- Email: [email protected]
Tribological properties of Cr–Si–N nanocomposite film adherent silicon under various environments
Thin Solid Films 518 (2010) 7509–7514
Contents lists available at ScienceDirect
Thin Solid Films j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / t s f
Tribological properties of Cr–Si–N nanocomposite film adherent silicon under various environments Hsin-Han Lin a, Chau-Chang Chou b,⁎, Jyh-Wei Lee c a b c
Institute of Nanotechnology and Microsystems Engineering, National Cheng Kung University, Tainan 70101, Taiwan, ROC Department of Mechanical and Mechatronic Engineering, National Taiwan Ocean University, Keelung 20224, Taiwan, ROC Department of Materials Engineering, Mingchi University of Technology, Taipei 24301, Taiwan, ROC
a r t i c l e
i n f o
Available online 10 May 2010 Keywords: Nanocomposite thin films Pulsed DC magnetron sputtering process Cr–Si–N films Tribological property
a b s t r a c t Chromium nitride thin films have good corrosion resistance and mechanical properties. However, their hardness is slightly lower than that of other hard coatings. The concept of nanocomposite thin films is employed by adding silicon to form Cr–Si–N thin films with enhanced hardness and wear resistance. In this study, Cr–Si–N films with various Si contents were coated on silicon wafer to enhance the tribological properties and anticorrosion by a bipolar symmetry pulsed DC reactive magnetron sputtering process. The tribological properties were studied by a pin-on-disk tester. The tests were conducted with the same operating condition under three different environments. They were performed in the ambient atmosphere (in 55% humid air), DI water, and 0.01 M NaCl aqueous solution, respectively. The wear tests revealed that, as the silicon content was increased, even though the Cr–Si–N films had a better anticorrosion property they had an inferior performance on wear resistance. The results were concluded to be mainly due to Cr–Si–N films’ microstructures and adhesion to the Si substrate rather than their hardness and toughness. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Chromium nitride (Cr–N) thin film has been widely applied in molding industries to prolong service life due to its excellent corrosion resistance and mechanical properties [1]. It also exhibits rather good adhesions when deposited on general steel substrates [2]. The Cr–N coating has also been noted for its good oxidation resistance up to 800 °C [3]. At higher temperatures, the formation of Cr2O3 on the surface can further promote the coating's wear resistance and its capability as a diffusion barrier. This is beneficial to the improvement of die casting molds [4,5]. According to the concept of superhard nanocomposite hard coatings proposed by Veprek et al. [6], the addition of silicon to the Cr–N thin film to form amorphous silicon nitride phase has been adopted to improve the mechanical properties and oxidation resistance [7–14]. It has been observed that CrN crystalline and amorphous Si3N4 phase are two major contributors in the Cr–Si–N nanocomposite thin film [8,14]. Lee and Chang [14] evaluated the mechanical properties of the Cr– Si–N films by microhardness as well as micro wear tests of atomic force microscopy (AFM). The antiwear capability and microhardness of the composite films were increased with Si contents of up to 12 at. %. The corrosion resistance of each thin film was evaluated by a potentiostat in 3.5 wt.% NaCl aqueous solution in our previous
⁎ Corresponding author. E-mail address: [email protected] (C.-C. Chou). 0040-6090/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2010.05.035
report [15], showing that it was enhanced directly by the amount of Si content of the composite film. Microwear tests in a corrosive environment (a NaCl solution) by AFM were also implemented to investigate the possible wear-corrosion synergism. It has been shown that the best wear resistance under the corrosive environment is obtained with Cr–Si–N films with a concentration of Si of 10.1 at.%, which also yields the maximum hardness of Cr–Si–N coatings. In this work, the same asymmetric bipolar pulsed DC reactive magnetron sputtering technique [15] was adopted to deposit Cr–Si–N thin films with different silicon contents on ptype (100) silicon substrates. The tribological properties of the coated samples rather than the composite thin films themselves were studied by a pin-on-disk tester. The tests were conducted with the same operating condition under three different environments to investigate the wear resistance and tribological behavior of Cr–Si–N coatings on Si substrates. 2. Experimental details 2.1. Deposition of nanocomposite films Cr–Si–N thin films were coated on a p-type (100) silicon wafer by using a bipolar asymmetry pulsed DC reactive magnetron sputtering system. Two DC power suppliers with a pulse controller (SPIK 2000A, Shen Chang Electric Co., Taiwan) were used to supply controlled power to Cr and Si targets, respectively. To obtain Cr–Si– N films with various Si contents, the powers supplied to Si target
7510
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514
Table 1 Elemental compositions of Cr–Si–N nanocomposite thin films produced by various pulse powers supplied to Si target. Code of silicon specimen
Cr target power (W)
Si target power (W)
Element content (at.%) Cr
N
Si
A1 A2 A3 A4
300 300 300 300
224 160 95 65
53.1 52.1 55.5 58.0
34.1 38.9 38.4 38.1
12.8 9.0 6.1 3.9
were kept at four values: 224, 160, 95, and 65 W, while, that to Cr target was fixed at 300 W. The purity and diameter of Cr and Si targets were both 99.99 wt.% and 76.2 mm. The pulse frequency was kept at 20 kHz with 100% reversed voltage and an 80% duty cycle during the sputtering process. A substrate bias of − 300 V was applied with a pulse unit (Sparcle V, Advanced Energies Industries, USA) with 80 kHz frequency, 15% reversing voltage and a 60% duty cycle. The base pressure, 5.6 × 10− 6 Pa, was achieved before sputtering. All substrates were sputtered clean for 10 min at 2.7 Pa argon pressure with a substrate bias of −500 V. A Cr–Si mixed film was deposited as an adhesion interlayer with around 150 nm thickness onto the substrate under 5 × 10− 3 Pa in pure Ar atmosphere. The flow rates of Ar/N2 mixture at a ratio of 1:1 were monitored by individual mass flow controllers. The pressure in the chamber during deposition was 8.0 × 10− 1 Pa. The deposition time was 165 min. All substrates were heated to 300 °C during the sputtering process. They were placed 100 mm from both targets and fixed to a holder which rotated continuously in order to obtain uniform nanocomposite thin films.
