TiN multilayer coatings deposited by pulsed dc magnetron sputtering

TiN multilayer coatings deposited by pulsed dc magnetron sputtering

Surface & Coatings Technology 276 (2015) 152–159 Contents lists available at ScienceDirect Surface & Coatings Technology journal homepage: www.elsev...

2MB Sizes 8 Downloads 107 Views

Surface & Coatings Technology 276 (2015) 152–159

Contents lists available at ScienceDirect

Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat

Mechanical and tribological properties of CrN/TiN multilayer coatings deposited by pulsed dc magnetron sputtering Y.X. Ou a,b, J. Lin b,c, H.L. Che a, W.D. Sproul b,d, J.J. Moore b, M.K. Lei a,⁎ a

Surface Engineering Laboratory, School of Materials Science and Engineering, Dalian University of Technology, Dalian, 116024, China Advanced Coatings and Surface Engineering Laboratory (ACSEL), Department of Metallurgical and Materials Engineering, Colorado School of Mines, Golden, CO 80401, USA Southwest Research Institute, San Antonio, TX 78238, USA d Reactive Sputtering, Inc, 2152 Goya Place, San Marcos, CA 92078, USA b c

a r t i c l e

i n f o

Article history: Received 30 March 2015 Revised 4 June 2015 Accepted in revised form 29 June 2015 Available online xxxx Keywords: CrN/TiN multilayer coatings Pulsed dc magnetron sputtering (PDCMS) Mechanical properties Tribological properties

a b s t r a c t CrN/TiN multilayer coatings were deposited by pulsed dc magnetron sputtering (PDCMS) in a closed field unbalanced magnetron sputtering system. The Ti target power was maintained at 2000 W while the Cr target power was varied from 70 to 1000 W. As the Cr target power was increased, CrN/TiN multilayer coatings with the Cr/(Cr + Ti) ratio of 0.037–0.573 had a single phase face-centered cubic structure with a texture evolution from (111) to (220). The residual stress of the coatings increased from −1.03 GPa to −5.65 GPa. The hardness and the H/E* and H3/E*2 ratios of the coatings exhibited an initial increase with the increasing Cr target power, then followed by a decrease. The coating with a Cr/(Cr + Ti) ratio of 0.30 reached the highest hardness and the H/E* and H3/E*2 ratios of 31 GPa, 0.0832 and 0.214, respectively. The coating also showed a dominant oxidative wear with the lowest friction coefficient and specific wear rate of 0.41 and 2.3 × 10−6 mm3 N−1 m−1, respectively. The increase in the H/E* and H3/E*2 ratios led to the increase in the toughness and cohesion/adhesion strength of the coatings with the increased HF levels and critical loads (LC1, LC2 and LC3). However, the coatings with similar H/E* and H3/E*2 ratios exhibited different tribological properties due to the different critical loads LC3 resulted from the increased compressive residual stress. The improvements in toughness and cohesion/adhesion strength promoted the reduction in the crack initiation and propagation, and oxidative wear during dry sliding tests. © 2015 Elsevier B.V. All rights reserved.

1. Introduction CrN/TiN multilayer coatings were widely used as an excellent protective candidate to improve the performance and lifetime of components in complex environments due to their increased mechanical properties and wear resistance in comparison with traditional monolithic coatings (e.g., TiN, CrN, TiCrN) [1–3]. The performance of CrN/ TiN multilayer coatings was found to be associated with the modulated period (Λ), grain size, and interface structure [4–7]. CrN/TiN multilayer coatings have been investigated in terms of the effects of the Cr/ (Cr + Ti) ratio and Λ on the mechanical and tribological properties [1, 6,8,9]. The explanations of hardening mechanisms were based on the modified Hall–Petch effect [10], variations of shear modulus [11], strain in the lattice-mismatched materials [12] and internal stress [13,14]. The tribological responses of hard coatings were usually complex [15,16]. Although hard coatings exhibited a lower plastic deformation and a higher resistance to scratching during wear tests, the high toughness and low friction were considered as important characteristics in tribological applications [14,17]. Generally, the coated surface was ⁎ Corresponding author. E-mail address: [email protected] (M.K. Lei).

http://dx.doi.org/10.1016/j.surfcoat.2015.06.064 0257-8972/© 2015 Elsevier B.V. All rights reserved.

