Tribological properties of N+ ion implanted ultrananocrystalline diamond films

Tribological properties of N+ ion implanted ultrananocrystalline diamond films

Tribology International 57 (2013) 124–136 Contents lists available at SciVerse ScienceDirect Tribology International journal homepage: www.elsevier...

3MB Sizes 4 Downloads 24 Views

Tribology International 57 (2013) 124–136

Contents lists available at SciVerse ScienceDirect

Tribology International journal homepage: www.elsevier.com/locate/triboint

Tribological properties of N þ ion implanted ultrananocrystalline diamond films Kalpataru Panda a, N. Kumar a,n, B.K. Panigrahi a, S.R. Polaki a, B. Sundaravel a, S. Dash a, A.K. Tyagi a, I-Nan Lin b a b

Materials Science Group, Indira Gandhi Centre for Atomic Research, Kalpakkam, 603102 Tamil Nadu, India Department of Physics, Tamkang University, Tamsui 251, Taiwan, ROC

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 April 2012 Received in revised form 9 July 2012 Accepted 16 July 2012 Available online 25 July 2012

Tribological properties of ultra nanocrystalline diamond (UNCD) films have chemically been modified by N þ ion implantation and subsequent annealing processes. Friction coefficient is found to be 0.15 in as-prepared film comparing to 0.09 and 0.05 in N þ ion implanted and post-annealed films, respectively. Such a modification of friction coefficient is a characteristic of the transformation of sp3 to graphitized/ amorphized sp2 bonded carbon network. Transformation of sp3 to sp2 carbon network causes conversion of higher surface energy state (hydrophilic) to lower (hydrophobic) one which results in ultra low friction coefficient. Graphitization/amorphization in wear track observed by micro Raman spectroscopy is found to be the prominent mechanism for the reduction in friction coefficient. & 2012 Elsevier Ltd. All rights reserved.

Keywords: Ultra nanocrystalline diamond Chemical structure Tribological properties

1. Introduction Understanding the composition and bonding structure of materials at the surface has fundamental scientific and technological interest in many disciplines. It is particularly interesting at the contact interface, where adhesion, friction and tribochemical reactions are influenced by the composition and bonding structures. At the nano/micrometer scales, the dominance of the surface compared to bulk renders such consideration which is essential for implementing reliable micro and nanoelectromechanical systems (MEMS/NEMS) which experience sliding contacts [1]. Crystal structure of diamond consists of two closely packed interpenetrating face centered cubic lattices. One lattice is shifted with respect to the other along the elemental cube space diagonally by one-quarter of its length. Due to the short carbon˚ diamond crystal exhibits the highest carbon distance of 1.54 A, atomic packing density. Very high bonding energy between carbon atoms and directionality of tetrahedral bonds are the reasons behind the high strength of diamond. Nanostructured specimen such as nanocrystalline and ultra nanocrystalline diamond (NCD/UNCD) films synthesized by Microwave Plasma Enhanced Chemical Vapour Deposition (MPECVD) technique has unique multifunctional properties compared to bulk diamond [2–5]. Depending on deposition parameters such as the gas mixture, temperature, pressure, and substrate seeding process,

n

Corresponding author. Tel. þ 91 4427480081. E-mail address: [email protected] (N. Kumar).

0301-679X/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.triboint.2012.07.016

MPECVD technique results in different types of nanostructured films such as microcrystalline (grain size 1–5 mm), nanocrystalline (grain size 1–50 nm) and ultra nanocrystalline (grain size E5 nm) films. These films exhibit special properties due to the small dimension of nanocrystallites. The significant increase in ratio of surface to volume atoms in these films result in distinct physical and chemical properties [6]. The application oriented interest in nanocrystalline diamond films are increasing due to its unique and specific properties such as high hardness, high elastic modulus, high stiffness, high fracture toughness, high thermal conductivity, low thermal expansion, chemical resistivity, chemical inertness, low friction coefficient and high wear resistance [7,8]. All these above properties make it an attractive material for many engineering applications including cutting tools and mechanical assembly devices which always requires high wear resistance with low friction [7–9]. Nevertheless, the friction and wear of nanostructured crystalline diamond is anisotropic and depends on the chemical reactivity of the surface, crystallite size, crystallographic orientation, dangling covalent bonds, sliding direction of oriented planes, surface roughness, transfer layer, test environment, and test parameters such as loading and sliding velocity [10–15]. In addition to all known important properties and parameters of these nanocrystalline diamond surfaces, which influence the tribological behavior, the amount of hydrogen and the sp3/sp2 bonding ratio play significant role on the friction and wear [16]. Nitrogen ion implantation can be used to tailor the sp3/sp2 ratio and formation of sufficient amount of sp2 bonded network by properly selecting the dose and energy of N þ ions [17]. This behavior was mainly studied on MCD and NCD films, which includes

K. Panda et al. / Tribology International 57 (2013) 124–136

efforts to improve the field emission properties. Presently, there is no report which deals with the tribological properties of nitrogen ion implanted UNCD films. Ion implantation to UNCD films causes desorption of hydrogen from the hydrocarbon network present at the grain boundaries resulting in depassivation of dangling bonds. Nitrogen incorporation into the carbon network is found to promote the sp2 phase fraction of carbon bonding, most likely due to the inherent sp2 hybridization of the N-electrons. This in turn promotes low friction coefficient, better wear resistance, durability, and reduced internal stresses [18]. Present study aims to explore the possible reasons for the improvement in tribological properties of nitrogen ion implanted/ post-annealed UNCD films. The improved tribological properties of N þ ion implanted/post-annealed UNCD films are explained in terms of surface conductivity and surface energy state by STS and contact angle measurements, respectively. Detailed structural characteristics have been analyzed by various analytical techniques which explain the nanomechanical and tribological behavior of N þ implanted/post-annealed UNCD films.