2.2. Film characterization The elemental compositions of the Cr–Si–N thin films were identified with a field emission electron probe microanalyzer (FEEPMA, JXA-8500F, JEOL, Japan) with a ZAF-corrected program. The phases of the coatings were explored with a grazing incidence (1°) X-ray diffraction (XRD) equipped with a Cu target source. The surface roughness analysis was performed by an AFM system (D3100, Veeco). A silicon-based probe (NSC15/AIBS, Micko Mash) was used to scan a 50 × 50 μm2 area of the composite thin films to derive their roughness average (Ra) and maximum peak-to-valley height (Rmax).
Fig. 2. Cross-sectional SEM images of Cr–Si–N coatings with (a) 3.9 at.% Si and (b) 12.8 at.% Si.
The mechanical properties, the hardness and reduced Young's modulus, were measured by micro indentation using a Triboindenter (Hysitron Inc.) equipped with a Berkovich pyramidal tip. The load– displacement curves were analyzed by applying the method of Oliver and Pharr [16]. The maximum load was 2.5 mN, applied at a loading– unloading rate of 1000 μN/s. Each specimen was tested more than three times. 2.3. Wear tests of Cr–Si–N samples The wear coefficients were evaluated by means of pin-on-disk tests in three environments. They were conducted in the ambient atmosphere (in 55% humid air), DI water, and 0.01 M NaCl aqueous solution at room temperature. The counter parts were cemented tungsten carbide (WC with 6.0 wt.% Co) balls of 1/4″ (6.350 mm)
Table 2 Hardness (H), reduced Young's modules (Er), and resistance to plastic deformation (H3/ E2r ) of Cr–Si–N samples with different Si contents. Code of silicon specimen
Fig. 1. XRD patterns of Cr–Si–N thin films deposited on Si substrates. (Note: A1: 12.8 at. % Si, A2: 9.0 at.% Si, A3: 6.1 at.% Si, and A4: 3.9 at.% Si).
A1 A2 A3 A4
Hardness (H), GPa
Reduced Young's modulus (Er), GPa
15.31 ± 0.26 11.54 ± 0.35 13.85 ± 1.48 8.89 ± 0.43
152.6 ± 2.1 147.6 ± 3.5 155.2 ± 7.4 131.4 ± 3.4
Resistance to plastic deformation (H3/Er2), GPa 0.15 0.07 0.11 0.04
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514 Table 3 Wear volume (Ws) and dimensional wear coefficient (WR) of Cr–Si–N samples with 12.8 at.% and 9.0 at.% Si contents. r = 1.4 mm, Fn = 0.49 N, S = 22 m Si content of Cr–Si–N film Ws (mm3)
WR (mm3 N−1 m−1)
Dry DI water 0.01 M NaCl Dry DI water 0.01 M NaCl
WS, and the dimensional wear coefficient, WR, were calculated using the equations [17] Ws =
A1 (12.8 at.% Si)
A2 (9.0 at.% Si)
6.14 × 10− 3 5.33 × 10− 3 6.63 × 10− 3 5.70 × 10− 4 4.94 × 10− 4 6.15 × 10− 4
3.49 × 10− 3 2.82 × 10− 3 3.60 × 10− 3 3.24 × 10− 4 2.62 × 10− 4 3.34 × 10− 4
7511
t 2 2 3t + 4b 2πr 6b
ð1Þ
Ws Fn ⋅S
ð2Þ
and WR =
where r is the radius of the wear track, S is the sliding distance, and Fn is the normal load. diameter. A normal load of 0.49 N was applied. The sliding speed was 0.015 m/s (100 rpm) along a circular wear track of 2.8 mm diameter. The total wear length was 22 m (2500 cycles) for each test. Worn samples were ultrasonically cleaned in alcohol for 2 min to remove wear debris after the wear tests. The surface morphologies and chemical composition of the wear scar were investigated by scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) analyzer. The depth of each wear track, t, and the width of the wear track, b, were determined by using a stylus surface profiler (SE500-58D, Kosaka lab). The wear volume,
3. Results and discussion 3.1. Composition and phases characteristics The elemental compositions of the various Cr–Si–N coatings are listed in Table 1. The Si concentration of these films increases as the sputtering power on the Si target is raised. Fig. 1 demonstrates the XRD patterns of Cr–Si–N thin films on the p-type (100) silicon wafer. For coatings with Si content of higher than 6.1 at.%, only pure CrN (200) phase is shown. On the other hand, a mixture of CrN (200),
Fig. 3. Wear tracks’ surface morphologies and EDX results of Cr–Si–N samples with (a) 9.0 at.% Si and (b) 6.1 at. % Si after pin-on-disk tests in the ambient atmosphere.
