Role of transfer layer on tribological properties of nanocrystalline diamond nanowire film sliding against alumina allotropes

Role of transfer layer on tribological properties of nanocrystalline diamond nanowire film sliding against alumina allotropes

Diamond & Related Materials 48 (2014) 6–18 Contents lists available at ScienceDirect Diamond & Related Materials journal homepage: www.elsevier.com/...

5MB Sizes 198 Downloads 67 Views

Diamond & Related Materials 48 (2014) 6–18

Contents lists available at ScienceDirect

Diamond & Related Materials journal homepage: www.elsevier.com/locate/diamond

Role of transfer layer on tribological properties of nanocrystalline diamond nanowire film sliding against alumina allotropes R. Radhika a, N. Kumar b,⁎, A.T. Kozakov c, K.J. Sankaran d, S. Dash b, A.K. Tyagi b, N.-H. Tai d, I.N. Lin e a

Crystal Growth Centre, Anna University, Chennai 600025, TN, India Materials Science Group, Indira Gandhi Centre for Atomic Research, Kalpakkam, TN, India Research Institute of Physics, Southern Federal University, Rostov-on-Don, Russian Federation d Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu, Taiwan, ROC e Department of Physics, Tamkang University, Tamsui, Taiwan, ROC b c

a r t i c l e

i n f o

Article history: Received 15 March 2014 Received in revised form 28 April 2014 Accepted 12 June 2014 Available online 20 June 2014 Keywords: Nanocrystalline DNW film Tribological properties Sliding counterbodies Al2O3 Sapphire Ruby Interface chemical behavior

a b s t r a c t The tribological properties of nanocrystalline diamond nanowire (DNW) film treated in CH4 atmosphere at 400 °C were studied in ambient atmosphere at room temperature using various allotropes of alumina ball as sliding counterbodies. Super low value of friction coefficient (~ 0.003) and extremely high wear resistance (~ 2.8 × 10−21 mm3/Nm) were observed when the Al2O3 ball slides against the film. In contrast, high friction coefficients with the values ~0.06 and ~0.07 were observed while using sapphire and ruby balls, respectively. Wear loss was also high ~4 × 10−19 mm3/Nm and 2.8 × 10−15 mm3/Nm in DNW/sapphire and DNW/ruby sliding pairs, respectively. Such a behavior is fundamentally explained in terms of the chemical nature of the sliding interfaces and surface energy of ball counterbodies. As a consequence, the chemical affinity of Al2O3 ball towards the carbon atoms is less, which resulted in the absence of carbonaceous transfer layer formation on the Al2O3 ball scar. However, in the case of sapphire and ruby balls, the wear track was found to be highly deformed and significant development of carbonaceous transfer layer was observed on respective ball scars. This phenomenon involving transfer layer formation is related to high surface energy and strong chemical affinities of sapphire and ruby balls towards carbon atoms. In such a condition, sliding occurs between film and the carbonaceous transfer layer formed on the ball exhibiting high energy due to covalent carbon bonds that chemically interact and enhance sliding resistance. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Tribological efficiency needs advanced materials for specific engineering applications where sliding occurs between alternate materials [1,2]. In this regard, it is technically important to investigate ceramic counterbodies sliding against carbon based films like nanocrystalline diamond (NCD) and DNW [3,4]. In NCD films, the typical size of the diamond crystallites falls in the range of 15–30 nm. However, in DNW films, ultrananocrystalline diamond grains with sizes 3–5 nm are embodied in the wires. These wires have typical dimensions of 100–130 nm in length and 10–15 nm in diameter. Such films are an advanced class of materials which have normally low/ultra-low friction coefficients with high wear resistance in a wide range of operating conditions [5–8]. It is revealed from several investigations that mainly, formation of carbonaceous transfer layer on sliding ceramic counterbodies often influences tribological properties of diamond-like carbon and NCD films [2,3]. Interestingly, high friction coefficient and high wear rate were measured when NCD slides against SiC and Si3N4 balls [3]. Such a characteristic was explained ⁎ Corresponding author. Tel./fax: +91 4427480081. E-mail address: [email protected] (N. Kumar).

http://dx.doi.org/10.1016/j.diamond.2014.06.005 0925-9635/© 2014 Elsevier B.V. All rights reserved.

by the formation of carbonaceous transfer layer on these balls that causes the development of covalent chemical bonding between the sliding interfaces, resulting in high friction. In contrast, ultra-low friction coefficient and high wear resistance were obtained while sliding against Al2O3, ruby and Zr2O3 balls [2,3]. Furthermore, super-low friction coefficient and high wear resistance were measured on H2/O2 plasma treated nanocrystalline DNW film while using Al2O3 balls [9,10]. This happens due to weak chemical interaction between sliding interfaces which restricts formation of covalently bonded transfer layer. In the early work, it is shown that the value of friction coefficient consistently decreases with increasing the humidity levels during the tribo-test performed on ultra-nanocrystalline diamond (UNCD) films [7]. This value is ~0.13 in low humidity which is significantly decreased to ~ 0.004 at highly humid conditions. Such a reduction in friction coefficient was ascribed to passivation of dangling covalent bonds of carbon atoms occurring due to the formation of chemical species such as C\COO, CH3COH and CH2\O. In another work, friction coefficient of as-deposited DNW film showed a high value of ~0.2 while sliding against Al2O3 ball [9]. However, this value was decreased to a super low value ~ 0.0001 in H2 plasma treated DNW film. Such super low valued friction coefficient is described by the passivation of uncompensated carbon dangling bonds by

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

hydrogen ingress and adsorption of H2O molecules. However, in the same H2 treated DNW film, the friction coefficient is found to increase at low humid atmospheres. This was explained by insufficient chemical passivation of carbon dangling bonds. In the light of abovementioned facts, it was observed that there are very few reports on tribological properties of NCD and DNW films sliding against ceramic balls [3,11]. The specific mechanism responsible for glaring variation in friction behavior remains a matter of study in DNW films [9–11]. In this regard, the present study aims to investigate tribological behavior of nanocrystalline DNW film sliding against three similar classes of ceramic balls such as Al2O3, sapphire and ruby. Frictional behavior was analyzed with regard to microstructure and chemical composition of the sliding counterbodies. In addition, surface topography and surface energy of the sliding counterbodies were investigated to bring out correlation with friction and wear mechanism. Wear tracks of the film were chemically characterized by micro-Raman spectroscopy and micro-XPS to determine the chemical changes and their effect on tribological properties.