2. Experimental method UNCD films were grown on n-type silicon substrates in a MPECVD system (IPLAS-Cyrannus) [19]. UNCD films were deposited on silicon (Si) substrates using CH4 (1%)/Ar plasma for 3 h with a microwave power of 1200 W. The pressure and the flow rate were maintained at 150 Torr and 100 sccm, respectively. The growth process was carried out at low temperatures ( o475 1C) without any intentional heating of the substrate. A 150 kV ion implanter was used to implant 75 keV nitrogen ions with a fluence of 5  1015 ions cm  2 on the UNCD films, at room temperature and a pressure below 2  10  7 mbar. After implantation, the films were annealed at 600 1C in N2 atmosphere for 30 min. The chemical bonding structures were investigated by XPS using SPECS make photoelectron spectrometer which uses monochromatic Al Ka radiation at 1486.74 eV as a probe. Local electrical conductivity measurements were performed using a commercial UHV-STM (150 Aarhus, SPECS GmbH, Germany) working at a base pressure of 10  10 mbar. The imaging was performed with a set current of 0.59 nA with a sample bias between 2.5 and 3.5 V. The current–voltage (I–V) spectra are obtained during scanning and the data presented are the average of 12 reproducible spectra acquired in subsequent scans. Nanoindentation measurements (CSM Instruments, Switzerland) were performed with a diamond Berkovich indenter with a loadingunloading rate of 4 mN min  1. This was performed up to a maximum load of 1.4 mN. Oliver and Pharr method [20] was used to calculate the elastic modulus and hardness of the films. Scratch tests were performed using a Revetest scratch tester (CSM Instruments, Switzerland). A spheroconical diamond indenter with a radius of curvature of 200 mm was used as the scratching element. Normal load was applied progressively from 1 to 10 N and scratch length was kept constant at 3 mm. Rotational mode of a ball on disk nanotribometer, NTR2 (CSM Instruments, Switzerland) was used to carry out tribological tests. Contacting 100Cr6 steel ball sliding against the static film is 1.5 mm in diameter. Sliding speed of the film against the ball and normal load were kept constant at 1.5 cm s  1 and 5 mN, respectively. Total sliding distance for each measurement was 8 m. In situ wear track depth was measured by LVDT sensor coupled with a nanotribometer. Tribological experiments were conducted in ambient atmospheric condition with relative humidity of 42%. The surface morphology and wear tracks on the UNCD films were analyzed using a Field Emission Scanning Electron Microscope (FESEM, carl zeiss supra 55) and an OXFORD Energy Dispersive

125

X-ray Spectrometer (EDX). The Raman spectra on the surface and wear track were recorded in back scattering geometry using 514.5 nm line of an Ar-ion laser using Renishaw micro-Raman spectrometer (Model INVIA). The contact angle of UNCD films was measured by sessile drop method with a Kruss EasyDrop contact angle instrument (EasyDrop DSA 100). The volume of the water droplets used was l mL. All measurements were carried out at room temperature and atmospheric pressure with relative humidity of  50%. Standard deviations of the measurements were typically 721. Wettability of the film surface was evaluated by measuring static contact angles for water. For simplicity, asprepared, N þ ion implanted and post implanted annealed films are designated as UNCD(I), UNCD(II) and UNCD(III), respectively.

3. Results and discussion 3.1. Surface morphology of UNCD film examined by FESEM FESEM micrographs [Fig. 1(a)–(c)] show significant change in surface morphology of UNCD film after the N þ ion implantation/ post annealing processes. Microstructure of the as-prepared films resemble large agglomerates of diamond grains in UNCD(I). Agglomerates are composed by fine diamond grains. Further, it can be noticed that such structure disappeared and a smoother surface appeared after the N þ ion implantation [UNCD(II)]. It seems that smaller particles agglomerated in the as-prepared films coalesced to form a finer and homogeneous layer of grains. Cauliflower-like morphology appears after the post-annealing process in [UNCD(III)]. 3.2. XPS analysis of UNCD films XPS measurements are carried out to understand how N ion implantation/annealing processes attribute to the chemical bonding state of these UNCD films. The C1s photoemission spectrum of UNCD(I), UNCD(II) and UNCD(III) films are shown in Fig. 2. Nitrogen ion implantation influences the sp2/sp3 ratio and affects the degree of ordering of sp2 bonding, inducing formation of amorphous carbon network [21]. The data are fitted with Lorentian peaks with binding energies at 284.4, 285.1, 286.0 and 286.2 eV corresponding to sp2 C ¼C, sp3 C–C, C¼N and C–N bonds, respectively. Relative intensities of these peaks are tabulated in Table 1. The background was subtracted using Shirley’s method [22]. The measurements were conducted with ion beam sputtering to avoid the surface residual contamination. In UNCD(I), sp3 C–C bonding is predominant at 285.1 eV with a peak intensity of 60.1% while sp2 C¼C present at 284.4 eV with a peak intensity of 37.8%. UNCD(I) film has negligible amount of nitrogen content, which may come during the synthesis process while it increases considerably in N þ ion implanted, UNCD(II) [Fig. 2(b)] and post-annealed, UNCD(III) [Fig. 2(c)] films. In UNCD(II), sp3 C–C peak intensity decreases to 44.5% and sp2 C¼C peak intensity increases to 42.4%. It is due to the implantation induced conversion of sp3 to sp2 bonding [17]. In particular, the ratio of Csp3/Csp2 decreases from 1.58 in UNCD(I) to 0.94 and 0.81 in UNCDII) and UNCD(III) films, respectively. It confirms the implantation induced graphitization/amorphization in case of N þ implanted and post-annealed films. For covalently bonded solids, high dose ion implantation usually leads to amorphization. However, for diamond, an additional level of complexity is involved since the bonds broken by ion impact may rearrange to form stable sp2 carbon network [17]. The N þ ion implanted films clearly shows the presence of C¼N bonding. This indicates that the chemical bonding between C and N atoms happen during the implantation process. However, the boundary phase as C–H bonding is absent in N þ implanted and post-annealed films. In

126

K. Panda et al. / Tribology International 57 (2013) 124–136

Fig. 1. FESEM surface morphology of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films.

Fig. 2. XPS spectra of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films.

UNCD(III), sp2 C¼ C peak intensity increases further to 53.2% with the decrease in sp3 C–C peak intensity to 38.2%. This is attributed to thermally driven transformation of sp3 to sp2 phase. It is also known that the thermal stability of sp2 is higher than sp3 bonding phase. In UNCD(III) films, the C¼ N peak intensity decreases compared to UNCD(II). Decrease in intensity of C ¼N peak upon annealing is due to the lower thermal stability of C¼N bonds [23]. In post-annealed films, there will be loss of nitrogen and rearrangement of carbon atoms to form sp2 C–C bonds.