subject to a plastic deformation and a scratch under sliding contact load in ball-on-disk wear tests. When the coated surface was unable to tolerate the deformation, the cracks would be rapidly initiated and propagated [16,18]. However, a sharp decrease in the coatings' toughness due to the crack initiation was unnecessary to result in the failure of the coatings [19,20]. The wear particles generated by the surface plastic deformation can possibly act as an abradant to accelerate the wear and to influence the wear mechanisms [16]. The resistance of the coating against cracking was benefited from the stratified structure, which hindered the crack propagation at interfaces between layers in the multilayered coatings [21,22]. In general, the compressive stresses in the coatings were favorable since they can act to close through thickness cracks and improve the density of the microstructure. However, at high stress levels the reduction in adhesion due to the build-up of stored elastic strain energy would give rise to the decrease in the wear resistance of the coatings [23]. The occurrence of the elastoplastic deformation during sliding led to wear [24]. Recently, the scratching technique has increasingly gained interest due to the numerous inherent properties implied (adherence, hardness, elasticity, visco-elasticity, cohesion, indenter radius, etc.) during tests [25–28]. The critical loads (LC1, LC2, LC3, LC4) of the coatings were evaluated by the scratch adhesion test [29,30]. It was found that

Y.X. Ou et al. / Surface & Coatings Technology 276 (2015) 152–159

153

Table 1 Summary of the composition, modulated periods and properties of CrN/TiN multilayer coatings.

(Cr + Ti)/N

1.67 3.11 6.56 13.43 22.49 24.98 27.95

43.53 41.02 38.32 35.82 23.76 21.41 19.83

54.77 55.87 55.12 54.81 53.75 53.61 52.22

0.0369 0.0705 0.1544 0.2972 0.4863 0.5385 0.5730

0.8258 0.7899 0.7787 0.8245 0.8605 0.8653 0.9150

cohesion strength of the coatings was related to the crack initiation resistance of LC1, while the adhesion strength was associated with the first edge chipping (LC2) and the first delamination within the track (LC3) [20,32] before the massive delamination within the track (LC4) [30]. The H/E* (elastic strain to failure resistance) and H3/E*2 ratios (the plastic deformation resistance) of the coated surface associated with the toughness were found to have a strong relationship with the wear resistance of the coatings [15,16,18–20]. The H3/E*2 ratio was considered as the surface plastic deformation response to sliding contact load, which was associated with the hardness of the coatings [1, 16,31]. The increase in the H/E* and H3/E*2 ratios led to the increase in critical load LC, which prevented crack initiation and propagation, and therefore improving the wear resistance of hard coatings [16,20,31,32]. The H/E* and H3/E*2 ratios and the critical load LC were also related to residual stress in hard coatings [16,31,32]. However, the influence of the H/E* and H3/E*2 ratios, critical load LC and residual stresses on tribological properties and the endurance of hard coatings was still limited and needed further investigations. In this work, CrN/TiN multilayer coatings were deposited at various Cr target powers by pulsed dc magnetron sputtering (PDCMS) in a closed field unbalanced magnetron sputtering configuration. The mechanical and tribological properties of CrN/TiN multilayer coatings were investigated and discussed.

2. Experimental details CrN/TiN multilayer coatings were deposited on AISI 304L austenite stainless steel (SS), Si (100) and cemented carbide substrates in a closed field unbalanced magnetron sputtering system by sputtering a Cr and a Ti target. The average surface roughness of AISI 304L SS and cemented carbide substrates was 8–10 nm. The targets (292 mm × 102 mm × 6.4 mm) with a purity of 99.95% were installed opposite each other at a 470 mm distance (the diameter of the chamber). Prior to the coating depositions, the substrates were ultrasonically degreased in denatured acetone and alcohol for 15 min, respectively. The substrates were mounted in the chamber with a substrate to target distance of 120 mm. The chamber was pumped down to a base pressure less than 4.0 × 10−4 Pa. In order to remove the surface contaminates, the substrates were cleaned by Ar+ plasma etching for 30 min with a substrate bias of −500 V pulsed at 100 kHz and 90% duty cycle. A Cr adhesion layer with a thickness of about 200 nm was firstly deposited with a −60 V dc substrate bias prior to all coating depositions. CrN/TiN multilayer coatings with a total thickness of 2.0–2.5 μm were synthesized by alternately depositing CrN and TiN nanolayers in an Ar/N2 atmosphere. The substrate holder was moving back and forth between the Cr and Ti targets using a rotation system with the staying periods of 2 s and 7 s in front of the Cr and Ti targets, respectively. The transition of the substrate holder between the Cr and Ti targets took 6 s. The Cr and Ti targets were powered by PDCMS using Advance Energy Pinnacle Plus power sources. The Cr target was powered at