7

3. Results and discussion 3.1. Film surface morphology The top-view FESEM micrograph and cross section of CH4 treated DNW film is shown in Fig. 1. It is observed that DNW film possesses dense and uniformly distributed wire-like granular morphology. These wires are composed of several NCD grains. The boundary of a wire consists of chemical impurities like a-C and sp2C_C bonded phases [13–15]. Thickness of the amorphous carbon and graphitic shield present around a nanowire varies from a few atomic layers to 5 nm [9]. The contact angle of such film is high (172°) and belongs to the hydrophobic class of nanostructured materials. Low surface energy can be attributed to the formation of microscopic cavities where air can be trapped and create the high pressure that repels the water droplet. In addition, surface energy of this film is lowered by chemical passivation of the surface by hydrogen atoms/molecules present during the nucleation/growth and CH4 treatment.

2. Experimental techniques 3.2. Tribology test DNW films were deposited on silicon (100) substrates in N2 (94%)/ CH4 (6%) plasma by an MPECVD (6 in. IPLAS-CYRANNUS) system, operating with microwave power and frequency of 1200 W and 2.45 GHz, respectively. During the deposition, the chamber pressure was kept at 70 mbar and total flow rate of gases was maintained at 100 sccm. An external heater was used to heat the substrate to a temperature of 700 °C. After the deposition of nanocrystalline DNW, the film was annealed in CH4 atmosphere at 400 °C for 30 min. Surface morphology of the film was characterized by field emission scanning electron microscope (FESEM, JEOL, JSM-6500F). Atomic force microscope (AFM, Park XE100) was used to investigate the topography of the film and corresponding local deformations in wear tracks. This was also used to investigate the topography and surface roughness (Ra) of the ball counterbodies. Chemical characteristics of the film and wear tracks were investigated by micro-Raman spectroscopy and micro-XPS. Raman spectra were recorded in back scattering geometry with a Renishaw micro-Raman spectrometer (Model INVIA), equipped with an Ar-ion laser operating at a wavelength of 514.5 nm. In these measurements, 100% laser power was used for 60 min of exposure time. More importantly, micro-XPS (ESCALAB 250) was carried out in wear tracks of nanocrystalline DNW film formed against Al2O3, sapphire and ruby balls to investigate C1s and O1s core levels and corresponding chemical compounds. Friction and wear behavior of films were measured by ball-on-disk tribometer (CSM, Switzerland) operating in a linear reciprocating mode. Al2O3, sapphire and ruby balls with 6 mm diameter were used as sliding counterbodies against nanocrystalline DNW film. The normal load and sliding speeds were kept constant at 5 N and 3 cm/s, respectively. A stroke length of 3 mm was used in each experiment. The tests were carried out in ambient (dry and unlubricated) atmospheric conditions with a relative humidity level of 66±3%. Each tribological experiment was performed three times and the data were found to be approximately similar. Wear profile after the test was measured by a Dektak 6M stylus profiler using 5 mg load with a scanning speed of 30 μm/s. In this method, tip of diamond stylus with a radius of curvature 12.5 μm was scanned across the wear track. Nanoindentation measurements were performed with a diamond Berkovich indenter at a loading–unloading rate of 2 mN/min. This was performed up to a maximum load of 2 mN. Oliver and Phar method was used to calculate the elastic modulus and hardness of the specimen [12]. Maximum penetration depth was kept less than 1/10 of the DNW film thickness to avoid the substrate effect during measurements. Similar parameters were used to measure the E and H of the ball used for tribology measurements. Measurements were repeated 5 times on each sample and no significant deviation was observed among experimental data.

Friction behavior of the nanocrystalline DNW film sliding against Al2O3, sapphire and ruby balls along with their wear profiles is shown in Fig. 2. Friction coefficient of DNW/Al2O3 sliding system is extremely low and its saturated value after a 300 m shows ~0.003. However, this value in DNW/sapphire and DNW/ruby is ~0.06 and ~0.07, respectively. These are more than one order of magnitude higher compared to DNW/ Al2O3 sliding system. However, the trend in friction behavior measured in the sliding systems of DNW/Al2O3 and DNW/sapphire is similar. Conversely, such trend is not observed in the sliding system of DNW/ruby. The wear profile follows the trend of friction coefficient as shown in the inset of Fig. 2. Extremely low wear depth ~20 nm was observed in the sliding system of DNW/Al2O3 after 500 m sliding distance at 5 N normal load. This increases to ~45 nm and ~200 nm in the sliding systems DNW/sapphire and DNW/ruby, respectively. In all the cases, wear depth is much less than the film thickness which clearly indicates that sliding occurs well within the DNW film. Therefore, it is confirmed that there is no effect of substrate in friction and wear measurements. On the basis of wear track dimension, wear rate was calculated which is extremely low ~2.8 × 10−21 mm3/Nm in DNW/Al2O3 sliding system. On the basis of data reported so far, this is the highest wear resistance of

Fig. 1. Surface morphology and contact angle of nanocrystalline DNW film.

8

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

Fig. 2. Friction coefficient of nanocrystalline DNW films sliding against (a) Al 2 O 3 , (b) sapphire and (c) ruby balls. Inset shows the wear track profile of corresponding wear tracks formed against these balls.

any film ever measured for any sliding systems. However, wear rates increase to ~ 4 × 10− 19 mm 3 /Nm and 2.8 × 10− 15 mm 3 /Nm in DNW/sapphire and DNW/ruby sliding pairs, respectively.

3.3. AFM analysis It is shown that AFM topographical features of the balls are similar in nature (Fig. 3). Size of the particles and surface roughness are also more or less similar in all the balls. However, particles are densely distributed in Al2O3 and ruby balls compared to sapphire. Average surface roughness Ra is ~ 8 ± 2 nm, ~ 10.2 ± 2.1 nm and ~ 6.5 ± 1.2 nm in Al2O3, sapphire and ruby balls, respectively. In the inset of AFM images, corresponding contact angles measured on the ball surface are given. High contact angle ~90° is observed on Al2O3 ball. However, this value is consistently decreased to ~62° and ~34° on sapphire and ruby ball, respectively. Such behavior of contact angle cannot be directly explained by the variation of topography and surface roughness alone because no trend exists between roughness vs. contact angle data. Therefore, there will be some other factor to influence such a behavior of contact angle which will be investigated below.