3.3. STM measurement of UNCD films The local electrical properties of the films were investigated by scanning tunneling spectroscopy (STS) to know more about the surface conductivity of these films after the N þ ion implantation/ post-annealing processes. N þ ion implantation introduces defects and sp2 phases and expels hydrogen from the grain boundaries and makes the film more conducting [24]. Introduction of defects may also change the grain size. Grain size is found to decreases in

K. Panda et al. / Tribology International 57 (2013) 124–136

implanted films [Fig. 3, UNCD(II)] as compared to as- deposited films [Fig. 3, UNCD(I)]. After annealing it is found to increase [Fig. 3, UNCD(III)]. Surface smoothening upon post-annealing occurs due to the graphitization of the surface. Smoothening is also affected by decrease in grain size in implanted films. For the characterization of the local electronic properties of UNCD(I), UNCD(II) and UNCD(III) films, I–V curves are taken at various sample positions, as on the grains and grain boundaries as shown in Fig. 3. The tunneling current under positive bias is lesser than that under negative bias implying that the films have n-type conductivity [24]. Only positive portion of the I–V curves are shown in this figure. However, the surface conductivity significantly enhanced in N þ ion implanted and post-annealed films than as- prepared one. Moreover, we observe a significant change in conductivity at the grain and grain boundaries for both UNCD(II) and UNCD(III) films compared to UNCD(I). We observe that in each case grain boundaries (gb) are more conducting compared to the grains (g). In addition, the conductivity gets significantly enhanced both at the grain and grain boundaries upon post implantation annealing [UNCD(IIIg)] and [UNCD(IIIgb)] films as compared with those for as implanted ones [UNCD(IIg)] and [UNCD(IIgb)] as shown in Fig. 3. This value is lowest in

Table 1 XPS results of UNCD(I), UNCD(II) and UNCD(III). Peak position (eV)

284.4 285.1 286.0 286.2

Chemical bonding

2

sp C ¼C sp3 C–C C¼ N C–N

Peak intensity (%) UNCD(I)

UNCD(II)

UNCD(III)

37.8 60.1 – 2.1

42.4 44.5 13.1 –

53.2 38.2 8.6 –

127

UNCD(I) both at grain and grain boundaries. The defects and sp2 phase produced by nitrogen ion implantation can introduce various defect levels at the diamond grain [24,25]. Interestingly, from the I–V curves, we found that the band gap for the grain boundary is much lesser compared to the grains. It is due to the presence of a mixture of sp2, amorphous carbon and nitrogen phases at the grain boundary [24,25], which will have higher conductivity. Post-implantation annealing reduces the defects, along with the conversion of amorphous carbon phase to sp2 one. The annealing process facilitates the diffusion of a certain fraction of nitrogen from diamond grains to the grain boundaries. Increase in conductivity upon annealing can be due to the introduction of sp2 phase in the diamond films [26]. Furthermore, the conductivity of the grain boundary showing nearly metallic behavior [UNCD(IIIgb)] indicating that it is mostly of sp2 content. After N þ ion implantation grain boundaries appear to be more susceptible to amorphization and graphitization giving rise to wider grain boundaries [24]. 3.4. Raman spectroscopy of UNCD films The Raman spectroscopy is a reliable analytical technique to measure local chemical structure of carbon based materials. The Raman spectra were de-convoluted using multi-peak Lorentzian fitting method as shown in Fig. 4. The spectra shows characteristic feature of UNCD films. A peak at 1369 cm  1, designated as D band that represents zone-edge A1g mode due to the disorder and a comparatively broad G peak centered around 1543 cm  1 corresponding to sp2 network is seen in all these films. These features are thoroughly studied by Ferrari et al. [27] Details of Raman spectra are presented in Table 2. Diamond Dn peak around 1331 cm  1 is observed in visible Raman spectra. According to Ferrari and Robertson, the D peak is due to the breathing modes belonging to A1g symmetry which involves phonons near the K

Fig. 3. Local I–V measurements on the surface of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films. The corresponding STM images showing the respective I–V curves of the UNCD films are indicated. I–V curve at the grain ‘g’ and grain boundary ‘gb’ is presented. The representation of grain ‘g’ and grain boundary ‘gb’ is also given in the corresponding STM images.

128

K. Panda et al. / Tribology International 57 (2013) 124–136

Fig. 4. Raman spectra of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films. Lorentzian peak fitting has been used and peak assignments are described in the corresponding text.

Table 2 Raman spectroscopic results of UNCD(I), UNCD(II) and UNCD(III). No

[Dn(cm  1)]

[D (cm  1)]

[G (cm  1)]

[I(D)/I(G)]

[FWHM(G) (cm  1)]

(a) (b) (c)

1331 – –

1374 1327 1353

1543 1527 1576

0.3 0.6 1.14

110 118 29

(a)UNCD(I) (b) UNCD(II), (c) UNCD(III).

zone boundary while the G peak is due to the bond stretching of all pairs of sp2 atoms present in both rings and chains [28,29]. In this context, UNCD(I) shows a G peak centered at 1543 cm  1 indicating prevalence of ordered sp2 phase. The broadening of the G peak in UNCD(II) film is related to the disordering of sp2 clusters due to N þ ion implantation which ultimately causes amorphization of the graphitic phase. Peak broadening is also related with internal stress, sp2 admixtures and defects produced due to the N þ ion implantation. N þ ion implantation also changes the bonding characteristics of the carbon atoms in UNCD films. In UNCD(III), the FWHM(G) is found to narrow compared to UNCD(I) and UNCD(II), which indicates the formation of ordered sp2 clusters. Broad FWHM(G) arises due to bond length and bond angle disorder in sp2 based aromatic rings. Such local structural alterations like sp2 clustering and ordering always shift the G peak position [28]. UNCD(III) exhibits highest G band shift (1576 cm  1), and this may be related with ordering of sp2 clusters and formation of crystalline graphite. In contrast to the G peak dispersion, the I(D)/I(G) ratio and the D peak shows maximum dispersion for microcrystalline and nanocrystalline graphites. This dispersion decreases with increasing disorder. The integrated intensity ratio [I(D)/I(G)] of the D and G peaks conventionally indicates the degree of long range ordering in

clustered aromatic sp2 bonding [27]. The D peak indicates disordering in graphitic phase but ordering of an amorphous carbon structure [28]. Transition from nanocrystalline graphite to graphite or vice versa will occur due to ion irradiation and post annealing of implanted UNCD films. The in-plane correlation length or cluster size can be estimated from La¼ C(l)/[I(D) (/I(G)], where C(l) for graphite at 514.5 is 4.4 nm [30]. In this context, results show that I(D)/I(G) ratio is 0.3, 0.6 and 1.12 for UNCD(I), UNCD(II) and UNCD(III), respectively which corresponds to sp2 cluster sizes of 14.6, 7.3 and 3.92 nm, respectively. Here, it is noted that the sp2 cluster size is different from the sp3 diamond grain size. Decrease in in-plane grain size is attributed to the clustering of sp2 bonding which is triggered by the implanted N þ ions. Increase in I(D)/I(G) values and shifting of G band towards higher frequency side implies the formation of nanographitic matrix and decrease in sp3 content in amorphous carbon. Three other Raman spectra peaks such as n1, n2 (C–H in plane bending) and n3 (C¼C stretch) bands in UNCD(I), centered at 1140, 1180 and 1479 cm  1, respectively, are observed due to the vibration of trans-polyacetylene (TPA) segment present at the grain boundaries [31]. Broader features of these bands signifies disorder TPA chain while higher frequency of v1 is a typical fingerprint of transformation of TPA to transpolymeric a-C:H at the grain boundaries [31]. These bands are absent in UNCD(II) and UNCD(III) films, possibly an indication of the breaking of TPA chains to transpolymeric a-C:H due to the N þ ion implantation and post annealing processes. As TPA states are unstable at high temperatures, these features disappear when UNCD films are annealed at high temperatures [32]. This causes disappearance of Raman spectra of TPA vibrational mode and modifies relative intensity of D and G bands [33]. Further, it may be concluded that the structure of the sp2 phases and sp2/sp3 ratio are significantly modified by the N þ ion implantation and post-annealing processes.

