosity played a role of lubricant in reducing friction coefficient. To summarize these considerations and experimental results, aluminium chloride hydrate is thought to be a plausible candidate for the tribofilm formed via a tribochemical reaction between chlorine-doped aC:H films and the aluminium alloy. The aluminium chloride hydrate was liquefied due to water adsorption from the atmosphere, forming a tribofilm at the sliding interface which acted as a lubricant. It is possible that this tribofilm prevented contact between the solid surfaces and reduced the friction at the sliding interface. This friction process is proposed as one of the mechanisms for the observed decrease in the friction coefficient with Cl addition. 4. Conclusion This study investigated two different types of a-C:H films (chlorinedoped and non-doped) deposited onto an aluminium alloy (ISO-AlMg1SiCu) via the PBII&D process. The friction properties of these films were studied using a reciprocating ball-on-disk sliding tester. Structural analyses using Raman spectroscopy and XPS, and surface analysis using FTIR and TOF-SIMS were performed to investigate the low friction mechanism of the chlorine-doped a-C:H films. Our findings are summarized as follows. (1) The chlorine-doped a-C:H films formed graphite-like structures, unlike the non-doped a-C:H films. In addition, the film hardness and Young's modulus decreased with increasing Cl/(C þ Cl) ratio in the films. (2) The chlorine-doped a-C:H films showed lower friction coefficients compared with the non-doped a-C:H films. The friction coefficients of chlorine-doped a-C:H films decreased with increasing Cl/(C þ Cl) ratio. For Cl/(C þ Cl) ratio up to 0.083, the wear areas of the disk and ball were reduced compared with the non-doped aC:H films, whereas for Cl/(C þ Cl) ratio 0.108, the wear properties deteriorated.
Fig. 12. Photographs showing the results of exposing aluminium chloride hexahydrate to atmospheric air. (a) Dehydrated reagent. (b) Reagent exposed to the atmosphere for 1 h.
Fig. 13. Cl 2p peaks detected from the (a) chlorine-doped a-C:H film surface and (b) wear track of aluminium alloy ball after a sliding test, and the (c) aluminium chloride hexahydrate reagent.
Fig. 14. Comparison of the viscosity of pure water, PAO4, and liquefied aluminium chloride hexahydrate. 578
Y. Tokuta et al.
Tribology International 115 (2017) 573–579
(3) Based on the surface observations and FT-IR analysis of the wear track of the aluminium alloy after sliding tests, the presence of a hydrate material was confirmed. TOF-SIMS and XPS analyses suggested that this material was aluminium chloride hydrate. (4) The viscosity of liquefied aluminium chloride hexahydrate (AlCl3(H2O)6) prepared by exposure to a humid environment was nearly the same value as that of PAO4.
[14] Ahn HJ, Kim JB, Hwang BH, Baik HK, Park JS, Kang D. Structural and chemical characterization of fluorinated amorphous carbon films (a-C: F) as a liquid crystal alignment layer. Diam Relat Mater 2008;17:2019–24. [15] Zhang P, Tay BK, Yu GQ, Lau SP, Fu YQ. Surface energy of metal containing amorphous carbon films deposited by filtered cathodic vacuum arc. Diam Relat Mater 2004;13:459–64. [16] Sheeja D, Tay BK, Sze JY, Yu LJ, Lau SP. A comparative study between pure and Alcontaining amorphous carbon films prepared by FCVA technique together with high substrate pulse biasing. Diam Relat Mater 2003;12:2032–6. [17] Yang L, Neville A, Brown A, Ransom P, Morina A. Friction reduction mechanisms in boundary lubricated W-doped DLC coatings. Tribol Int 2014;70:26–33. [18] Davey W. The extreme pressure lubricating properties of some chlorinated compounds as assessed by the four-ball machine. J Inst Petrol 1945;31:73–88. [19] Gao F, Kotvis PV, Tysoe WT. The surface and tribological chemistry of chlorine- and sulfur- containing lubricant additives. Tribol