Moreover, topography of film surface and corresponding wear tracks formed against Al2O3, sapphire and ruby balls were observed by AFM (Fig. 4). For this purpose, several images from the film surface and different locations of wear tracks were obtained and they were found to be similar. Topography of the nanocrystalline DNW film is shown in Fig. 4a. In this image, the nanowire feature is clearly visible and it is typically marked by an arrow. The surface roughness Ra value of this film is 22 ± 2 nm. It is observed that each nanowire contains several diamond nanocrystallites [9,14]. However, the sheath of the nanowire is constituted by sp2C_C and a-C phase constituent. This fact was well studied by HRTEM analysis [9]. Nanowire features of the nanocrystalline DNW film are deformed in the wear track while sliding against Al2O3, sapphire and ruby balls, as shown in Fig. 4b–d. Deformation of nanowire sheath was observed after exposure to hydrogen and oxygen plasma treatment [9,10]. Deformation of sheath is driven by chemical energy when ionic hydrogen and oxygen recombine with sp2C_C and a-C sites of nanowire sheath leading to etching off of these constituents. However, in the present study, mechanically driven energy acts to deform the nanowire features due to multi-cycle sliding of ball against nanocrystalline DNW film. Due to this process, nanowire like features disappear and nanograins are clearly seen in the corresponding image when the Al2O3 ball slides against nanocrystallite DNW film (Fig. 4b). In this image, deformation of the grains is not observed and roughness is decreased to 18 ± 2 nm. However, more deformed wear tracks are observed when the sapphire ball slides against the nanocrystalline DNW film (Fig. 4c). In this image, local plastic deformation of the grains is observed as shown by an arrow mark. Such plastic yielding acts to decrease the surface roughness to 12 ± 1.3 nm. Further, more deformation is observed when the ruby ball slides against the film (Fig. 4d). In this case, large areas of the grains are plastically deformed resulting in significant decrease in roughness to a value of ~ 5 nm. Such a low valued roughness is due to the presence of plastically polished nanodiamond grains. 3.4. Raman spectra The micro-Raman spectrometer was used to investigate the chemical characteristics of the film surface, wear tracks and corresponding ball scar regions. In this respect, Fig. 5 shows characteristic features of DNW film [9,16]. Lorentzian fits were used to describe the peak

Fig. 3. AFM topography of (a) alumina, (b) sapphire and (c) ruby ball and behavior of water contact angle on these balls.

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

9

Fig. 4. AFM topography of (a) nanocrystalline DNW film surface and wear track formed on the same film using (b) alumina, (c) sapphire and (d) ruby balls.

parameters. Peaks assigned as ν1 and ν3 at 1130 cm−1 and 1517 cm−1, respectively, originated from C\H bending and C_C stretching modes of trans-polyacetylene (t-PA or poly-CHx) [17–19]. Shift of ν3 band to higher wave number indicates the presence of stain in C_C stretching mode. The t-PA is an alternate chain of sp2C_C carbon atoms in which a single hydrogen is bonded to each carbon atom. This occurs in the grain boundaries of the nanocrystallites and sheath of the DNW film. These peaks are the signature of nanodiamond phase [18,19]. Furthermore, the pair of peaks assigned as D and G bands at 1350 cm−1 and 1592 cm−1, respectively, correspond to a-C/sp2C_C bonded carbon. The G band has E2g symmetry which appears due to bond-stretching vibration of a pair of sp2C_C sites in the form of olefinic chains or aromatic rings [17,20]. The D band belongs to A1g breathing mode of a six-fold aromatic ring which is activated by the presence of disorder. Furthermore, micro-Raman spectra were obtained from the wear tracks of nanocrystalline DNW film and corresponding deformed ball scars to investigate the tribo-induced chemical changes. In this respect, Fig. 5a–c shows deformed wear track of nanocrystalline DNW film sliding against Al2O3 ball. Corresponding Raman spectra obtained from different locations (a), (b) and (c) have more or less similar spectral characteristics. Moreover, these features are nearly similar as compared to the virgin nanocrystalline DNW film. However, the FWHM of D and G bands obtained from wear tracks is narrow as compared to that of virgin DNW film. This indicates increase in cluster size [17,18]. These spectral behaviors explain the absence of many chemical changes in the particular wear track due to less dissipation of frictional energy. It is investigated that chemical changes and corresponding phase conversion are possible when threshold mechanical/thermal energy is adequate [21]. However, corresponding Raman spectral features such as peak position, intensity and shape on the scar of Al2O3 ball are quite different as compared to these parameters obtained from wear track (Fig. 6). In the location (a), the D and G bands are observed at 1336 cm−1 and 1500 cm−1, respectively. The shift towards lower wavenumber is due to strain [17] which is acquired in tribological process. In location (b), the t-PA peaks such as ν1 and ν3 are present at 1110 cm−1 and 1428 cm−1, respectively. The D and G bands are observed at 1329 cm−1 and 1538 cm−1, respectively. These spectral features correspond to more like a-C

constitutes on locations (a) and (b) of ball scar. This is clear evidence that the transfer layer does not contain nanocrystalline DNW features on the Al2O3 ball scar. Several other peaks observed in the spectra are related to functional groups containing molecular oxygen and water [15,22–24]. A more deformed wear track is observed when the sapphire ball is made to slide against nanocrystalline DNW film (Fig. 7). The spectra acquired from different locations (a), (b) and (c) of a particular wear track depicted altered features. Most significant is the appearance of new peak ν2, which belongs to the doubly degenerated t-PA molecule [17,18] due to tribochemical reaction. These peaks, observed at 1218 cm− 1, 1220 cm−1 and 1217 cm− 1, belong to locations (a), (b) and (c), respectively. In all these locations belonging to wear track, an increase in G band intensity is observed which signifies enhancement in sp3C\C bonding [17,20]. This is further related to deformation induced wear loss of softer phase with sp2C_C/a-C bonding. Moreover, enhancement in sp3C\C fraction favors increase in t-PA phase. This happens when the alternate chains of sp2C_C carbon atoms in which single hydrogen is bonded to each carbon atom are increased. Such features are related to not only an increase in intensities of ν1 and ν3 peaks but also to the appearance of ν2 band. In all the locations of wear track, the Raman spectral characteristics are more or less similar. On the scar of the sapphire ball, the Raman spectral behavior has changed with locations (Fig. 8). This ball is found to deform more as compared to the Al2O3 ball. On the edge of the ball scar, in location (a), the ν1, ν2 and ν3 peaks including D and G bands are observed. Intensity of the G band is enhanced as compared to the Al2O3 ball scar (Fig. 6). At high wave number, two well developed peaks at 1787 cm− 1 and 1906 cm−1 belong to the molecular oxygen and water molecules [15, 22,24]. This is a typical characteristic of oxidation of ball scar. The abovementioned bands disappear in location (b) where a large amount of wear debris is found to be present. The spectra in this location are typical to DNW film where ν1, ν2 and ν3 peaks occur at 1120 cm−1, 1215 cm− 1 and 1445 cm− 1, respectively. The D, G and long chained polyethylene peaks are observed at 1328 cm−1, 1537 cm−1 and 1657 cm− 1, respectively [17,20]. However, in the center of the ball scar, at location (c) where deformation is significant, the presence of