K. Panda et al. / Tribology International 57 (2013) 124–136

129

Fig. 5. Contact angle measurement of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films.

3.5. Surface free energy and wettability of UNCD films The wettability of a film surface is influenced by the surface chemical composition, surface energy, texture and morphology [34]. Wettability of carbon materials is sensitive to the sp2/sp3 bonding ratio and formation of C–N polar radicals. In particular, the surface energy of the basal plane of graphite is relatively lower than that of the sp3 plane of the diamond surface [35,36]. Cleavage and softness of the sp2 bonding is also known to reduce the surface energy state [37]. Presence of C–N and high amount of sp3 bonding in UNCD(I) may reduce the significance of hydrophobic behavior. N þ ion implantation and post annealing processes change the surface chemical state, structure and morphology of the film surface and have different influence on the wettability. In UNCD(II) and UNCD(III), C¼N bond has high polarity which causes to reduce the contact angle. However, high fraction of sp2 bonding reduces the surface energy and surface electronegativity and hence CA increases to 1111 and 1471 in UNCD(II) and UNCD(III), respectively, as shown in [Fig. 5(b) and 5(c)]. CA of 1471 is known as sub superhydrophobic behavior.

4. Mechanical properties of UNCD films 4.1. Nanoindentation Nanoindentation is a useful technique to characterize the mechanical properties of films [38]. In this technique, loaddisplacement curves, considered as mechanical fingerprints of materials, are recorded. The curves obtained for UNCD(I), (II) and (III) films are shown in Fig. 6. Shift of load–displacement curves towards lower penetration depths in UNCD(I) is observed. A increase in maximum penetration depth, at the same indentation load, points towards a decrease in penetration resistance of UNCD(II) and UNCD(III) films. The ratio of the elastic work to total work, characterizes the elastic fraction of the total work done during an indentation cycle [38]. A larger value of this ratio indicates that the material is stiffer which in turn indicates the higher hardness of the material. In this context, UNCD(I) shows highest hardness (H) and elastic modulus (E) as 28 and 610 GPa, respectively. These values decreases to 21 and 482 GPa in UNCD(II) films. In UNCD(III), these values are minimum as 16 and 378 GPa, respectively. The key parameters in bonding of

Fig. 6. Nanoindentation load-displacement curves of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films.

UNCD films are the sp3/sp2 content, clustering and orientation of the sp3/sp2 phase. These characteristics lead to influence the H and E values of the films [39]. The decrease in H and E values is related to the increase in sp2/sp3 bonding ratio. However, the sp3 content is directly related with the hardness and elastic constants of UNCD films [39–41]. As it is evidenced from XPS and Raman spectroscopy that sp3/sp2 fraction is higher in UNCD(I) where the H and E values are higher. 4.2. Micro scratch test of UNCD films The scratch test was performed under progressive loading to determine the scratch resistance of the films. In Fig. 7, the applied normal load [Fn (N)] vs. resulting friction coefficient and acoustic emission [A(E)] peaks are displayed as a function of the scratch track length. It is seen that the critical load for failure of the film UNCD(II) and UNCD(III) occurs after 5 N of normal force. However, UNCD(I) shows adhesive failure after 2.5 N which may be due to residual stress at the interface which causes to break the interlocking of the bonds. This implies that the film is damaged by interfacial spallation and ploughing at critical load. The trend in friction coefficient during scratch test matches with the evolution of AE curves. This value is low up to  0.08 in UNCD(III) films up

130

K. Panda et al. / Tribology International 57 (2013) 124–136

Fig. 7. Scratch resistance test of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films, optical images are shown with the corresponding film.

Fig. 8. Coefficient of friction recorded on (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) film surfaces. In inset of this figure, the evolution of the coefficient of friction during the initial passes is shown. Optical images of ball sliding face are shown. In inset friction coefficient of initial passes are drawn.

to a normal load of 3 N, after that this increases upto 0.6. In UNCD(I) and UNCD(II), friction coefficients are found to be higher compared to UNCD(III). Low value of friction coefficient is caused by elastic deformation, while high values points to the formation of cracks, fracture and spallation of the films. The onset of film spallation is observed by the rapid change in acoustic response as well as by visual inspection of the scratch. Similar behavior was studied by Shane et al. and Buijnsters et al. [42,43]. The fluctuation of the friction coefficient and acoustic emission can be attributed to the trapping of the broken sharp asperities from the films. At low region of normal load small scale cracks and partially distributed shell shaped spallation are observed in all the films. This region in graph is indicated as D1, which corresponds to D1 in the optical image. However, large scale fracture and interfacial spallation are observed in different regions of the films during the loading as shown by D2. 4.3. Tribological properties of UNCD films Coefficient of friction is found to be 0.15 in UNCD(I) while this values reduce to 0.09 and 0.05 in UNCD(II) and UNCD(III), respectively, as shown in Fig. 8. The corresponding micrograph of ball contact surfaces are shown in inset of this figure.