7512
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514
(220) and Cr2N (002) phases is found in coatings with Si content of less than 3.9 at.%. The phase transition range of Si concentrations, 3.9– 6.1 at.%, is higher than the values of 3.1–4.1 at.% Si, derived from the authors’ previous work (Lee and Chang [14]). The different Si contents of their Cr–Si–N thin films were obtained due to the difference of the stationary substrate's position between the Cr and Si targets. In this study, the substrates were rotating and their Si contents were changed by varying the sputtering power on the Si targets. In addition, as the substrate bias was supplied with a higher pulse frequency (80 kHz) than that (50 kHz) used in [14], the light nitride atom was much more easily removed by the higher excitation energy [18,19]. The composition of Cr–Ns in coatings with low Si content thus tended to be in the form of Cr2N instead of CrN. A similar result has been reported by Lee et al. [10]. They explain that Si plays the role as a substitutional solid solution element in the Cr–Si–N film and the extra Cr ions will react with the Cr–N phase to form the Cr2N. As the Si content is higher than the solubility limit, amorphous SiNx phase will thus produce and the Cr2N phase will disappear [10].
but those of other films with lower Si content was particulate. Among the samples, the Cr–Si–N thin film with the lowest Si content (3.9 at.%) had the highest surface roughness (2.188 nm Ra and 102.73 nm Rmax). The cross-sectional SEM images of coatings in Fig. 2 show that the Cr–Si–N thin films’ thickness was determined to be in the range of 1.2– 1.5 μm. A very dense and compact columnar structure is found on the coating with 3.9 at.% Si as shown in Fig. 2a. When the Si content increased, a featureless morphology was observed (Fig. 2b) for the film containing 12.8 at.% Si. No columnar structure could be identified by SEM anymore. The cross-sectional transmission electron microscopy (TEM) images of coatings from the previous work of Lee and Chang [14] revealed that the addition of Si reduced the width of columnar structure and further formed CrN nanocrystallites surrounded by amorphous SiNx matrix in Cr–Si–N coatings. In Fig. 1, the rather broad and low intensity reflections of CrN (200) are observed for coatings A1 containing 12.8 at. % Si, which also verifies the very small crystallite size of coatings. In the TiN/a-Si3N4 coating system [20], the crystallite size is also seen for the TiN/a-Si3N4 coating with more than 8 at.% Si.
3.2. Morphology and microstructure analysis
3.3. Evaluation of films’ toughness properties
From the surface roughness determination by AFM, Cr–Si–N thin film with 12.8 at.% Si was very smooth (1.052 nm Ra and 13.599 nm Rmax),
The values of hardness of the Cr–Si–N film, H, and reduced Young's modulus, Er, are listed in Table 2. They were measured at a depth of
Fig. 4. Wear tracks’ surface morphologies and EDX results of Cr–Si–N samples with (a) 6.1 at.% Si and (b) 3.9 at. % Si after pin-on-disk tests in the DI water.
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514
50–100 nm from the coating surface, where it was believed that no influence of substrate was included. The values of both properties were in the order of composite films with 12.8 at.%, 6.1 at.%, 9.0 at.%, and 3.9 at.% of Si in this study. The resistance to plastic deformation (H3/E2r ), or toughness of Cr–Si–N composite films, which was derived from H and Er [21,22] followed the same order as listed above. The main difference between Lee and Chang [14] and this work is the trend of the Cr–Si–N films’ hardness with different Si contents. When the Si content is higher than 1.6 at.% in [14], the hardness increases with increasing Si concentration and a maximum hardness is achieved for coatings with around 10.1 at.% Si and a minimum grain size around 5.2 nm, caused by the grain boundary hardening, i.e., the Hall–Petch relation [7,20]. However, in this work the phase transition range of Si concentrations is higher and the grain boundary hardening is found only in the range of around 6.1–9.0 at.% Si. Different coating setups, e.g. the variable impulse powers and rotating holder, could account for the different microstructures of the films with the same Si content. 3.4. Tribological results The wear volume, WS, and dimensional wear coefficient, WR, of Cr– Si–N samples under three environments are shown in Table 3. Due to the resolution of the adapted stylus profiler, only coatings with
7513
12.8 at.% and 9.0 at.% Si contents were measured and evaluated. It can be seen that the coated samples with higher hardness had higher wear coefficients. This is contrary to the micro tribological behavior of the composite films [14]. The values of WR indicate that the wear coefficient of the coated specimens can be reduced from the dry contact to a lubrication state by the de-ionized (DI) water. However, when the corrosive 0.01 M NaCl aqueous solution was implemented, the wear coefficients increased and were even higher than the ones tested in the atmosphere. To investigate the wear mechanism involved in the pin-on-disk test, SEM and EDX analysis of the wear track region were performed. SEM observation of the worn surface of each coating under three test environments reveals that, when tested in the atmosphere, the failure of the coatings occurred for Cr–Si–N films with Si contents higher than 6.1 at.% (Fig. 3). While the samples were tested in DI water and 0.01 M NaCl aqueous solution, the films with Si contents higher than 3.9 at.% failed (Figs. 4 and 5). The wear resistance of the composite films on Si substrates was thus found to be the highest when the Si content was 3.9 at.% and declined as the Si content increased. Among these tests, the amorphous SiNx phase of the Cr–Si–N films with higher Si contents is believed beneficial to reduce the friction on the surface when conducted in the ambient atmosphere. This also has been reported from the discussions of friction coefficient in [8,14]. When the wear tests were carried out in
Fig. 5. Wear tracks’ surface morphologies and EDX results of Cr–Si–N samples with (a) 6.1 at.% Si and (b) 3.9 at. % Si after pin-on-disk tests in the 0.01 M NaCl aqueous solution.