Int 2004;37:87–92. [20] Furlong O, Gao F, Kotvis P, Tyose T. Understanding the tribological chemistry of chlorine-, sulfur and phosphorus-containing additives. Tribol Int 2008;40:699–708. [21] Tokuta Y, Itoh T, Shiozaki T, Kawaguchi M, Sasaki S. Effect of chlorine doping on tribological properties of amorphous carbon films deposited by plasma-based ion implantation and deposition. Tribol Int 2017;113:377–82. [22] Riedo E, Comin F, Chevrier J, Schmithusen F, Decossas S, Sancrotti M. Structural properties and surface morphology of laser-deposited amorphous carbon and carbon nitride films. Surf Coating Tech 2000;125:124–8. [23] Oliver WC, Pharr GM. An improved technique for determining hardness and elasticmodulus using load and displacement sensing indentation experiments. J Mater Res 1992;7:1564–83. [24] Ferrari AC, Robertson J. Resonant Raman spectroscopy of disordered, amorphous, and diamond-like carbon. Phys Rev B 2001;64:075414–513. [25] Ferrari AC, Robertson J. Raman spectroscopy of amorphous, nanostructured, diamond–like carbon, and nanodiamond. Philos Trans Roy Soc A 2004;362: 2477–512. [26] Nakamizo M, Kammereck R, Walker PL. Laser Raman studies on carbons. Carbon 1974;12:259–67. [27] Ferrari AC, Robertson J. Interpretation of Raman spectra of disordered and amorphous carbon. Phys Rev B 2000;61:14095–107. [28] Nakazawa H, Kawabata T, Kudo M, Mashita M. Structural changes of diamond-like carbon films due to atomic hydrogen exposure during annealing. Appl Surf Sci 2007;253:4188–96. [29] Kalish R, Lifshitz Y, Nugent K, Prawer S. Thermal stability and relaxation in diamond-like-carbon. A Raman study of films with different sp3 fraction (ta-C to aC). Appl Phys 1999;74:2936–8. [30] Li H, Xu T, Wang C, Chen J, Zhou H, Liu H. Annealing effect on the structure, mechanical and tribological properties of hydrogenated diamond-like carbon films. Thin Solid Film 2006;515:2153–60. [31] Hind A, Al A, Grassian VH. FT-IR study of water adsorption on aluminium oxide surface. Langmuir 2003;19:341–7.
Acknowledgment Funding: This work was supported by JSPS KAKENHI [grant number JP15K17964]. References [1] Grill A. Electrical and optical properties of diamond-like carbon. Thin Solid Film 1999;355:189–93. [2] Trucio-Ortega D, Rodil SE, Muhl S. Corrosion behavior of amorphous carbon deposited in 0.89% NaCl by electrochemical impedance spectroscopy. Diam Relat Mater 2009;18:1360–8. [3] Grill A. Diamond-like carbon coatings as biocompatible materials-an over view. Diam Relat Mater 2003;12:166–70. [4] Kano M. Diamond-like carbon coating applied to automotive engine components. Tribol Online 2014;9:135–42. [5] Kalin M, Vizintin J. The tribological performance of DLC-coated gears lubricated with biodegradable various pinion/gear material combinations. Wear 2005;259: 1270–80. [6] Dai M, Zhou K, Yuan Z, Ding Q, Fu Z. The cutting performance of diamond and DLC coated cutting tools. Diam Relat Mater 2000;9:1753–7. [7] Aisenberg S, Chabot R. Ion-beam deposition of thin films of diamond-like carbon. Appl Phys 1971;42:2953–8. [8] Grill A. Diamond-like carbon: state of the art. Diam Relat Mater 1999;8:428–34. [9] Erdemir A. Genesis of super low friction and wear in diamond-like carbon films. Tribol Int 2004;37:1005–12. [10] Kano M. Super low friction of DLC applied to engine cam follower lubricated with ester-containing oil. Tribol Int 2006;39:1682–5. [11] Camargo SS, Neto ALB, Santos RA, Feire Jr FL, Finger F. Improved high-temperature stability of Si incorporated a-C: H films. Diam Relat Mater 1998;7:1155–62. [12] Donnet C. Recent progress on the tribology of doped diamond-like and carbon alloy coatings: a review. Surf Coating Tech 1998;100–101:180–6. [13] Hakovirta M, He XM, Nastasi M. Optical properties of fluorinated diamond-like carbon films produced by pulsed glow discharge plasma immersion ion processing. Appl Phys 2000;88:1456–9.
579