Residual stress [GPa]

COF

Wear rate [10−6 mm3 N−1 m−1]

28.4 29.2 29.5 31.0 29.1 28.4 26.5

326 339 340 353 336 323 309

0.0823 0.0816 0.0819 0.0832 0.0820 0.0819 0.0834

0.184 0.194 0.198 0.214 0.195 0.196 0.185

−1.03 – −1.95 −3.12 – – −5.65

0.88 0.76 0.69 0.41 0.55 0.59 0.65

11.0 9.1 8.7 2.3 4.1 6.3 7.5

70–1000 W and the Ti target was powered at 2000 W, respectively. The target pulsing parameters were 100 kHz and 90% duty cycle. During the depositions, the working pressure was maintained at 0.67 Pa with 40% (23 sccm) N2 flow rate to total gas flow rate. A −60 V negative dc substrate bias (Vs) voltage (AXIS, Zpulser LLC) was used for the depositions of CrN/TiN multilayer coatings. The element concentration in CrN/TiN multilayer coatings was detected using an EPMA-1600 type electron probe microanalyzer (EPMA). The crystallographic structure was investigated using a highangle X-ray diffraction (HAXRD) in a PHILIPS X-pet diffractometer from 30° to 90° in the θ-2θ configuration with the Cu Kα radiation. The FWHM of HAXRD pattern was calculated using X'Pert High Score Plus software. The instrumental broadening was excluded and the strain broadening was neglected when simply using the Debye– Scherrer equation to estimate mean crystalline size [33,34]. The Λ of 4.4–8.2 nm was calculated from the deposition time of 110–70 min and the thickness of CrN/TiN multilayer coatings, as shown in Table 1. The residual stresses of CrN/TiN multilayer coatings were estimated by sin2ψ method using a D8-Discover grazing incident X-ray diffraction (GIXRD) (Cu Kα radiation). The principle and details of the GIXRD sin2ψ method can be found in Refs. [35,36]. The microstructure and thickness of the coatings were examined on the fractural cross-section of a coated silicon wafer using a JEOL JSM-7000F field emission scanning electron microscope (FESEM). The cross-sectional microstructure and Λ of the coatings were further investigated using a Philips/FEI CM200 transmission electron microscope (TEM) operated at 200 kV. The hardness (H) and elastic modulus (E) of the coatings were measured in a nanoindenter (Nanoindenter XPTM, MTS Systems Corporation) equipped with a Berkovich diamond indenter. The indentation depth

SS C-(311)

Cr/(Cr + Ti)

H3/E*2

C-(220)

N [at.%]

H/E*

SS

Ti [at.%]

E [GPa]

SS Cr

4.4 4.5 4.8 5.4 6.2 7.0 8.2

Cr [at.%]

H [GPa]

C-(111)

70 100 200 400 600 800 1000

Composition

C-(200)

Λ [nm]

Intensity (a.u.)

Cr target power [W]

70 W 100 W 200 W 400 W 600 W 800 W 1000 W

30

40

50

60

70

80

90

2θ θ (degree) Fig. 1. HAXRD patterns of CrN/TiN multilayer coatings deposited at a Cr target power of 70–1000 W.

154

Y.X. Ou et al. / Surface & Coatings Technology 276 (2015) 152–159

was kept constantly below 10% of the coating thickness to minimize the substrate effect. For each sample, sixteen effective measurements separated by a distance of 200 μm were made to obtain the statistical results. A constant Poisson's ratio of ν = 0.235 was assumed for the coatings as an average of ν = 0.22 for c-CrN [37] and ν = 0.25 for c-TiN [38]. The adhesion strength of the coatings deposited on cemented carbide substrates was evaluated by a Micro Scratch-tester (CSM Instrument) using a 200 μm radius Rockwell C indenter. The coatings were scratched with increasing the load to a maximum of 30 N at a speed of 50 N/min. The scratch length was 3 mm. The acoustic emission (AE) signal was measured simultaneously and recorded for each test. The critical loads were determined based on specific damages observed by optical

microscopy and AE changes. The adhesion strength of the coatings deposited on AISI 304L SS substrates was evaluated by Rockwell C indentation test (HRC) using a 200 μm radius HRC indenter. The load was 150 kg. The indented coatings were examined by FESEM to evaluate the adhesion level according to VDI guidelines 3198, (1991) [39]. The dry sliding wear tests were performed on a ball-on-disk microtribometer (Center for Tribology, Inc.), where the surface of a stationary top mounted ball was rubbed against a reciprocating sample flat at room temperature and with a relative humidity of about 22%. The counterpart was a cemented carbide ball with a diameter of 1.6 mm. The applied normal load was 1 N and the sliding speed was 0.0471 m/s for a total sliding time of 1 h. After the wear tests, the