7514
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514
DI water or NaCl aqueous solution, the friction on the wear tracks was reduced mainly by the lubrication of water. The influence of the shear stress beneath the contact surface could thus extend to the interface of the coating and the substrate, which resulted in the failure of Cr–Si–N coatings with Si contents only higher than 3.9 at.%. This is inconsistent with the results of micro wear test in [14], which concluded that a composite film with higher hardness would reveal a lower wear coefficient. It should be notified that the wear tracks in [14] did not penetrate through the coatings and only the antiwear capability of the composite film itself was considered. Nevertheless, the results obtained from pin-on-disk tests in this work are the same as those of [21], which reported that the hardness and toughness of the Cr–Si– N films dominate the coatings’ antiwear performance. However, in this work, the higher hardness and toughness of the composite films did not guarantee better wear resistance of the coatings. As depicted in Figs. 3a, 4a, and 5a, the wear track from various environments show that the cracks and grooves induced by abrasive wear were more significant under a higher wear coefficient, e.g., coatings with 12.8 at.% Si (Fig. 3a). EDX results show that the wear debris still attached inside the wear track was all heavily oxidized. According to the analysis of [21], this can be attributed to Cr2O3. The worn surfaces, as shown in Figs. 3b, 4b, and 5b, illustrate that the element compositions inside and outside the wear tracks remained nearly unchanged and free of oxide as long as the films were sustained on the substrate. Compared to the large wear coefficient of the failed coatings, it is summarized that the hard wear debris accelerates the wear coefficient when the films are removed from the wear tracks. 4. Conclusion Cr–Si–N composite thin films with Si contents ranging from 3.9 to 12.8 at.% were coated on a p-type (100) silicon wafer by a bipolar symmetry pulsed DC reactive magnetron sputtering process. The tribological properties of the coated samples were studied by pin-on-disk tests conducted under the same operating condition in three different environments. For Cr–Si–N composite films with the same Si content, the wear coefficient obtained in DI water was the lowest. Wear tests in 0.01 M NaCl aqueous solution show more severe wear than did those in the atmosphere. All wear coefficients increased as the Si content of the coating increased. The amorphous SiNx phase of Cr–Si–N composite films was beneficial to reduce the friction while
rubbing in ambient atmosphere. However, in DI water and NaCl aqueous solution, the lubrication of water was suggested to decrease the friction and extend the influence of the shear stress beneath the contact surfaces. The adhesion between the films and substrate, which could be enhanced by the columnar microstructure of the composite film, was the main factor of the coated sample's wear resistance.
Acknowledgment The authors gratefully acknowledge the financial support of the National Science Council of the Republic of China through grant nos. NSC 97-2221-E-019-008 and NSC 98-2221-E-019-005.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
[16] [17] [18] [19] [20] [21] [22]
B. Navinsek, P. Panjan, I. Milosev, Surf. Coat. Technol. 97 (1-3) (1997) 182. S.S. Kim, J.G. Han, S.Y. Lee, Thin Solid Films 334 (1–2) (1998) 133. J.A. Sue, T.P. Chang, Surf. Coat. Technol. 76-77 (1995) 61 Part 1. O. Knotek, F. Löffler, B. Bosserhoff, Surf. Coat. Technol. 62 (1–3) (1993) 630. J. Vetter, R. Knaup, H. Dwuletzki, E. Schneider, S. Vogler, Surf. Coat. Technol. 86–87 (1996) 739. S. Veprek, S. Reiprich, L. Shizhi, Surf. Coat. Technol. 86–87 (1996) 394. E. Martinez, R. Sanjinés, A. Karimi, J. Esteve, F. Lévy, Surf. Coat. Technol. 180–181 (2004) 570. J.H. Park, W.S. Chung, Y.-R. Cho, K.H. Kim, Surf. Coat. Technol. 188–189 (2004) 425. G. Kim, B. Kim, S. Lee, Surf. Coat. Technol. 200 (5–6) (2005) 1814. H.Y. Lee, W.S. Jung, J.G. Han, S.M. Seo, J.H. Kim, Y.H. Bae, Surf. Coat. Technol. 200 (1–4) (2005) 1026. S.Y. Lee, B.S. Kim, S.D. Kim, G.S. Kim, Y.S. Hong, Thin Solid Films 506–507 (2006) 192. C.S. Sandu, R. Sanjinés, M. Benkahoul, F. Medjani, F. Lévy, Surf. Coat. Technol. 201 (7) (2006) 4083. I.-W. Park, D.S. Kang, J.J. Moore, S.C. Kwon, J.J. Rha, K.H. Kim, Surf. Coat. Technol. 201 (9–11) (2007) 5223. J.-W. Lee, Y.-C. Chang, Surf. Coat. Technol. 202 (4–7) (2007) 831. H.-H. Lin, C.-C. Chou, J.-C. Huang, J.-W. Lee, Y.-C. Chang, Y.-B. Lin, The 36th International Conference on Metallurgical Coatings and Thin Films, Secession F2/ B7-7, American Vacuum Society, San Diego, USA, 2009, April 25th–May 1st, 2009. W.C. Oliver, G.M. Pharr, J. Mater. Res. 7 (6) (1992) 1564. S. Ma, J. Procházka, P. Karvánková, Q. Ma, X. Niu, X. Wang, D. Ma, K. Xu, S. Veprek, Surf. Coat. Technol. 194 (1) (2005) 143. P.J. Kelly, R.D. Arnell, Vacuum 56 (3) (2000) 159. R.D. Arnell, P.J. Kelly, J.W. Bradley, Surf. Coat. Technol. 188–189 (2004) 158. S. Veprek, S. Reiprich, Thin Solid Films 268 (1–2) (1995) 64. M. Benkahoul, P. Robin, L. Martinu, J.E. Klemberg-Sapieha, Surf. Coat. Technol. 203 (8) (2009) 934. J. Musil, M. Jirout, Surf. Coat. Technol. 201 (9–11) (2007) 5148.