Corresponding wear tarck morphologies are shown in Fig. 9. Severly deformed film surface and formation of wider wear tracks are seen in UNCD(I) [Fig. 9(a)]. Generally, softer the film, the greater becomes the contact area as the number of revolutions increases. This mechanism does not hold if the film is under high stress. UNCD(I) film has high hardness and elastic modulus which also contains high compressive stress. By applying external stress during the tribo test, the film severly deforms and partially peels off from the substrate. During the course of sliding, diamond surface asperities are worn down and trapped in the wear contact, leading to abrasive wear of the steel ball. Magnified region of the wear track represents closer view of the deformed surface and nature of abrasive wear. This type of wear mechanism is possible, when interface has high stress instead of poor interfacial adhesion of film with substrate. Insitu wear depth of ball counterface as measured against sliding distnace during the tribology test is shown in Fig. 10. This clearly demosntrate that the total penetration depth is 196 nm [Fig. 10(a)] which corresponds to a high wear rate of 1.2  10  9 mm3 N m  1 in UNCD(I) as shown in inset of Fig. 10. In UNCD(II) [Fig. 10(b)], the film has less deformed but deformation of ball is severe as magnified portion represents clear morphology of metallic oxide [Fig. 9(b)]. In this case, adhesive and abrasive wear mechanism is dominated by ball counterface. Ball has lower hardness and elastisity while the film is hard with high elastic mudulus which resist for plastic deformation and wear of the film during sliding. In UNCD(II), penetration depth of ball and wear rate is found to 178 nm and 4.8  10  11 mm3 N m  1. In UNCD(III), deformation of the wear tarck is partial and minor [Fig. 9(c)]. Formation of fine abrasive particle of the ball and metallic oxide is visible over the track. In this condition, penetration depth is negligible [Fig. 10(c)] and a high wear resistance 2.8  10  11 mm3 N m  1 is obtained. It is known that the deformation and wear rate not only depends on the hardness and elasticity of the UNCD film but also on the contribution of sp2/sp3 ratio, hydrogen concentration and surface energy [39,40]. High sp2 concentration in UNCD can reduce the internal stress which leads to deform the specimen less. This in turn lowers the surface energy and hence enhances the slip behavior. Contact area of ball is observed to be less in UNCD(III) which increases in UNCD(I) and UNCD(II). The wear rate of the sliding ball is calculated by the method proposed by Qi et al. [44]. Calculated wear rate of ball is found to be high 3.8  10  6 mm3 N m  1 in ball sliding against UNCD(I). However, this values are 5.6  10  7 and 1.2  10  7 mm3 N m  1 in

K. Panda et al. / Tribology International 57 (2013) 124–136

131

Fig. 9. FESEM surface morphology of wear track formed on (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) film surfaces sliding with steel ball. Inset shows the magnified images of the corresponding films.

Fig. 10. Sliding distance vs penetration depth of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films, in inset penetration depth and wear rate of these films are shown. Geometry of sliding ball is also shown in this figure.

UNCD(II) and (UNCD(III), respectively. From these results it is found that the wear loss from the ball surface is higher than compared to UNCD film surface. High wear rate of ball due to plugging and micro abrasion is possible if the film has high hardness and elastic modulus compared to the ball. Several physical, chemical, contact mechanical and topographical phenomena are involved for analyzing the coefficient of friction of UNCD films. As it is explained by XPS [Fig. 2] that sp3/sp2 bonding ratio is dominated in UNCD(I) film, which is found to decrease in UNCD(II) and UNCD(III) due to transformation of sp3 into sp2 bonding. Increase in sp2/sp3 bonding ratio and improved crystalline phase in UNCD(II) and UNCD(III) compared

to UNCD(I) is examined by Raman spectroscopic results [Fig. 4]. It has been shown that, higher volume fraction of grain boundaries with sufficient amount of a-C:H easily passivates the dangling s bonds present in the nanocrystalline diamond film, causing ultra low friction and extremely low wear in ambient atmospheric conditions [45]. Concentration of hydrogen is high in UNCD(I) due to formation of TPA chain observed by XPS and Raman spectroscopy. In this specific film, mechanism of chemical passivation works, but due to high stress in the film the deformation is easy during tribo test by which free active sites of dangling bonds get enhanced which cause to increase the adhesion. High conductive specimen can provide an additional channel for energy dissipation besides the excitation of phonons, giving rise to an electronic contribution to friction [46]. The increase in conductivity in N þ ion implanted and post-annealed films is due to the transformation of sp3 to sp2 bonding network of carbon atoms in grain boundary which act as a lubricant phase. Intrinsically, high conductivity of the grain and grain boundaries are responsible for reduction in the electrostatic adhesion [47]. This is phenomenologically valid for UNCD(II) and UNCD(III). If the conductivity of the diamond surface is low, this can generate tribo induced charging and hence locally increase the electrostatic interaction of the charged particles leading to increase the electrostatic component of adhesion [47,48]. Interaction of electrostatic adhesion is known to be stronger compared to adhesive force, generated by the electronic excitation of phonon. In this context, electrostatic component of friction coefficient is high in UNCD(I). Besides the reduced electronic vibration of phonons in low conductive medium, the accumulation of charge may produce the tribo-plasma which leads to increase the adhesive force at sliding surfaces. This has been extensively studied by Nakayama [49]. It has also been demonstrated from STS study that conductivity increases in UNCD(III) and UNCD(II) compare to UNCD(I). However, in each specimen, conductivity at the grain boundary is high compared to grain [Fig. 3].

132

K. Panda et al. / Tribology International 57 (2013) 124–136

Fig. 11. EDAX analysis of the wear track formed in (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films.

Fig. 12. Raman spectroscopy of wear track on films (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III).

K. Panda et al. / Tribology International 57 (2013) 124–136

Surface roughness and microstructure of the films have significant effect on the friction and wear of diamond, nanocrystalline diamond and DLC films [50]. Measured value of surface roughness in UNCD(I), UNCD(II) and UNCD(III) by STM are 8, 3 and 6 nm over a measured area of 0.5  0.5 mm2, respectively. Even high roughness value of UNCD(III) shows low magnitude of initial friction. This is related with the softness of surface asperities having more sp2 bonding and low surface energy. Effect of surface roughness acts in the initial cycles and after a few cycles surface gets smoothened. During the initial few passes, maximum value of friction coefficient is observed due to surface roughness of UNCD(I) and UNCD(II) as shown in inset of Fig. 8. For smooth surface of UNCD(III), high friction is not observed. Replication of these trends was excellent. Microstructure is found to be one of the important factors that influence the friction and wear of carbon based materials. In UNCD(II) and (III), hardness and elastic modulus are less compared to UNCD(I) film. We have discussed that hardness and elastic modulus decreases due to N þ ion implantation and subsequent post annealing. This causes internal stress relaxation in the film. However, high hardness and high elastic modulus have normally high wear resistance. In this case, UNCD(I) should show less friction and high wear resistance. But it has high friction and high wear rate due to the presence of internal and interfacial stress in the

Table 3 Raman spectroscopic results obtained from the wear tracks of UNCD films. No

[Dn(cm  1)]

[D (cm  1)]

[G (cm  1)]

[I(D)/I(G)]

[FWHM(G) (cm  1)]

(a) (b) (c)

1332 – –

1365 1351 1351

1557 1531 1596

0.82 0.56 0.62

123 92 78

(a)UNCD(I) (b) UNCD(II), (c) UNCD(III).