10

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

Fig. 5. Micro-Raman spectra inside the wear track formed on nanocrystalline DNW film using alumina ball at locations (a), (b) and (c).

D, G, molecular oxygen and water bands is similar to location (a). This is direct evidence that explains the transfer layer formation in highly deformed region of the ball scar. Chemical characteristics of transfer layer indicate that its constituents belong to the sp3C\C and sp2C_C/ a-C phases as evident from the peak positions and shapes. The widest and highly deformed wear track is observed while the ruby ball was made to slide against nanocrystalline DNW film (Fig. 9). The spectral behavior in this wear track is comparatively similar to the wear track formed against the sapphire ball but there are slight changes in peak position, shape and intensity. In this particular track, Raman characteristics are more or less similar in all the locations (a), (b) and (c). In these locations ν1, ν2 and ν3 including D and G bands are observed which have small difference in peak positions, shapes and intensities. The I(D)/I(G) ratio is rather low 1.18 as compared to the virgin DNW film 1.37. This ratio is also high at 1.32 and 1.34 in locations (a) and (b), respectively of the wear track formed against the Al2O3 ball. This is characteristic of enhancement in nanocrystalline DNW phase. Consequently, the significantly deformed and fractured ruby ball scar was observed (Fig. 10). Raman spectral behavior in the scarred locations is different as compared to the corresponding wear track. In the more deformed locations (a) and (c), the intensity of D and G lines is strong.

This shows more transfer layer formation on the highly deformed scarred locations. However, at the edge of the scar, location (b) where wear debris is present shows weak behavior of these spectral lines. Several other strong peaks belong to molecular oxygen and water molecules which is an evidence of oxidation of abrasive wear debris [15,22–24]. In contrast, t-PA peaks are present in all the locations. Moreover, these peaks become strong in locations (a) and (b) due to oxidation. Chemical characteristic of transfer layer on the scar of the ruby ball is more or less similar to the scar of the sapphire ball. Development of such transfer layer is related to high surface energy of the sapphire and ruby balls and their high chemical affinity towards carbon atoms/molecules. This is also evident from the contact angle measurement (Fig. 3). 3.5. X-ray photo-electron spectroscopy For a more detailed chemical analysis of the wear tracks formed against the Al2O3, sapphire and ruby balls, the XPS analysis was carried out. In this respect, C1s and O1s spectra obtained from different tracks differed significantly. Firstly, broad C1s spectra deconvoluted into components indicated by A, B, B*, C, D and E, corresponding to specific chemical bonds present in nanocrystalline DNW film (Fig. 11). It is clearly

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

11

Fig. 6. Micro-Raman spectra inside the wear scar formed on alumina ball at locations (a) and (b).

shown that the binding energy peak positions of several components differ from spectrum to spectrum. A component with a binding energy of 284.1–284.2 eV refers to the sp2C_C bonding, typical for graphite, and belongs to grain boundaries and sheath of the DNW matrix [25]. Furthermore, chemical component B with binding energy of 284.8–284.9 eV corresponds to the sp3C\C bonding typical for nanocrystalline diamond. The ratio of sp2C_C/sp3C\C is similar and significantly high on the virgin and wear track formed by DNW/Al2O3 sliding pairs. A high amount of sp2C_C is available in such film because the sheath of the DNW including the nanocrystalline diamond grain boundaries contains a-C and graphitic contents [9]. The results are supported by the Raman spectra of the wear track formed with sliding pair of DNW/Al2O3 (Fig. 5). The results are similar to the virgin film surface. However, such chemical ratio is significantly decreased in the wear track formed with sliding pair of DNW/sapphire and DNW/ruby (Fig. 11). This is possibly related to more wearing of the sp2C_C phase due to its softness. Such chemical changes in the respective wear tracks are also evident by Raman spectra. The energy range 285.3–285.9 eV indicated by components B* and C signifies formation of carboxylic and hydroxyl group of C\COO and CH3COH chemical

compounds, respectively. Peak position and shape of such chemical components differ from track to track which signifies differences in wear track's chemical reactivity and oxidation state. Carboxylic C\COO group [26,27] is observed on the wear track formed by sliding pair of DNW/sapphire. However, formation of hydroxylic group CH3COH is absent in this particular wear track. Furthermore, this group is present on the virgin film surface and wear track formed in sliding pairs of DNW/Al2O3 and DNW/ruby balls. Components D and E, with binding energy in the range of 286–282.2 eV refers to molecular fragments containing C\COO, CH3COH and CH2\O. Such compounds form when chemically activated carbon reacts with the atmospheric moisture containing H2O molecules. These chemical compounds are negligibly present on the virgin film surface but absent on the wear track formed by sliding pair of DNW/Al2O3 due to the lower chemical reactivity of the particular wear track. In contrast, these chemical compounds are observed on the wear tracks of DNW/sapphire and DNW/ ruby due to high chemical affinity towards carbon atoms. The details are summarized in Table 1 showing energy component of C1s and the corresponding chemical components on the virgin film surface and wear track formed against the Al2O3, sapphire and ruby balls.

12

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

Fig. 7. Micro-Raman spectra inside the wear track formed on nanocrystalline DNW film using sapphire ball locations (a), (b) and (c).