Contents lists available at ScienceDirect
Thin Solid Films j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / t s f
Tribological properties of Cr–Si–N nanocomposite film adherent silicon under various environments Hsin-Han Lin a, Chau-Chang Chou b,⁎, Jyh-Wei Lee c a b c
Institute of Nanotechnology and Microsystems Engineering, National Cheng Kung University, Tainan 70101, Taiwan, ROC Department of Mechanical and Mechatronic Engineering, National Taiwan Ocean University, Keelung 20224, Taiwan, ROC Department of Materials Engineering, Mingchi University of Technology, Taipei 24301, Taiwan, ROC
a r t i c l e
i n f o
Available online 10 May 2010 Keywords: Nanocomposite thin films Pulsed DC magnetron sputtering process Cr–Si–N films Tribological property
a b s t r a c t Chromium nitride thin films have good corrosion resistance and mechanical properties. However, their hardness is slightly lower than that of other hard coatings. The concept of nanocomposite thin films is employed by adding silicon to form Cr–Si–N thin films with enhanced hardness and wear resistance. In this study, Cr–Si–N films with various Si contents were coated on silicon wafer to enhance the tribological properties and anticorrosion by a bipolar symmetry pulsed DC reactive magnetron sputtering process. The tribological properties were studied by a pin-on-disk tester. The tests were conducted with the same operating condition under three different environments. They were performed in the ambient atmosphere (in 55% humid air), DI water, and 0.01 M NaCl aqueous solution, respectively. The wear tests revealed that, as the silicon content was increased, even though the Cr–Si–N films had a better anticorrosion property they had an inferior performance on wear resistance. The results were concluded to be mainly due to Cr–Si–N films’ microstructures and adhesion to the Si substrate rather than their hardness and toughness. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Chromium nitride (Cr–N) thin film has been widely applied in molding industries to prolong service life due to its excellent corrosion resistance and mechanical properties [1]. It also exhibits rather good adhesions when deposited on general steel substrates [2]. The Cr–N coating has also been noted for its good oxidation resistance up to 800 °C [3]. At higher temperatures, the formation of Cr2O3 on the surface can further promote the coating's wear resistance and its capability as a diffusion barrier. This is beneficial to the improvement of die casting molds [4,5]. According to the concept of superhard nanocomposite hard coatings proposed by Veprek et al. [6], the addition of silicon to the Cr–N thin film to form amorphous silicon nitride phase has been adopted to improve the mechanical properties and oxidation resistance [7–14]. It has been observed that CrN crystalline and amorphous Si3N4 phase are two major contributors in the Cr–Si–N nanocomposite thin film [8,14]. Lee and Chang [14] evaluated the mechanical properties of the Cr– Si–N films by microhardness as well as micro wear tests of atomic force microscopy (AFM). The antiwear capability and microhardness of the composite films were increased with Si contents of up to 12 at. %. The corrosion resistance of each thin film was evaluated by a potentiostat in 3.5 wt.% NaCl aqueous solution in our previous
⁎ Corresponding author. E-mail address: [email protected] (C.-C. Chou). 0040-6090/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2010.05.035
report [15], showing that it was enhanced directly by the amount of Si content of the composite film. Microwear tests in a corrosive environment (a NaCl solution) by AFM were also implemented to investigate the possible wear-corrosion synergism. It has been shown that the best wear resistance under the corrosive environment is obtained with Cr–Si–N films with a concentration of Si of 10.1 at.%, which also yields the maximum hardness of Cr–Si–N coatings. In this work, the same asymmetric bipolar pulsed DC reactive magnetron sputtering technique [15] was adopted to deposit Cr–Si–N thin films with different silicon contents on ptype (100) silicon substrates. The tribological properties of the coated samples rather than the composite thin films themselves were studied by a pin-on-disk tester. The tests were conducted with the same operating condition under three different environments to investigate the wear resistance and tribological behavior of Cr–Si–N coatings on Si substrates. 2. Experimental details 2.1. Deposition of nanocomposite films Cr–Si–N thin films were coated on a p-type (100) silicon wafer by using a bipolar asymmetry pulsed DC reactive magnetron sputtering system. Two DC power suppliers with a pulse controller (SPIK 2000A, Shen Chang Electric Co., Taiwan) were used to supply controlled power to Cr and Si targets, respectively. To obtain Cr–Si– N films with various Si contents, the powers supplied to Si target
7510
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514
Table 1 Elemental compositions of Cr–Si–N nanocomposite thin films produced by various pulse powers supplied to Si target. Code of silicon specimen
Cr target power (W)
Si target power (W)
Element content (at.%) Cr
N
Si
A1 A2 A3 A4
300 300 300 300
224 160 95 65
53.1 52.1 55.5 58.0
34.1 38.9 38.4 38.1
12.8 9.0 6.1 3.9
were kept at four values: 224, 160, 95, and 65 W, while, that to Cr target was fixed at 300 W. The purity and diameter of Cr and Si targets were both 99.99 wt.% and 76.2 mm. The pulse frequency was kept at 20 kHz with 100% reversed voltage and an 80% duty cycle during the sputtering process. A substrate bias of − 300 V was applied with a pulse unit (Sparcle V, Advanced Energies Industries, USA) with 80 kHz frequency, 15% reversing voltage and a 60% duty cycle. The base pressure, 5.6 × 10− 6 Pa, was achieved before sputtering. All substrates were sputtered clean for 10 min at 2.7 Pa argon pressure with a substrate bias of −500 V. A Cr–Si mixed film was deposited as an adhesion interlayer with around 150 nm thickness onto the substrate under 5 × 10− 3 Pa in pure Ar atmosphere. The flow rates of Ar/N2 mixture at a ratio of 1:1 were monitored by individual mass flow controllers. The pressure in the chamber during deposition was 8.0 × 10− 1 Pa. The deposition time was 165 min. All substrates were heated to 300 °C during the sputtering process. They were placed 100 mm from both targets and fixed to a holder which rotated continuously in order to obtain uniform nanocomposite thin films.