Fig. 2. Cross-sectional and top view SEM micrographs of CrN/TiN multilayer coatings deposited at different Cr target powers.

Y.X. Ou et al. / Surface & Coatings Technology 276 (2015) 152–159

wear tracks were examined using a Dektak surface profilometer to measure the wear rate of the coatings. The worn surface morphologies were analyzed using a FESEM.

3. Results and discussion 3.1. Composition and microstructure of CrN/TiN multilayer coatings The Cr, Ti and N element concentrations in CrN/TiN multilayer coatings deposited at various Cr target powers were summarized in Table 1. When the Cr target power was increased from 70 W to 1000 W, the Cr concentration increased from 1.7 at.% to 28.0 at.%, and the Ti concentration decreased from 43.5 at.% to 19.8 at.% due to a higher sputtering rate of Cr compared to that of Ti. However, the N concentration showed similar values in the range of 52.2–55.9 at.%. This was because the working pressure (0.67 Pa) and the nitrogen to total gas flow rate (40%) were maintained constantly for all the depositions. Besides, Ti can react much more readily with N2 than Cr due to a higher TiN formation standard heat of 80.8 kcal/mol than that of 29.8 kcal/mol for CrN [40]. The Cr/(Cr + Ti) ratio increased from 0.037 to 0.573. The (Cr + Ti)/N ratio of 0.778–0.915 (Me/N b 1) indicated stoichiometric CrN and TiN cubic phases were probably obtained in the coatings with such a high N content [41,42]. Fig. 1 shows HAXRD patterns of CrN/TiN multilayer coatings deposited at a Cr target power of 70–1000 W. The peaks from AISI 304L SS substrate (JCPDS 52-0513) and Cr peaks (JCPDS 89-4055) from the adhesion layer were detected and labeled as ‘SS’ and ‘Cr’ in the patterns, respectively. The typical peaks from cubic phase were correspondingly labeled as C-(111), C-(200), C-(220) and C-(311) in the patterns, respectively. All the coatings exhibited a single phase face-centered cubic (fcc) structure, while a texture evolution from (111) to (220) was developed with increasing Cr target power from 70 W to 1000 W. No distinct peaks from Cr2N (JCPDS 35-0803) phase were detected in the coatings [43]. At 70 W Cr target power, the coating exhibited a strong (111) preferred orientation, indicating CrN well-epitaxial growth on TiN [14,42] due to the lowest strain energy (U200 N U220 N U111) [44]. As the Cr target power was increased to 400 W, randomly orientated (111), (200) and (220) peaks were obtained due to the balance between the effects of thermodynamics (surface energy and strain energy) and kinetics (ion channeling) [45–47]. Upon further increasing to 1000 W, a distinct (220) broad peak was obtained for the coatings. It indicated that the increased CrN component was responsible for the development of (110) texture in CrN/TiN multilayer coatings [48] owing to the ion channeling effect [49]. The broad (220) peak was related to the decreased grain size and inhomogeneous stress. The estimated mean grain size was decreased from 25 nm to 6 nm with increasing Cr target power from 70 W to 1000 W. The stress states in CrN/TiN multilayer coatings were determined to be compressive stress ranging from − 1.03 GPa to − 5.65 GPa with increasing Cr target power from 70 W to 1000 W. The increase in Cr target power led to the increase in CrN layer thickness of CrN/TiN multilayer coatings. The increased number of interfaces would possibly result in decreasing stress since the interfaces were sites for the energy dissipation [50]. Similar results were reported earlier for CrN/AlN superlattice coatings deposited using PDCMS [51]. Fig. 2 shows the cross-sectional and top view SEM micrographs of CrN/TiN multilayer coatings deposited at different Cr target powers. At 70 W Cr target power, the polycrystalline CrN/TiN multilayer coatings exhibited a nano-columnar structure. Meanwhile, the dense packed nano-sized grains grew between the adjacent large columnar grains. When the Cr target power was increased to 400 W, the coating showed a dense columnar structure with smooth surface and a decreased grain size. With further increasing the Cr target power to 1000 W, the coating showed a decrease in columnar grain size. The refined grain size was related to the broadening of the (220) peak in HAXRD patterns. Also, the