Consequently, Fig. 12 shows O1s spectra from the surfaces of wear tracks formed against the Al2O3, sapphire and ruby balls. In addition, O1s spectra from the DNW film surface are also given for comparison with wear tracks. Broad spectra of O1s are deconvoluted into four individual peaks. In this context, it is mentioned that O1s and C1s spectra from the surfaces of nanoscale catalysts on graphite have been studied in a number of papers [28–30]. Based on these data, the component A with binding energy of 530.5–531.0 eV observed in the spectra of all three tracks is attributed to carbonate structures. Intensity counts of such structures are high in the wear track formed by DNW/sapphire and DNW/ruby sliding pairs. The binding energy shifts to 285.3–285.9 eV, if oxygen interacts with carbon atoms. Moreover, energy component of the B at 531.7–531.9 eV is attributed to weakly bounded molecular oxygen having covalent bonding characteristic. Such oxygen is related to hydroxylic groups [28] or typical molecular oxygen such as superoxide, peroxide and ozonide. In the C1s spectrum this peak corresponds to component D. Component C indicates the presence of OH groups with binding energy in the range of 532.5–533.0 eV. Finally, component D with binding energy from 533.6 to 534.5 indicates the presence of graphite network apart from OH groups and H2O molecules. Summary of O1s spectra and the corresponding components of chemical

network are given in Table 2. Table 3 shows ratio of various chemical compounds such as sp2_C/sp3C\C and O/C ratio on the virgin and wear tracks formed with sliding pairs DNW/Al2O3, DNW/sapphire and DNW/ruby. sp2_C/sp3C\C ratio is high on the virgin film surface and wear track formed by the Al2O3 ball. However, this ratio decreases in the sliding pairs of DNW/sapphire and DNW/ruby. The component C/B which is the ratio of CH3COH/sp3 and C\COO/sp3 is higher with values 0.66 and 0.47 in the tracks formed by sliding pairs DNW/sapphire and DNW/ruby, respectively. However, these are less at 0.12 and 0.24 in the virgin film surface and wear track of DNW/Al2O3 ball. Similarly, O/C ratio is high in wear track formed by sliding pairs of DNW/sapphire and DNW/ruby. This is significantly less in virgin film surface and wear track of DNW/Al2O3 sliding pair. Such less chemical compounds in the wear track of DNW/Al2O3 are due to less chemical reactivity. Evidence of chemical reactions and formation of several chemical compounds on the film surface and in wear track of DNW film formed by the Al2O3, ruby and sapphire balls was analyzed from O1s spectra. In this respect, Table 4 shows the relative values of the peak areas of components A, B, C and D (Fig. 12). These data partially overlapped with the data presented in Table 3. For example, component C of C1s spectrum and component A of O1s spectrum characterize similar

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

13

Fig. 8. Micro-Raman spectra inside the wear scar formed on sapphire ball at locations (a), (b) and (c).

chemical networks (C\COO, CH3COH). However, chemical relationships defined by C1s and O1s spectra do not coincide completely, because O1s spectrum contributes OH groups, water, etc., without forming chemical bonds with the carbon of the DNW film. Therefore, they may not be reflected in the fine structure of the C1s spectrum. At the same time, these chemical bonds are modified in the wear track formed by these balls. In this respect, transformation of chemical bonds related to water/OH groups is observed (Table 4). A relative amount of water molecules (peak D) decreases while OH groups (peak C) increase in the wear track formed by Al2O3 ball compared to the virgin film surface. A molecule described by peak B is weakly chemisorbed on the surface and intensity of peak A comprising C\COO, CH3COH is not found to change much. Wear track on DNW film formed by ruby and sapphire balls also results in further increase of the surface molecules containing OH group (peak C) by reducing the molecular structures of water (peak D), while reducing the chemisorbed species (peak area B). In general, areas of peaks such as C

and D of the wear tracks formed by ruby and sapphire balls increase resulting in decreases in chemisorbed molecules (peak B). In addition, it is necessary to take into account that the ratio of O/C also increases in the case of wear tracks formed by ruby and sapphire balls. Interestingly, the bonds C\COO, CH3COH, defined by O1s spectra are not found to change while the analysis of C1s spectra shows a significant increase in such components in the wear track formed by ruby and sapphire balls, respectively (Table 3). We attribute this increase to the increase of O/C ratio. Since tribo-test is performed in an ambient atmosphere, the influence of moisture on the chemical state of DNW wear track cannot be ignored. The sapphire and ruby balls contain a small amount (2–3%) of Cr and Mg atoms, respectively. Magnesium can be recorded by Mg1s (Ebinding = 1305 eV), Mn2s (Ebinding = 90 eV) and Mg2p (Ebinding = 51 eV) lines. However, Cr can be registered as Cr2p3/2, 1/2(Ebinding = 577 (2p3/2) and 586 (2p1/2) eV, Cr3p (Ebinding = 45 eV) lines. However, the lines of these elements are not visible in Figs. 13 to 15. The insets in these figures show a large region of the

14

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

Fig. 9. Micro-Raman spectra inside the wear track formed on nanocrystalline DNW film using ruby ball at locations (a), (b) and (c).

spectra with binding energies of 0 to 200 eV. In Fig. 13, Al2p, Si2p and Si2s are clearly observed. These peaks can still be discerned, but they are weak in Figs. 14 and 15. Si2p and Si2s peaks were derived from the silicon substrate. Al2p and Al2s peaks on the wear track originate from the transfer of wear particles from the Al2O3 ball. It should be noted that the transfer particles from the sapphire and ruby balls contain less Al and Si, than in the case of the Al2O3 ball. This is evident from the fact that the spectral lines Al2p, Al2s, and Si2p, Si2s are hardly distinguishable from the background. Therefore, it is difficult to expect the appearance of lines in the overview spectrum of Cr and Mg content from the sapphire and ruby balls, respectively. 3.6. Tribology mechanism In the present work, friction mechanism of CH4 treated DNW film is studied using their different allotropes of alumina balls. In this regard, results show extremely low valued friction coefficient and less wear of DNW film sliding against Al2O3 ball (Fig. 2). However, as-deposited DNW film (without CH4 treated DNW film) showed high friction

coefficient due to high energy of sliding interfaces [9]. In the present study, friction coefficient and wear are increased while sliding against sapphire and ruby balls. Such a behavior is explained by the microstructure and chemical nature of the sliding interfaces. Firstly, the influence of microstructure is described below. It is shown in Table 5 that elastic modulus (E), hardness (H) and H3/E2 value are high in DNW film as compared to the Al2O3, sapphire and ruby balls. Sliding system of DNW/Al2O3 is considered as hard/soft combination where E and H values of the Al2O3 ball are significantly less compared to DNW film. Less deformation and wear were observed in the DNW film and scar formed on the Al2O3 ball. However, sliding combinations become slightly harder using sapphire and ruby balls, but these combinations (DNW/ sapphire and DNW/ruby) can also be considered as hard/soft film/ball sliding interfaces. Comparatively, high deformation and wear were observed in these sliding interfaces including high friction coefficient. In contrast, high friction coefficient was observed when soft ball was used to slide against DNW film [11]. From this observation, microstructures such as H, E and H3/E2 values fail to explain such difference in friction behavior. In this context, the chemical nature of sliding interfaces is