2.2. Film characterization The elemental compositions of the Cr–Si–N thin films were identified with a field emission electron probe microanalyzer (FEEPMA, JXA-8500F, JEOL, Japan) with a ZAF-corrected program. The phases of the coatings were explored with a grazing incidence (1°) X-ray diffraction (XRD) equipped with a Cu target source. The surface roughness analysis was performed by an AFM system (D3100, Veeco). A silicon-based probe (NSC15/AIBS, Micko Mash) was used to scan a 50 × 50 μm2 area of the composite thin films to derive their roughness average (Ra) and maximum peak-to-valley height (Rmax).
Fig. 2. Cross-sectional SEM images of Cr–Si–N coatings with (a) 3.9 at.% Si and (b) 12.8 at.% Si.
The mechanical properties, the hardness and reduced Young's modulus, were measured by micro indentation using a Triboindenter (Hysitron Inc.) equipped with a Berkovich pyramidal tip. The load– displacement curves were analyzed by applying the method of Oliver and Pharr [16]. The maximum load was 2.5 mN, applied at a loading– unloading rate of 1000 μN/s. Each specimen was tested more than three times. 2.3. Wear tests of Cr–Si–N samples The wear coefficients were evaluated by means of pin-on-disk tests in three environments. They were conducted in the ambient atmosphere (in 55% humid air), DI water, and 0.01 M NaCl aqueous solution at room temperature. The counter parts were cemented tungsten carbide (WC with 6.0 wt.% Co) balls of 1/4″ (6.350 mm)
Table 2 Hardness (H), reduced Young's modules (Er), and resistance to plastic deformation (H3/ E2r ) of Cr–Si–N samples with different Si contents. Code of silicon specimen
Fig. 1. XRD patterns of Cr–Si–N thin films deposited on Si substrates. (Note: A1: 12.8 at. % Si, A2: 9.0 at.% Si, A3: 6.1 at.% Si, and A4: 3.9 at.% Si).
A1 A2 A3 A4
Hardness (H), GPa
Reduced Young's modulus (Er), GPa
15.31 ± 0.26 11.54 ± 0.35 13.85 ± 1.48 8.89 ± 0.43
152.6 ± 2.1 147.6 ± 3.5 155.2 ± 7.4 131.4 ± 3.4
Resistance to plastic deformation (H3/Er2), GPa 0.15 0.07 0.11 0.04
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514 Table 3 Wear volume (Ws) and dimensional wear coefficient (WR) of Cr–Si–N samples with 12.8 at.% and 9.0 at.% Si contents. r = 1.4 mm, Fn = 0.49 N, S = 22 m Si content of Cr–Si–N film Ws (mm3)
WR (mm3 N−1 m−1)
Dry DI water 0.01 M NaCl Dry DI water 0.01 M NaCl
WS, and the dimensional wear coefficient, WR, were calculated using the equations [17] Ws =
A1 (12.8 at.% Si)
A2 (9.0 at.% Si)
6.14 × 10− 3 5.33 × 10− 3 6.63 × 10− 3 5.70 × 10− 4 4.94 × 10− 4 6.15 × 10− 4
3.49 × 10− 3 2.82 × 10− 3 3.60 × 10− 3 3.24 × 10− 4 2.62 × 10− 4 3.34 × 10− 4
7511
t 2 2 3t + 4b 2πr 6b
ð1Þ
Ws Fn ⋅S
ð2Þ
and WR =
where r is the radius of the wear track, S is the sliding distance, and Fn is the normal load. diameter. A normal load of 0.49 N was applied. The sliding speed was 0.015 m/s (100 rpm) along a circular wear track of 2.8 mm diameter. The total wear length was 22 m (2500 cycles) for each test. Worn samples were ultrasonically cleaned in alcohol for 2 min to remove wear debris after the wear tests. The surface morphologies and chemical composition of the wear scar were investigated by scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) analyzer. The depth of each wear track, t, and the width of the wear track, b, were determined by using a stylus surface profiler (SE500-58D, Kosaka lab). The wear volume,
3. Results and discussion 3.1. Composition and phases characteristics The elemental compositions of the various Cr–Si–N coatings are listed in Table 1. The Si concentration of these films increases as the sputtering power on the Si target is raised. Fig. 1 demonstrates the XRD patterns of Cr–Si–N thin films on the p-type (100) silicon wafer. For coatings with Si content of higher than 6.1 at.%, only pure CrN (200) phase is shown. On the other hand, a mixture of CrN (200),
Fig. 3. Wear tracks’ surface morphologies and EDX results of Cr–Si–N samples with (a) 9.0 at.% Si and (b) 6.1 at. % Si after pin-on-disk tests in the ambient atmosphere.