155

residual compressive stress state in the coatings promoted the decrease in grain size according to the Hall–Petch relationship [52]. Fig. 3 shows the cross-sectional TEM (XTEM) micrograph, selected area electron diffraction (SAED) patterns and high resolution TEM micrograph of CrN/TiN multilayer coatings deposited at 400 W Cr target power. It can be seen that a stratified architecture consisting of successive distinct CrN and TiN nanolayers was obtained in the coating. A large amount of dense fine nano-columnar grains grew along with the large columnar grains. These features agreed well with the morphologies observed in SEM investigations as well as the estimated mean grain size from HAXRD patterns. The CrN nanolayers appeared to be darker than TiN nanolayers in the XTEM micrograph due to the higher scattering factor of Cr compared to that of Ti. The Λ of the coating measured from TEM micrographs was about 5.4 nm (TiN layer thickness of 4.3 nm, CrN layer thickness of 1.1 nm). The fcc crystal structure of the coating can be further identified because cubic (111), (200), (220) and (311) reflections were clearly detected in the SAED patterns. No hexagonal Cr2N phase was observed in the coatings. 3.2. Mechanical properties The hardness and Young's modulus of CrN/TiN multilayer coatings deposited at different Cr target powers of 70–1000 W were shown in Table 1. The hardness (H) and effective Young's modulus (E*) were determined by nanoindentation. The Young's modulus (E) can be calculated by the formula of E* = E / (1 − v2) (where v was the Poisson's ratio) [16,18]. The H and E of CrN/TiN multilayer coatings increased from 28.4 GPa and 326 GPa to 31 GPa and 351 GPa with an increase in the Cr target power from 70 W to 400 W and then decreased to 26.5 GPa and 309 GPa at 1000 W, respectively. The increased hardness of the coating can be correlated with the optimized Cr/(Cr + Ti) ratio, the increased number of interfaces and the dense microstructure [5,9,12,51]. The calculated H/E* and H3/E*2 ratios of CrN/TiN multilayer coatings deposited at various Cr target powers of 70–1000 W were summarized in Table 1. The H/E* ratio was an experimental measure of the elastic limit of strain [15]. The toughness of the material was its ability to absorb the energy during deformation up to its fracture [19]. The toughness of hard coatings can be improved in terms of a high hardness and

(b)

(a)

CrN/TiN(222) CrN/TiN(311) CrN/TiN(220) CrN/TiN(200) CrN/TiN(111)

(c)

CrN TiN Λ=5.4 nm

Fig. 3. (a) Cross-sectional TEM (XTEM) micrographs, (b) SAED paterns, (c) high resolution TEM micrograph of CrN/TiN multilayer coatings deposited at 400 W Cr target power.

Y.X. Ou et al. / Surface & Coatings Technology 276 (2015) 152–159

1.0 70 W

Coefficient of Friction

156

0.8

100 W 200 W 1000 W

0.6

800 W 600 W 400 W 304L SS

0.4

0.2

0.0

0

10

20

30

40

50

60

Time (min) Fig. 5. Coefficient of friction curves of CrN/TiN multilayer coatings deposited at different Cr target powers as compared with that of AISI 304L SS substrate.