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

15

Fig. 10. Micro-Raman spectra inside the wear scar formed on ruby ball at locations (a), (b) and (c).

important in describing the friction behavior. Raman spectra obtained from the wear track formed against Al2O3 ball showed not much change in spectral parameters as compared to virgin film surface (Fig. 5). This means less deformation of wear track as evident from the optical images of the sliding interfaces. In this condition, carbonaceous transfer layer on the scar of Al2O3 ball was found to be significantly less as evident from Raman spectra (Fig. 6). Such behavior is explained by less chemical affinity of Al2O3 ball against the carbon atoms. It is seen that in the case of Al2O3 ball, the contact angle is high indicating less chemical affinity towards water. This is an indication that weak chemical interaction may exist between Al2O3 ball and DNW film which attributes less dissipation of frictional energy. All the three balls used for friction measurement are α-alumina in which a small fraction of the Al3 + ions is replaced by Cr3 + ions. Each Cr3 + is surrounded octahedrally by six O2− ions. In Al2O3 ball, the impurity constitutes only Cr3+ ion. The results are supported by the mechanism proposed by Cui et al. in DLC films which shows less friction coefficient in the absence of carbon

transfer layer on the ceramic balls such as Al2O3 and ZrO2 [2]. In the absence of transfer layer, sliding occurs between film and Al2O3 ball. This could be related to low surface energy and less deformation of Al2O3 ball. The role of transfer layer formation using SiC ball sliding against UNCD films is well described by Radhika et al. [31]. The chemical nature of C1s and O1s spectra observed by XPS in this particular wear track did not show much change as compared to the virgin film surface (Figs. 11 and 12). In addition, formation of functional group and O/C ratio is less in the particular wear track (Table 3). However, chemical changes are significant as observed by Raman and XPS in the wear tracks and corresponding scars formed on the sapphire and ruby balls, as discussed in the previous section. Carboxylic, hydroxylic and O/C ratios are high in these wear tracks due to high surface energy. Such a high surface energy arises due to fractured wear track resulting in unsaturated carbon bonds. Such carbon bonds form covalent networking in the counterface and increase the adhesive interactions resulting in an increase in friction [21]. This causes formation of carbonaceous transfer layer on the

16

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

Fig. 11. C1s XPS spectra inside the wear track formed on nanocrystalline DNW film using (a) Al2O3, (b) sapphire and (c) ruby balls; (d) spectra obtained from virgin film surface.

corresponding balls which is quite significant as observed by Raman spectra (Figs. 8 and 10). It is shown that the formation and breaking of covalent bonds during sliding result in the development of transfer layer. If a covalent bond forms, friction increases and local temperature originating from the interface decreases [21,32]. Deposition of carbonaceous transfer layer is caused by high chemical affinity of sapphire and ruby balls towards carbon atoms. High chemical affinity was possibly derived from additional impurities present in sapphire and ruby balls. In sapphire 0.18 at.% of copper impurity is present. However, in ruby, the amount of titanium and Mg is 0.67 at.% and 0.02% at.%, respectively. This could be related to high surface energy of the sapphire and ruby balls as evident from the low contact angle measurement (Fig. 3). High surface energy causes strong interaction between the sliding surfaces by which more frictional energy dissipates. Such energy results in deformation and development of chemically activated carbonaceous transfer layer. In such a condition, sliding occurs between fractured film Table 1 Binding energy of C1s and corresponding chemical networking lines on virgin surface and in wear track formed against (a) Al2O3, (b) sapphire and (c) ruby balls. Components

Chemical network

Background surface

Al2O3

Ruby

Sapphire

A B B* C

sp2C_C sp3C\C – C\COO CH3COH CH2\O CH2\O

284.1 284.9 – 285.8

284.2 284.8 – 285.5

284.1 284.8 – 285.6

284.2 284.9 285.3 –

286.0 –

– –

286.5 288.7

286.2 287.2

D E

Fig. 12. O1s XPS spectra inside the wear track formed on nanocrystalline DNW film using (a) Al2O3, (b) sapphire and (c) ruby balls; (d) spectra obtained from virgin film surface.

and carbonaceous transfer layer forms on the sapphire and ruby balls. Due to interaction between the fractured surface, the dimension of ball scar becomes larger and deformation is also high on the sapphire and ruby balls. These are shown in optical images in Figs. 8 and 10. Strong chemical bonds form due to high chemical reactivity at sliding interfaces that highly resists sliding and increases the friction and wear. Other aspects that increase the friction in sapphire and ruby balls could be related to the formation of a large amount of polar chemical compounds such as CH3COH, C\COO, OH and H2O in the sliding interface resulting in higher ratio of O/C in corresponding wear tracks. These compounds can raise the forces like hydrogen bonding and capillarity [32,33] and increase the friction and wear. Table 2 Binding energy of O1s and corresponding chemical networking lines on virgin surface and in wear track formed against (a) Al2O3, (b) sapphire and (c) ruby balls. Components Chemical network

Background Binding energy of O1s surface line

A

530.6

530.2 530.5 530.8

531.9

531.7 531.6 531.8

533.0 533.7

533.0 532.5 532.7 534.5 533.6 534.0

Binding energy of C1s line

Al2O3 Ruby

B

C D

C\COO CH3COH Weakly bonded oxygen atom in the form of superoxide, peroxide and ozonide OH groups OH groups, H2O

Sapphire

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

17

Table 3 Ratio of sp2/sp3 and C\COO(CH3COH)/sp3 (carboxyl and hydroxyl groups to sp3 diamond) on the virgin and wear track formed by Al2O3, sapphire and ruby balls. System

Ratio of chemical components

DNW/Al2O3

A/B sp2/sp3 C/B CH3COH/sp3 C\COO/sp3 A/B sp2/sp3 C/B CH3COH/sp3 C\COO/sp3 A/B sp2/sp3 C/B CH3COH/sp3 C\COO/sp3 A/B sp2/sp3 C/B CH3COH/sp3 C\COO/sp3

DNW/sapphire

DNW/ruby

Virgin film surface

Ratio of O/C concentration 4

0.03

0.24

1.5

0.13

0.66

2.9

0.07

0.47

4

0.03

0.12

Fig. 14. Overview of spectrum obtained from the wear track formed by sapphire ball.