133

film. Despite the low hardness and low elasticity, wear tracks of UNCD(II) and UNCD(III) are not found to be severely deformed under the tribological sliding conditions. This is explained by low internal stress present in the film which resists to pill off from the substrate. In UNCD(I), high friction coefficient is also dominated by severe adhesive and abrasive wear mechanism and formation of larger sliding contact area. However, slippery behavior of surface due to high sp2 bonding lowers the adhesion between sliding surfaces. Slipperiness behavior of surface reduces sliding resistance and lowers the adhesive/abrasive wear [51]. If the adhesive strength of sliding surfaces is low, the frequent movement is possible. Track containing oxides layer is observed by FESEM and Raman spectroscopy. However, formation of severe metallic oxide is known to increase the friction in UNCD(II) film. Lower extent of deformation caused by ball sliding with UNCD(III), effectively contributes to the lower friction coefficient.

4.4. EDAX anlalysis of wear track EDAX elmental analysis of the wear tracks is shown in Fig. 11. Basically, five elements such as C, O, Si, Fe and N are observed in all these three tracks. Percentage of C is highest (85.12 wt%) in UNCD(III), [Fig. 11(c)]. However, carbon content is more or less similar in UNCD(I) and UNCD(II) films [Fig. 11(a) and Fig. 11(b)]. Tribological tests performed in the ambient atmospheric conditon which may account for the presence of oxygen constituent in the wear tracks. The wt% of O and Fe are 6.01 and 8.37 wt%, respectively. The wt% of Fe is 7.34 in UNCD(III) track. Little amount of silicon is also detected from the substrate exposing during the tribo test. Silicon may come form the matting ball, as it is known that 100Cr6 steel ball contains 0.35 wt% of silicon. However, nitrogen is high 2.11 wt% in N þ implanted [UNCD(II)] films. After anneling nitrogen reduces to 1.74 wt% in UNCD(III).

Fig. 13. Raman spectroscopy of sliding ball with (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films.

134

K. Panda et al. / Tribology International 57 (2013) 124–136

4.5. Raman spectral analysis of wear track and ball counterbody It is seen from Fig. 12 that the trend in Raman spectra of the wear tracks are similar with the spectra recorded from the surface of the films. However, the intensity of spectra such as D, G, n1, n2 and n3 peaks in wear track formed on UNCD(I) films [Fig. 12(a)] decreases significantly compared to the spectral intensity obtained from the virgin film surface [Fig. 4(a)]. Significant decrease in spectral intensity of diamond related peak such as Dn is observed in wear track of UNCD(I) films. This may correspond to short range ordering in carbon lattice and decrease in TPA chain length due to the wear of the film. It is observed that the D (1365 cm  1) and G (1557 cm  1) band positions in wear track of UNCD(I) films are shifted towards lower and higher frequencies, respectively, comparing to the virgin film surface. This behavior indicates formation of strain in sp2 and sp3 sites [27]. This may also reduce the crystalline properties of sp2 clusters. TPA related peaks such as n1 (1135 cm  1), and n3 (1470 cm  1) are also shifted towards lower frequency side which is an indication of disordered and formation of short range chains of TPA constituents in grain boundaries of UNCD(I) films. Increase in I(D)/I(G) values in wear track compared to film surface of

Table 4 Raman spectroscopic results obtained from the ball counter face. No

[D (cm  1)]

[G (cm  1)]

[I(D)/I(G)]

[FWHM(G) (cm  1)]

(a) (b) (c)

1371 1366 1358

1556 1553 1553

3.7 3.1 2.4

0.9 1.8 2.2

(a)UNCD(I) (b) UNCD(II), (c) UNCD(III).

UNCD(I), shows disorder while broader FWHM(G) reveals amorphization in the wear tracks. In case of wear tracks, formed on UNCD(II), the D and G bands shifted towards lower wave number sides due to tribochemical reaction [Fig. 12(b)]. The detail of the spectra is given in Table 3. The shift of G band towards higher frequency side is an indication of improved crystallinity of sp2 clusters in UNCD(III). This also occurs due to the strain localization in sp2 and sp3 sites during sliding [28]. Decrease of width and peak shift of G band towards higher frequency is attributed to a progressive reduction of defects such as bond-angle and bondbending disorder in the sp2 amorphous carbon cluster [28]. Low value of FWHM(G) and higher value of I(D)/I(G) is an indication of graphitization. I(D)/I(G) value is found to be low in wear track of UNCD(II) and UNCD(III) films, due to disorder accompanied by strain. In case of UNCD(III), the D band position is more or less similar to the film surface but G peak position gets shifted towards high frequency side (1596 cm  1). This indicates clustering and ordering of sp2 carbon lattice and subsequent transformation to nanocrystalline graphite [Fig. 12(c)] [28]. On the other hand, polymerization of a-C:H, may reduce the FWHM(G), consistent with the disorder. In UNCD(III) films, graphitization is a prominent feature which causes to decrease the I(D)/I(G) values to 0.62 from 1.14. This leads to the formation of larger clusters due to the tribochemical reaction. These results are characteristic of carbon with crystalline features [52]. D band on the ball surface is shifted towards the higher frequency side when sliding against the UNCD(I) and UNCD(II) films comparing to UNCD(III) films [Fig. 13]. Similarly D band position shifted towards higher frequencies for ball counter face sliding with UNCD(I) and UNCD(II) compared to D band position observed on the virgin surface of UNCD and in wear track. Here, the I(D)/I(G) value is high and FWHM(G) is narrow as shown in Table 4. Above mentioned characteristic of Raman spectroscopy points to the

Fig. 14. (I) Raman spectroscopy (200–1000 cm  1) of virgin film surfaces (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III), (II) Raman spectroscopy (200–1000 cm  1) obtained in wear track of (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films, (III) Raman spectroscopy (200–1000 cm  1) of steel ball sliding with (a) UNCD(I) (b) UNCD(II) and (c) UNCD(III) films.