7512
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514
(220) and Cr2N (002) phases is found in coatings with Si content of less than 3.9 at.%. The phase transition range of Si concentrations, 3.9– 6.1 at.%, is higher than the values of 3.1–4.1 at.% Si, derived from the authors’ previous work (Lee and Chang [14]). The different Si contents of their Cr–Si–N thin films were obtained due to the difference of the stationary substrate's position between the Cr and Si targets. In this study, the substrates were rotating and their Si contents were changed by varying the sputtering power on the Si targets. In addition, as the substrate bias was supplied with a higher pulse frequency (80 kHz) than that (50 kHz) used in [14], the light nitride atom was much more easily removed by the higher excitation energy [18,19]. The composition of Cr–Ns in coatings with low Si content thus tended to be in the form of Cr2N instead of CrN. A similar result has been reported by Lee et al. [10]. They explain that Si plays the role as a substitutional solid solution element in the Cr–Si–N film and the extra Cr ions will react with the Cr–N phase to form the Cr2N. As the Si content is higher than the solubility limit, amorphous SiNx phase will thus produce and the Cr2N phase will disappear [10].
but those of other films with lower Si content was particulate. Among the samples, the Cr–Si–N thin film with the lowest Si content (3.9 at.%) had the highest surface roughness (2.188 nm Ra and 102.73 nm Rmax). The cross-sectional SEM images of coatings in Fig. 2 show that the Cr–Si–N thin films’ thickness was determined to be in the range of 1.2– 1.5 μm. A very dense and compact columnar structure is found on the coating with 3.9 at.% Si as shown in Fig. 2a. When the Si content increased, a featureless morphology was observed (Fig. 2b) for the film containing 12.8 at.% Si. No columnar structure could be identified by SEM anymore. The cross-sectional transmission electron microscopy (TEM) images of coatings from the previous work of Lee and Chang [14] revealed that the addition of Si reduced the width of columnar structure and further formed CrN nanocrystallites surrounded by amorphous SiNx matrix in Cr–Si–N coatings. In Fig. 1, the rather broad and low intensity reflections of CrN (200) are observed for coatings A1 containing 12.8 at. % Si, which also verifies the very small crystallite size of coatings. In the TiN/a-Si3N4 coating system [20], the crystallite size is also seen for the TiN/a-Si3N4 coating with more than 8 at.% Si.
3.2. Morphology and microstructure analysis
3.3. Evaluation of films’ toughness properties
From the surface roughness determination by AFM, Cr–Si–N thin film with 12.8 at.% Si was very smooth (1.052 nm Ra and 13.599 nm Rmax),
The values of hardness of the Cr–Si–N film, H, and reduced Young's modulus, Er, are listed in Table 2. They were measured at a depth of
Fig. 4. Wear tracks’ surface morphologies and EDX results of Cr–Si–N samples with (a) 6.1 at.% Si and (b) 3.9 at. % Si after pin-on-disk tests in the DI water.
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514
50–100 nm from the coating surface, where it was believed that no influence of substrate was included. The values of both properties were in the order of composite films with 12.8 at.%, 6.1 at.%, 9.0 at.%, and 3.9 at.% of Si in this study. The resistance to plastic deformation (H3/E2r ), or toughness of Cr–Si–N composite films, which was derived from H and Er [21,22] followed the same order as listed above. The main difference between Lee and Chang [14] and this work is the trend of the Cr–Si–N films’ hardness with different Si contents. When the Si content is higher than 1.6 at.% in [14], the hardness increases with increasing Si concentration and a maximum hardness is achieved for coatings with around 10.1 at.% Si and a minimum grain size around 5.2 nm, caused by the grain boundary hardening, i.e., the Hall–Petch relation [7,20]. However, in this work the phase transition range of Si concentrations is higher and the grain boundary hardening is found only in the range of around 6.1–9.0 at.% Si. Different coating setups, e.g. the variable impulse powers and rotating holder, could account for the different microstructures of the films with the same Si content. 3.4. Tribological results The wear volume, WS, and dimensional wear coefficient, WR, of Cr– Si–N samples under three environments are shown in Table 3. Due to the resolution of the adapted stylus profiler, only coatings with
7513
12.8 at.% and 9.0 at.% Si contents were measured and evaluated. It can be seen that the coated samples with higher hardness had higher wear coefficients. This is contrary to the micro tribological behavior of the composite films [14]. The values of WR indicate that the wear coefficient of the coated specimens can be reduced from the dry contact to a lubrication state by the de-ionized (DI) water. However, when the corrosive 0.01 M NaCl aqueous solution was implemented, the wear coefficients increased and were even higher than the ones tested in the atmosphere. To investigate the wear mechanism involved in the pin-on-disk test, SEM and EDX analysis of the wear track region were performed. SEM observation of the worn surface of each coating under three test environments reveals that, when tested in the atmosphere, the failure of the coatings occurred for Cr–Si–N films with Si contents higher than 6.1 at.% (Fig. 3). While the samples were tested in DI water and 0.01 M NaCl aqueous solution, the films with Si contents higher than 3.9 at.% failed (Figs. 4 and 5). The wear resistance of the composite films on Si substrates was thus found to be the highest when the Si content was 3.9 at.% and declined as the Si content increased. Among these tests, the amorphous SiNx phase of the Cr–Si–N films with higher Si contents is believed beneficial to reduce the friction on the surface when conducted in the ambient atmosphere. This also has been reported from the discussions of friction coefficient in [8,14]. When the wear tests were carried out in
Fig. 5. Wear tracks’ surface morphologies and EDX results of Cr–Si–N samples with (a) 6.1 at.% Si and (b) 3.9 at. % Si after pin-on-disk tests in the 0.01 M NaCl aqueous solution.