Table 4 Percentages of the various chemical compounds in the wear track and virgin film surface defined by O1s spectra. Components

A B

C D

Chemical network

C\COO CH3COH Weakly bonded oxygen atom in the form of superoxide, peroxide and ozonide OH groups OH groups, H2O

Relative area of O1s spectra on the virgin film surface (%)

Relative area of O1s spectra in wear track (%) Al2O3

Ruby

Sapphire

12.9

10.4

10.7

14.1

53.3

54.7

41.7

43.8

17.0 16.8

28.4 6.5

35.5 12.1

33.5 8.6

4. Conclusions Fig. 15. Overview of spectrum obtained from the wear track formed by ruby ball.

Tribological properties of CH4 treated nanocrystalline DNW film were investigated in ambient atmosphere using Al2O3, sapphire and ruby balls as sliding counterbodies. Super low friction coefficient ~ 0.003 and

extremely high wear resistance ~2.8 × 10−21 mm3/Nm were measured when the Al2O3 ball slides against the film. Such a low value of wear rate is being reported for the first time. Significant improvement in tribological properties is related to the low surface energy of the Al2O3 ball as corroborated from contact angle measurements. Deformation of the sliding interfaces such as film/Al2O3 ball was marginal and the absence of carbonaceous transfer layer was observed on the Al2O3 ball scar. This is related to weak chemical affinity of Al2O3 towards carbon atoms. In this case, no chemical change in the wear track could be detected by Raman spectroscopy and XPS. However, comparatively high values of friction coefficients 0.05 and 0.07 were measured for sapphire and ruby balls, respectively. Increase in friction coefficient increases the wear rate to ~4 × 10−19 mm3/Nm and 2.8 × 10−15 mm3/Nm in DNW/sapphire

Table 5 Hardness, elastic modulus and H3/E2 value of nanocrystalline DNW film, and Al2O3, sapphire and ruby balls.

Fig. 13. Overview of spectrum obtained from the wear track formed by alumina ball.

Material

H (GPa)

E (GPa)

H3/E2

DNW Al2O3 Sapphire Ruby

43 17 19 23

478 372 410 440

3.47 0.03 0.04 0.06

± ± ± ±

3 2 1.8 2.2

± ± ± ±

12 7 8 11

18

R. Radhika et al. / Diamond & Related Materials 48 (2014) 6–18

and DNW/ruby sliding pairs, respectively. Sliding interfaces such as film/ sapphire and film/ruby were found to be highly deformed and carbonaceous transfer film was formed on these balls. This is related to high surface energy and strong chemical affinity of sapphire and ruby balls towards carbon atoms. In this condition, sliding interfaces exhibit high energy due to covalent carbon bonds that chemically interact and enhance sliding resistance. Enhancement in sliding resistance leads to pronounced chemical changes in severely deformed sliding interfaces as evident from Raman and XPS measurement. Prime novelty In the present manuscript, we investigated the tribological properties of CH4 treated diamond nanowire film for the first time. Films show super low friction coefficient ~0.003 and extremely high wear resistance ~2.8 × 10−21 mm3/Nm against Al2O3 ball. This lowest value of wear rate is the first to be reported in any materials so far. Scientific novelty is systematically justified in the present paper. Technically, such material is useful for several real applications in machine tools and mechanical components where low friction coefficient and high wear resistance are desirable for energy efficiency and component's longer life. Acknowledgments The authors would like to sincerely thank Dr. R. Pandian for SEM and Ashok Bahuguna for the measurement of wear track depth profile. References [1] B. Shi, O.O. Ajayi, G. Fenske, A. Erdemir, H. Liang, Tribological performance of some alternative bearing materials for artificial joints, Wear 255 (2003) 1015–1021. [2] L. Cui, Z. Lu, L. Wang, Toward low friction in high vacuum for hydrogenated diamondlike carbon by tailoring sliding interface, ACS Appl. Mater. Interfaces 5 (2013) 5889–5893. [3] N. Sharma, N. Kumar, S. Dhara, S. Dash, A. Bahuguna, M. Kamruddin, A.K. Tyagi, B. Raj, Tribological properties of ultra nanocrystalline diamond film-effect of sliding counterbodies, Tribol. Int. 53 (2012) 167–178. [4] S.Y. Luoa, J.K. Kuo, B. Yeh, J.C. Sung, C.W. Dai, T.J. Tsai, The tribology of nanocrystalline diamond, Mater. Chem. Phys. 72 (2001) 133–135. [5] A.R. Konicek, D.S. Grierson, P.U.P.A. Gilbert, W.G. Sawyer, A.V. Sumant, R.W. Carpick, Origin of ultralow friction and wear in ultra nano-crystalline diamond, Phys. Rev. Lett. 100 (2008) 235502–235506. [6] K.J. Sankaran, N. Kumar, J. Kurian, R. Radhika, H.C. Chen, S. Dash, A.K. Tyagi, C.Y. Lee, N.H. Tai, I.N. Lin, Improvement in tribological properties by modification of grain boundary and microstructure of ultrananocrystalline diamond films, ACS Appl. Mater. Interfaces 5 (2013) 3614–3624. [7] N. Kumar, R. Radhika, A.T. Kozakov, K.J. Sankaran, S. Dash, A.K. Tyagi, N.H. Tai, I.N. Lin, Humidity-dependent friction mechanism in an ultrananocrystalline diamond film, J. Phys. D. Appl. Phys. 46 (2013) 275501–275508. [8] A.R. Konicek, D.S. Grierson, A.V. Sumant, T.A. Friedmann, J.P. Sullivan, P.U.P.A. Gilbert, W.G. Sawyer, R.W. Carpick, Influence of surface passivation on the friction and wear behavior of ultrananocrystalline diamond and tetrahedral amorphous carbon thin films, Phys. Rev. B 85 (2012) 155448–155460. [9] K.J. Sankaran, N. Kumar, H.C. Chen, C.L. Dong, A. Bahuguna, S. Dash, A.K. Tyagi, C.Y. Lee, N.H. Tai, I.N. Lin, Near frictionless behavior of hydrogen plasma treated diamond nanowire films, Sci. Adv. Mater. 5 (2013) 1–12.