a low elastic modulus. Both H/E* and H3/E*2 ratios were used to evaluate the toughness of hard coatings, which was considered as important parameters to evaluate tribological properties of hard coatings [15,16, 32,51]. As the Cr target power was increased from 70 to 1000 W, the H/E* ratio exhibited closed values of 0.0816–0.0834, while the H3/E*2 ratio increased from 0.184 to 0.214, and then decreased to 0.185. The coating deposited at 400 W Cr target power exhibited high toughness with high H/E* and H3/E*2 ratios of 0.0832 and 0.214, respectively. The improved toughness of CrN/TiN multilayer coatings was found to be associated with the crack deflection at interfaces and dissipation of crack energy by plastic deformation in the crack tip by introducing a large number of interfaces with different elastic moduli [19]. 3.3. Cohesion and adhesion The cohesion and adhesion of CrN/TiN multilayer coatings were evaluated by microscratch and Rockwell C tests. Fig. 4 shows acoustic emission curves and optical morphologies of scratch tracks for CrN/TiN multilayer coatings deposited on cemented carbide substrates. The SEM images of HRC indents performed on AISI 304L SS substrates were inserted in the figures. As the Cr target power was increased from 70 W to 1000 W, LC1 (LC1 = LC2) increased from 5.0 N to 10.8 N and LC3 varied from 7.2 N to 20.3 N. No LC4 was detected for the coatings in the tests. The critical loads (LC1, LC2 and LC3) reached the highest value for the coating deposited at 400 W Cr target power. The critical loads showed a similar tendency as those of the H/E* and H3/E*2 ratios. The coatings deposited at a Cr target power of 70 W and 1000 W showed similar LC1 and LC2 values due to the similar H/E* and H3/E*2 ratios, but the increase in the residual stress in the coatings led to the decrease in LC3 [31,32]. In progressive scratch tests, once the coated surface was unable to tolerate the deformation, cracks were rapidly initiated and propagated. The increased critical loads were benefited from the combined effect of the increased H/E* and H3/E*2 ratios and appropriate residual stress in the coatings [19,31,32]. As shown in SEM images of HRC indents on AISI 304L SS substrates, the coatings deposited at a Cr target power of 400 W and 600 W showed a better adhesion strength of HF1 than those of HF3 at 70 W and 1000 W in that no cracks and delamination of the coatings were observed. When the contact pressure levels were well-above the plastic yielding onset of

Fig. 4. Acoustic emission curves and optical images of scratch tracks for CrN/TiN multilayer coatings deposited at a Cr target power of 70–1000 W on cemented carbide substrates after the microscratch test along with SEM images of HRC indents on AISI 304L SS substrates inserted.

Y.X. Ou et al. / Surface & Coatings Technology 276 (2015) 152–159

the substrate when the applying contact load was increased, the tensile radial stresses and strains existing in the vicinity of the residual imprints became large enough for inducing circumferential cracks at the coating surface [53,54]. It was noted that the coatings deposited at various Cr target powers present the same HF level due to the similar LC1 (LC1 = LC2) of the crack initiation resistance, which was also related to the similar value of the H/E* and H3/E*2 ratios [17,30,31]. However, the LC3 values of the coatings were different because of the combined effect of the H/E* and H3/E*2 ratios and residual stress states. It was reported that the adhesion strength has a strong relationship with LC3 [31,32], which can be improved by appropriate residual compressive stress levels in the coatings [16,23].

157

3.4. Tribological properties Fig. 5 shows the coefficient of friction (COF) curves of CrN/TiN multilayer coatings deposited at a Cr target powers of 70–1000 W as compared with that of AISI 304L SS. The COF of AISI 304L SS showed an unstable and fluctuating feature indicating the occurrence of severe adhesion tendency between the slider and the coated surface. In the initial sliding period, the COF curves of CrN/TiN multilayer coatings exhibited a rapid increase at the beginning of the test (0–10 min), followed by a steady value of 0.41–0.88 (as shown in Table 1). The short running-in period with a low and unstable COF resulted from adventitious surface contamination such as impurities, oxide films, bulges, and pits. For the

Fig. 6. Typical SEM micrographs of worn surfaces of CrN/TiN multilayer coatings deposited at different Cr target powers as compared with that of AISI 304L SS substrate.