7514
H.-H. Lin et al. / Thin Solid Films 518 (2010) 7509–7514
DI water or NaCl aqueous solution, the friction on the wear tracks was reduced mainly by the lubrication of water. The influence of the shear stress beneath the contact surface could thus extend to the interface of the coating and the substrate, which resulted in the failure of Cr–Si–N coatings with Si contents only higher than 3.9 at.%. This is inconsistent with the results of micro wear test in [14], which concluded that a composite film with higher hardness would reveal a lower wear coefficient. It should be notified that the wear tracks in [14] did not penetrate through the coatings and only the antiwear capability of the composite film itself was considered. Nevertheless, the results obtained from pin-on-disk tests in this work are the same as those of [21], which reported that the hardness and toughness of the Cr–Si– N films dominate the coatings’ antiwear performance. However, in this work, the higher hardness and toughness of the composite films did not guarantee better wear resistance of the coatings. As depicted in Figs. 3a, 4a, and 5a, the wear track from various environments show that the cracks and grooves induced by abrasive wear were more significant under a higher wear coefficient, e.g., coatings with 12.8 at.% Si (Fig. 3a). EDX results show that the wear debris still attached inside the wear track was all heavily oxidized. According to the analysis of [21], this can be attributed to Cr2O3. The worn surfaces, as shown in Figs. 3b, 4b, and 5b, illustrate that the element compositions inside and outside the wear tracks remained nearly unchanged and free of oxide as long as the films were sustained on the substrate. Compared to the large wear coefficient of the failed coatings, it is summarized that the hard wear debris accelerates the wear coefficient when the films are removed from the wear tracks. 4. Conclusion Cr–Si–N composite thin films with Si contents ranging from 3.9 to 12.8 at.% were coated on a p-type (100) silicon wafer by a bipolar symmetry pulsed DC reactive magnetron sputtering process. The tribological properties of the coated samples were studied by pin-on-disk tests conducted under the same operating condition in three different environments. For Cr–Si–N composite films with the same Si content, the wear coefficient obtained in DI water was the lowest. Wear tests in 0.01 M NaCl aqueous solution show more severe wear than did those in the atmosphere. All wear coefficients increased as the Si content of the coating increased. The amorphous SiNx phase of Cr–Si–N composite films was beneficial to reduce the friction while
rubbing in ambient atmosphere. However, in DI water and NaCl aqueous solution, the lubrication of water was suggested to decrease the friction and extend the influence of the shear stress beneath the contact surfaces. The adhesion between the films and substrate, which could be enhanced by the columnar microstructure of the composite film, was the main factor of the coated sample's wear resistance.
Acknowledgment The authors gratefully acknowledge the financial support of the National Science Council of the Republic of China through grant nos. NSC 97-2221-E-019-008 and NSC 98-2221-E-019-005.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
[16] [17] [18] [19] [20] [21] [22]
B. Navinsek, P. Panjan, I. Milosev, Surf. Coat. Technol. 97 (1-3) (1997) 182. S.S. Kim, J.G. Han, S.Y. Lee, Thin Solid Films 334 (1–2) (1998) 133. J.A. Sue, T.P. Chang, Surf. Coat. Technol. 76-77 (1995) 61 Part 1. O. Knotek, F. Löffler, B. Bosserhoff, Surf. Coat. Technol. 62 (1–3) (1993) 630. J. Vetter, R. Knaup, H. Dwuletzki, E. Schneider, S. Vogler, Surf. Coat. Technol. 86–87 (1996) 739. S. Veprek, S. Reiprich, L. Shizhi, Surf. Coat. Technol. 86–87 (1996) 394. E. Martinez, R. Sanjinés, A. Karimi, J. Esteve, F. Lévy, Surf. Coat. Technol. 180–181 (2004) 570. J.H. Park, W.S. Chung, Y.-R. Cho, K.H. Kim, Surf. Coat. Technol. 188–189 (2004) 425. G. Kim, B. Kim, S. Lee, Surf. Coat. Technol. 200 (5–6) (2005) 1814. H.Y. Lee, W.S. Jung, J.G. Han, S.M. Seo, J.H. Kim, Y.H. Bae, Surf. Coat. Technol. 200 (1–4) (2005) 1026. S.Y. Lee, B.S. Kim, S.D. Kim, G.S. Kim, Y.S. Hong, Thin Solid Films 506–507 (2006) 192. C.S. Sandu, R. Sanjinés, M. Benkahoul, F. Medjani, F. Lévy, Surf. Coat. Technol. 201 (7) (2006) 4083. I.-W. Park, D.S. Kang, J.J. Moore, S.C. Kwon, J.J. Rha, K.H. Kim, Surf. Coat. Technol. 201 (9–11) (2007) 5223. J.-W. Lee, Y.-C. Chang, Surf. Coat. Technol. 202 (4–7) (2007) 831. H.-H. Lin, C.-C. Chou, J.-C. Huang, J.-W. Lee, Y.-C. Chang, Y.-B. Lin, The 36th International Conference on Metallurgical Coatings and Thin Films, Secession F2/ B7-7, American Vacuum Society, San Diego, USA, 2009, April 25th–May 1st, 2009. W.C. Oliver, G.M. Pharr, J. Mater. Res. 7 (6) (1992) 1564. S. Ma, J. Procházka, P. Karvánková, Q. Ma, X. Niu, X. Wang, D. Ma, K. Xu, S. Veprek, Surf. Coat. Technol. 194 (1) (2005) 143. P.J. Kelly, R.D. Arnell, Vacuum 56 (3) (2000) 159. R.D. Arnell, P.J. Kelly, J.W. Bradley, Surf. Coat. Technol. 188–189 (2004) 158. S. Veprek, S. Reiprich, Thin Solid Films 268 (1–2) (1995) 64. M. Benkahoul, P. Robin, L. Martinu, J.E. Klemberg-Sapieha, Surf. Coat. Technol. 203 (8) (2009) 934. J. Musil, M. Jirout, Surf. Coat. Technol. 201 (9–11) (2007) 5148.