[10] R. Radhika, N. Kumar, K.J. Sankaran, R. Dumpala, S. Dash, M.S.R. Rao, D. Arivuoli, A.K. Tyagi, N.H. Tai, I.N. Lin, Extremely high wear resistance and ultra-low friction behavior of oxygen plasma treated nanocrystalline diamond films, J. Phys. D. Appl. Phys. 46 (2013) 425304–425313. [11] K. Panda, N. Kumar, K.J. Sankaran, B.K. Panigrahi, S. Dash, H.C. Chen, I.N. Lin, N.H. Tai, A.K. Tyagi, Tribological properties of ultrananocrystalline diamond and diamond nanorod films, Surf. Coat. Technol. 207 (2012) 535–545. [12] W.C. Oliver, G.M. Pharr, An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments, J. Mater. Res. 7 (1992) 1564–1583. [13] L. Sun, J. Gong, D. Zhu, Z. Zhu, S. He, Diamond nanorods from carbon nanotubes, Adv. Mater. 16 (2004) 1849–1853. [14] I.I. Vlasov, O.I. Lebedev, V.G. Ralchenko, E. Goovaerts, G. Bertoni, G.V. Tendeloo, V.I. Konov, Hybrid diamond–graphite nanowires produced by microwave plasma chemical vapor deposition, Adv. Mater. 19 (2007) 4058–4062. [15] N. Shang, P. Papakonstantinou, P. Wang, A. Zakharov, U. Palnitkar, I.N. Lin, M. Chu, A. Stamboulis, Self-assembled growth, microstructure, and field-emission highperformance of ultrathin diamond nanorods, ACS Nano 3 (2007) 1032–1038. [16] R. Arenal, P. Bruno, D.J. Miller, M. Bleuel, J. Lal, D.M. Gruen, Diamond nanowires and the insulator–metal transition in ultrananocrystalline diamond films, Phys. Rev. B 75 (2007) 195431–195441. [17] A.C. Ferrari, J. Robertson, Resonant Raman spectroscopy of disordered, amorphous, and diamondlike carbon, Phys. Rev. B 64 (2001) 75414–75427. [18] R.H. Pfeiffer, K.P. Knoll, S. Bokova, N. Salk, B. Gunther, Evidence for transpolyacetylene in nano-crystalline diamond films, Diamond Relat. Mater. 12 (2003) 268–271. [19] H. Kuzmany, R. Pfeiffer, N. Salk, B. Gunther, The mystery of the 1140 cm−1 Raman line in nanocrystalline diamond films, Carbon 42 (2004) 4911–4917. [20] A.C. Ferrari, J. Robertson, Interpretation of Raman spectra of disordered and amorphous carbon, Phys. Rev. B 61 (2000) 14095–14107. [21] J.D. Schall, G. Gao, J.A. Harrison, Effects of adhesion and transfer film formation on the tribology of self-mated DLC contacts, J. Phys. Chem. C 114 (2010) 5321–5330. [22] K. Sugimoto, H. Fujiwara, S. Koda, Dynamical structure of water by Raman spectroscopy, Fluid Phase Equilib. 144 (1998) 323–330. [23] Y. Tominaga, A. Fujiwara, Y. Amo, Dynamical structure of water by Raman spectroscopy, Fluid Phase Equilib. 144 (1998) 323–330. [24] H. Cybulski, J. Sadlej, On the calculations of the vibrational Raman spectra of small water clusters, Chem. Phys. 342 (2007) 163–172. [25] Y.K. Chang, H.H. Hsieh, W.F. Pong, M.H. Tsai, F.Z. Chien, P.K. Tseng, L.C. Chen, T.Y. Wang, K.H. Chen, D.M. Bhusari, J.R. Yang, S.T. Lin, Quantum confinement effect in diamond nanocrystals studies by X-ray absorption spectroscopy, Phys. Rev. Lett. 82 (1999) 5377–5380. [26] T. Xu, S. Yang, J. Lu, Q. Xue, J. Li, W. Guo, Y. Sun, Characterization of nanocrystalline diamond films implanted with nitrogen ions, Diamond Relat. Mater. 10 (2001) 1441–1447. [27] F. Fusallba, N.E.I. Mehdi, L. Breau, D. Belanger, Physicochemical and electrochemical characterization of polycyclopenta [2,1-b;3, 4-b′]dithiophen-4-one as an active electrode for electrochemical supercapacitors, Chem. Mater. 11 (1999) 2743–2753. [28] L.S. Kibis, A.I. Stadnochenko, E.M. Pajetnov, S.V. Koscheev, V.I. Zaykovskii, A.I. Boronin, The investigation of oxidized silver nanoparticles prepared by thermal evaporation and radio-frequency sputtering of metallic silver under oxygen, Appl. Surf. Sci. 257 (2010) 404–413. [29] V.I. Bukhtiyarov, A.V. Carley, L.A. Dollard, M. Roberts, XPS study of oxygen adsorption on supported silver: effect of particle size, Surf. Sci. Lett. 381 (1997) L605–L608. [30] M.G. Mason, Electronic structure of supported small metal clusters, Phys. Rev. B27 (1983) 748–762. [31] R. Radhika, N. Kumar, S. Dash, A.K. Tyagi, I.N. Lin, Anomalous behavior of loaddependent friction on ultra-nanocrystalline diamond film, Adv. Eng. Mater. (2014), http://dx.doi.org/10.1002/adem.201300515. [32] G. Zilibotti, M.C. Righi, M. Ferrario, Ab initio study on the surface chemistry and nanotribological properties of passivated diamond surfaces, Phys. Rev. B 79 (2009) 75420–75430. [33] K.B. Jinesh, J.W.M. Frenken, Capillary condensation in atomic scale friction: how water acts like a glue, Phys. Rev. Lett. 96 (2006) 166103–166106.