158

Y.X. Ou et al. / Surface & Coatings Technology 276 (2015) 152–159

coating deposited at 70 W Cr target power, the oscillation of COF revealed a slight adhesion tendency. When the Cr target power was increased to 400 W, the coating showed a decreased and stable COF of 0.41. An oscillation of COF was observed for the coating deposited at 1000 W Cr target power. As shown in Table 1, the calculated specific wear rate of the coatings varied from 2.3 × 10− 6 mm3 N− 1 m− 1 to 11.0 × 10− 6 mm3 N− 1 m− 1, which was lower than that of 16.0 × 10−6 mm3 N−1 m−1 for AISI 304L SS substrate. The coating deposited at 400 W Cr target power showed the lowest COF and specific wear rate of 0.41 and 2.3 × 10−6 mm3 N−1 m−1, respectively. Fig. 6 shows the typical SEM micrographs of the worn surfaces of CrN/TiN multilayer coatings deposited at a Cr target powers of 70–1000 W as compared with that of AISI 304L SS substrate. In progressive sliding wear test, the AISI 304L SS substrate suffered severe plastic deformation and surface material transfer. Plastic grooves, wear debris and small cracks were observed on the worn surface, resulting in an unstable and fluctuating tendency in COF curve. The wear mechanism of AISI 304L SS substrate was dominated by the severe adhesive wear. At 70 W Cr target power, non-uniform black oxide films covered the worn surface. As a result, the generated wear debris was accumulated on the wear track resulting in a slight adhesion feature consistent with the oscillation of COF. The small cracks were also observed in the wear track. The coatings were dominated by a severe oxidative wear. As the Cr target power was increased to 400 W, a smooth surface was observed due to the uniform oxide films formed on the worn surface, and the decreased amount of wear debris accumulated on the wear tracks. The uniform oxide films and some fine metal oxide debris covered the worn surface. This indicated that tribochemical oxidation occurred in the atmosphere environment, which resulted in a small decrease in the stable period of the COF during the wear testing. The wear mechanism of the coating was dominated by a mild oxidative wear. Further increasing to 1000 W, the coating suffered severe plastic deformation and surface material transfer with the formation of a large number of cracks and non-uniform black oxide films on the worn surface. The coating was subject to a severe oxidative wear. As the Cr target power was increased from 70 W to 400 W, the decreased width and depth of wear tracks can be explained as the increased hardness and toughness with the increased H/E* and H3/E*2 ratios. The elastic and plastic deformations resulted in crack initiation and propagation to produce wear debris and the delamination of the coatings. Additionally, the high compressive residual stress levels in the coatings with increasing Cr target power to 1000 W gave rise to the decrease in wear resistance due to the decrease in toughness and cohesion/adhesion strength [16,23,32]. The increased crack initiation resistance was related to a higher LC1, which was thought to be due to the high H/E* and H3/E*2 ratios. The increased LC3 was also benefited from the increased H/E* and H3/E*2 ratios, while the increase in residual stress state resulted in the change of oxidative wear from severe to mild, and then to severe. Although the initiated cracks led to the decrease in toughness, the coatings deposited at 400 W Cr target power showed the increased endurance of the coatings during dry sliding tests due to the enhanced adhesion strength with the increased LC3 [19,20,32]. Therefore, the enhanced toughness and cohesion/adhesion strength with the increased HF levels and critical loads of CrN/TiN multilayer coatings prevented the coating failure and the formation of wear debris during dry sliding tests. 4. Conclusions 1. CrN/TiN multilayer coatings were deposited by pulsed dc magnetron sputtering in a closed field unbalanced magnetron sputtering system. As the Cr target power was increased from 70 W to 1000 W, the Cr/(Cr + Ti) ratio of the coatings increased from 0.037 to 0.573. The coatings exhibited a single phase face-centered cubic structure with a texture evolution from (111) to (220). The coatings exhibited

a dense columnar structure. The coatings deposited at 400 W Cr target power showed distinct multilayers. 2. As the Cr target power was increased from 70 W to 1000 W, the hardness, the H/E* and H3/E*2 ratios, HF levels and LC (LC1, LC2 and LC3) of CrN/TiN multilayer coatings initially increased and then decreased. The coating deposited at 400 W Cr target power showed the highest hardness, the H/E* and H3/E*2 ratios of 31 GPa, 0.0832 and 0.214, respectively. Correspondingly, the coatings exhibited high toughness and cohesion/adhesion strength with high HF levels and LC. 3. As the Cr target power was increased from 70 W to 1000 W, the coefficient of friction and specific wear rate showed an initial decrease from 0.88 to 0.41 and from 11 × 10− 6 mm3 N−1 m− 1 to 2.3 × 10−6 mm3 N−1 m−1, followed by an increase to 0.62 and 7.5 × 10−6 mm3 N−1 m− 1, respectively. The coating deposited at 400 W Cr target power showed the lowest coefficient of friction and specific wear rate. 4. The transition of the dominant oxidative wear mechanism of CrN/TiN multilayer coatings from severe to mild, and then to severe was observed as increasing Cr target powers from 70 W to 1000 W. The coatings with similar H/E* and H3/E*2 ratios exhibited different tribological properties due to the different critical loads LC3 resulted from the increase in compressive residual stress. The improvements in the toughness and cohesion/adhesion strength of the coatings promoted the reduction in crack initiation and propagation, and oxidation wear during dry sliding tests. Acknowledgments Y. X. Ou acknowledges the financial support of the State Scholarship Fund awarded by the China Scholarship Council in 2011. This work is supported by the National Science Foundation of China under Grant Nos. 51271048 and 51321004. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