K. Panda et al. / Tribology International 57 (2013) 124–136

formation of high disorder in sp2 phase on the ball counter face. This behavior concludes that a poor carbonaceous transfer layer gets formed on the ball counter face. Crystalline silicon peak at 520 cm  1 is observed only in UNCD(I) film surface and wear track [Fig. 14(I) and Fig. 14(II)], respectively. In the wear track, the intensity of this peak is found to be higher [Fig. 14(II)]. However, in other two films, UNCD(II) and UNCD(III), the silicon peak is absent. This signifies homogeneities and smoothness of the surface due to the increase in area to volume ratio of carbon atoms after the N þ implantation and post annealing processes. The Raman peaks at low frequencies occur due to the oxidation of metallic iron from the counter face materials. Prominent and broad feature in Raman spectra at 667 and 676 cm  1 points to the formation of amorphous Fe3O4 phase in wear tracks [Fig. 14(II) and Fig. 14(III)] and on counterbody sliding against the UNCD(II) and UNCD(III) films [Fig. 14(II) and Fig. 14(III)], respectively [53]. Other peaks with weak intensities at 226, 297 and 412 cm  1 correspond to the formation of amorphous a-Fe2O3 phase in UNCD(II) films [54]. Spectral characteristic of metallic oxide formed on the wear track is similar to the spectra observed on the ball surface. The origin of these peaks can be attributed to the tribochemical reaction of steel ball with the surrounding atmosphere. These oxide scales are absent on UNCD(I). This may be related to severe deformation of the film.

5. Conclusions The chemical bonding characteristic of as- deposited UNCD films are modified due to N þ ion implantation and subsequent post annealing. Csp2/Csp3 is found to be 0.81 for as- deposited film and this ratio increases to 0.94 and 1.58 while modifying the film with implantation and annealing process, respectively. This modification process decreases the hardness and elastic modulus, related to the increase in sp2/sp3 ratio in the films. Chemical structure of UNCD films changed and I(D)/I(G) value are found to low in wear tracks of modified UNCD surfaces. Hydrophilic property of as- deposited film transforms to hydrophobic and sub-super hydrophobic one after the implantation and subsequent annealing of implanted film, respectively. The chemical state, structure and microstructure of these films are well correlated with tribological properties. Synergetic combination of all the above factors and properties found to influence the friction coefficient.

Acknowledgements We thank Dr. R. Ramaseshan for nanoindentation measurements, Mr. Nanda Gopala Krishna for XPS experiments. We are thankful to Dr. S. Dhara for providing the Raman facility. Authors are grateful to Dr. C.S. Sundar, Director, Materials Science Group, IGCAR and Shri S.C. Chetal, Director, IGCAR for encouragement and support. References [1] Sumant AV, Grierson DS, Gerbi JE, Carlisle JA, Auciello O, Carpick RW. Surface chemistry and bonding configuration of ultrananocrystalline diamond surfaces and their effects on nanotribological properties. Physical Review B 2007;76:235429–39. [2] Popov C, Kulisch W, Jelinek M, Bock A, Strnad J. Nanocrystalline diamond/ amorphous carbon composite films for applications in tribology, optics and biomedicine. Thin Solid Films 2006;494:92–7. [3] Askari SJ, Chen GC, Akhtar F, Lu FX. Adherent and low friction nano-crystalline diamond film grown on titanium using microwave CVD plasma. Diamond and Related Materials 2008;17:294–9.

135

[4] Wang CS, Chen HC, Chen HF, Lin IN. Synthesis of diamond using ultrananocrystalline diamonds as seeding layer and their electron field emission properties. Diamond and Related Materials 2009;18:136–40. [5] Orlando A, Anirudha SV. Status review of the science and technology of ultrananocrystalline diamond (UNCD) films and application to multifunctional devices. Diamond and Related Materials 2010;19:699–718. [6] Kloss FR, Najam-Ul-Haq M, Rainer M, Gassner R, Lepperdinger G, Huck CW, Bonn G, Klauser F, Liu X, Memmel N, Bertel E, Garrido JA, Steinmuller-Nethl D. Nanocrystalline diamond—an excellent platform for life science application. Journal of Nanoscience and Nanotechnology 2007;7:4581–7. [7] Olszyna A, Smolik J. Nanocrystalline diamond-like carbon coating produced on the Si3N4–TiC composites intended for the edges of cutting tools. Thin Solid Films 2004;459:224–7. [8] Krauss AR, Auciello O, Gruen DM, Jayatissa A, Sumant A, Tucek J, Mancini DC, Moldovan N, Erdemir A, Ersoy D, Gardos MN, Busmann HG, Meyer EM, Ding MQ. Ultrananocrystalline diamond thin films for MEMS moving mechanical assembly devices. Diamond and Related Materials 2001;10: 1952–61. [9] Hollman P, Wanstrand O, Hogmark S. Friction properties of smooth nanocrystalline diamond coatings. Diamond and Related Materials 1998;7: 1471–7. [10] Pastewka L, Moser S, Gumbsch P, Moseler M. Anisotropic mechanical amorphization drives wear in diamond. Nature Materials 2011;10:34–8. [11] Grillo SE, Field JE, Van Bouwelen FM. Diamond polishing the dependency of friction and wear on load and crystal orientation. Journal of Physics D: Applied Physics 2000;33:985–90. [12] Konicek AR, Grierson DS, Gilbert PUPA, Sawyer WG, Sumant AV, Carpick RW. Origin of ultralow friction and wear in ultra nano-crystalline diamond. Physical Review Letters 2008;100:235502–6. [13] Zilibotti G, Righi MC, Ferrario M. Ab initio study on the surface chemistry and nanotribological properties of passivated diamond surfaces. Physical Review B 2009;79:75420–30. [14] Hayward IP. Effect of roughness on the friction of diamond on CVD diamond coatings. Wear 1992;157:215–27. [15] Luoa SY, Kuo JK, Yeh B, Sung JC, Dai CW, Tsai TJ. The tribology of nanocrystalline diamond. Materials Chemistry and Physics 2001;72:133–5. [16] Grierson DS, Carpick RW. Nanotribology of carbon-based materials. Nanotoday 2007;2:12–21. [17] Prawer S, Kalish R. Ion-beam-induced transformation of diamond. Physical Review B 1995;51:15711–22. [18] Khurshudov A, Kato K, Sawada D. Tribological and mechanical properties of carbon nitride thin coating prepared by ion-beam-assisted deposition. Tribology Letters 1996;2:13–21. [19] Joseph PT, Tai NH, Lee CY, Niu H, Pong WF, Lin IN. Field emission enhancement in nitrogen-ion-implanted ultrananocrystalline diamond films. Journal of Applied Physics 2008;103:43720–7. [20] Oliver WC, Pharr GM. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. Journal of Materials Research 1992;7:1564–83. [21] Ma KL, Zhang WJ, Zou YS, Chong YM, Leung KM, Bello I, Lee ST. Electrical properties of nitrogen incorporated nanocrystalline diamond films. Diamond and Related Materials 2006;15:626–30. [22] Janos V. The Shirley-equivalent electron inelastic scattering cross-section function. Surface Science 2004;563:183–90. [23] Li JJ, Zheng WT, Wu HH, Sun L, Gu GG, Bian HJ, Lu XY, Jin ZS. Compositional and structural modifications of amorphous carbon nitride films induced by thermal annealing. Journal of Physics D: Applied Physics 2011;36:2001–5. [24] Panda K, Sundaravel B, Panigrahi BK, Magudapathy P, Krishna DN, Nair KGM, Chen HC, Lin IN. Structural and electronic properties of nitrogen ion implanted ultra nanocrystalline diamond surface. Journal of Applied Physics 2011;110:44304–13. [25] Zapol P, Sternberg M, Curtiss LA, Frauenheim T, Gruen DM. Tight binding molecular-dynamics simulation of impurities in ultrananocrystalline diamond grain boundaries. Physical Review B 2002;65:45403–14. [26] Govindaraju N, Das D, Singh RN, Kosel PB. High-temperature electrical behavior of nanocrystalline and microcrystalline diamond films. Journal of Materials Research 2008;23:2774–86. [27] Matthews MJ, Pimenta MA, Dresselhaus G, Dresselhaus MS, Endo M. Origin of dispersive effects of the Raman D-band in disordered carbon materials. Physical Review B 1999;59:6585–8. [28] Ferrari AC, Robertson J. Interpretation of Raman spectra of disordered and amorphous carbon. Physical Review B 2000;61:14095–107. [29] Ferrari AC, Robertson J. Resonant Raman spectroscopy of disordered, amorphous, and diamondlike carbon. Physical Review B 2001;64:075414–27. [30] Ferrari AC, Robertson J. Raman spectroscopy of amorphous, nanostructured, diamond-like carbon, and nanodiamond. Philosophical Transactions of the Royal Society of London, Series A 2004;362:2477–512. [31] Kuzmany H, Pfeiffer R, Salk N, Gunther B. The mystery of the 1140 cm  1 Raman line in nanocrystalline diamond films. Carbon 2004;42:911–7. ¨ [32] Pfeiffer R, Kuzmany H, Knoll P, Bokova S, Salk N, Gunther B. Evidence for trans-polyacetylene in nano-crystalline diamond films. Diamond and Related Materials 2003;12:268–71. [33] Ballutaud D, Jomard F, Kociniewski T, Rzepka E, Girard H, Saada S. sp3/sp2 character of the carbon and hydrogen configuration in microand nano-crystalline diamond. Diamond and Related Materials 2008;17: 451–60.

136

K. Panda et al. / Tribology International 57 (2013) 124–136

[34] Azevedo AF, Matsushima JT, Vicentin FC, Baldan MR, Ferreira NG. Surface characterization of NCD films as a function of sp2/sp3 carbon and oxygen content. Appled Surface Science 2009;255:6565–70. [35] Ostrovskaya L, Perevertailo V, Ralchenko V, Saveliev A, Zhuravlev V. Wettability of nanocrystalline diamond films. Diamond and Related Materials 2007;16:2109–13. [36] Ostrovskaya LY, Dementiev AP, Kulakova II, Ralchenko VG. Chemical state and wettability of ion-irradiated diamond surfaces. Diamond and Related Materials 2005;14:486–90. [37] Huu TL, Schmitt M, Paulmier D. Diamond films control for tribological applications. Surface Science 1999;433–435:690–5. [38] Ayesha JH, Munroea PR, Hoffmana M, Martina PJ, Bendavida A. Berkovich indentation of diamond like carbon coatings on silicon substrates. Journal of Materials Research 2008;23:1862–9. [39] Ferrari AC, Robertson J, Pastorelli R, Beghi MG, Bottani CE. Elastic constants of tetrahedral amorphous carbon films by surface Brillouin scattering. Applied Physics Letters 1999;75:1893–5. [40] Bai M, Kato K, Umehara N, Miyake Y. Nanoindentation and FEM study of the effect of internal stress on micro/nano mechanical property of thin CNx films. Thin Solid Films 2000;377–378:138–47. [41] Lifshitz Y. Hydrogen-free amorphous carbon films: correlation between growth conditions and properties. Diamond and Related Materials 1996;5: 388–400. [42] Shane AC, James B, Vohra YK, Lacefield WR, Lemons JE. Nanoindentation hardness and adhesion investigations of vapor deposited nanostructured diamond films. Journal of Applied Physics 2002;91:5347–52. [43] Buijnsters JG, Shankar P, Enckevort van WJP, Schermer JJ, Meulen ter JJ. Adhesion analysis of polycrystalline diamond films on molybdenum by means of scratch, indentation and sand abrasion testing. Thin Solid Films 2005;474:186–96.

[44] Qi L, Stanishevsky A, Vohra YK. Tribological properties of undoped and boron-doped nanocrystalline diamond films. Thin Solid Films 2008;517: 800–4. [45] Sharma N, Kumar N, Sundaravel B, Panda K, David W, Kamarrudin M, Dash S, Panigrahi BK, Tyagi AK, Lin IN, Raj B. Effect of CH4/H2 plasma ratio on ultralow friction of nano-crystalline diamond coating deposited by MPECVD technique. Tribology International 2011;44:980–6. [46 Marcin K, Enrico G, Urs G, Laurent M, Simon R, Ernst M. Suppression of electronic friction on Nb films in the superconducting state. Nature Materials 2011;10:119–22. [47] Erdemir A, Donnet C. Tribology of diamond-like carbon films: recent progress and future prospects. Journal of Physics D: Applied Physics 2006;39:311–27. [48] Knorr N. Squeezing out hydrated protons: low-frictional-energy triboelectric insulator charging on a microscopic scale. AIP Advances 2001;1:22119–41. [49] Nakayama K. Triboemission of electrons, ions, and photons from diamondlike carbon films and generation of tribomicroplasma. Surface & Coatings Technology 2004;188–189:599–604. [50] Slawomir K, Jaromir P, Franciszek R. Spontaneous decrease of high surface electrical conductivity in diamond exposed to atmospheric air. Chemical Physics Letters 2004;391:56–9. [51] Song Y, Premachandran RN, Zou M, Wang YA. Adhesion and friction properties of micro/nano-engineered superhydrophobic/hydrophobic surfaces. Thin Solid Films 2010;518:3801–7. [52] Qun-feng Z, Guang-neng D, You-bai X. Influence of sulfidation treatment on the structure and tribological properties of nitrogen-doped diamond-like carbon films. Applied Surface Science 2008;254:3859–66. ¨ [53] Gasparov LV, Rush A, Guntherodt G, Berger H. Electronic Raman scattering in magnetite: spin versus charge gap. Physical Review B 2009;79:144303–7. [54] Sang-Heon S, Thomas DS. Raman spectroscopy of Fe2O3 to 62 GPa. American Mineralogist 2001;87:318–26.