Q. Yang, L.R. Zhao, Surf. Coat. Technol. 173 (2003) 58. P.C. Yashar, W.D. Sproul, Vacuum 55 (1999) 179. C. Subramanian, K.N. Strafford, Wear 165 (1993) 85. X.T. Zeng, S. Zhang, C.Q. Sun, Y.C. Liu, Thin Solid Films 424 (2003) 99. J. Paulitsch, P.H. Mayrhofer, W.-D. Münz, M. Schenkel, Thin Solid Films 517 (2008) 1239. M. Nordin, M. Larsson, Surf. Coat. Technol. 116 (1999) 108. H.C. Barshilia, A. Jainb, K.S. Rajama, Vacuum 72 (2003) 241. Y.M. Zhou, R. Asaki, K. Higashi, W.H. Soe, R. Yamamoto, Surf. Coat. Technol. 130 (2000) 9. S.Y. Lee, G.S. Kim, J.H. Hahn, Surf. Coat. Technol. 177 (2004) 426. S.L. Lehoczky, J. Appl. Phys. 49 (1978) 5479. X. Chu, S.A. Barnett, J. Appl. Phys. 77 (1995) 4403. M. Kato, T. Mori, L.H. Schwartz, Acta Metall. 28 (1980) 285. Q. Yang, C. He, L.R. Zhao, J.-P. Immarigeon, Scr. Mater. 46 (2002) 293. C. Mendibide, J. Fontaine, P. Steyer, C. Esnouf, Tribol. Lett. 17 (2004) 779. A. Leyland, A. Matthews, Wear 246 (2000) 1. A. Matthews, S. Franklin, K. Holmberg, J. Phys. D. Appl. Phys. 40 (2007) 5463. J.F. Archard, J. Appl. Phys. 24 (1953) 981. J. Musil, M. Jirout, Surf. Coat. Technol. 201 (2007) 5148. S. Zhang, D. Sun, Y. Fu, H. Du, Surf. Coat. Technol. 198 (2005) 2. S. Zhang, D. Sun, Y. Fu, H. Du, Thin Solid Films 447 (2004) 462. Q. Luo, W.M. Rainforth, W.-D. Münz, Wear 225 (1999) 74. C. Mendibide, P. Steyer, J. Fontaine, P. Goudeau, Surf. Coat. Technol. 201 (2006) 4119. S.J. Bull, D.S. Rickerby, Proceedings of the 16th Leeds-Lyon Symposium on Tribology held at The lnstitut National des Sciences Appliquées, 17 1990, p. 337. A.A. Torrance, Wear 200 (1996) 45. Y. Rudermann, A. Iost, M. Bigerelle, Tribol. Int. 44 (2011) 585. S.J. Bull, Tribol. Int. 30 (1997) 491. H. Ichimura, Y. Ishii, Surf. Coat. Technol. 165 (2003) 1. W.W. Gerberich, N.I. Tymiak, J.C. Grunlan, M.F. Horstemeyer, M.I. Baskes, J. Appl. Mech. 69 (2002) 433. S.J. Bull, E.G. Berasetegui, Tribol. Int. 39 (2006) 99. J. Stallard, S. Poulat, D.G. Teer, Tribol. Int. 39 (2006) 159. B.D. Beake, V.M. Vishnyakov, R. Valizadeh, J.S. Colligon, J. Phys. D. Appl. Phys. 39 (2006) 1392. D. Philippon, V. Godinho, P.M. Nagy, M.P. Delplancke-Ogletree, A. Fernández, Wear 270 (2011) 541. B.D. Cullity, Elements of X-ray Diffraction, 2nd ed. Addison-Wesley, Reading, MA, 1978.

Y.X. Ou et al. / Surface & Coatings Technology 276 (2015) 152–159 [34] E. Vallat-Sauvain, U. Kroll, J. Meier, A. Shah, J. Pohl, J. Appl. Phys. 87 (2000) 3137. [35] A.J. Perry, J.A. Sue, P.J. Martin, Surf. Coat. Technol. 81 (1996) 17. [36] I.C. Noyan, J.B. Cohen, Residual Stress Measurement by Diffraction and Interpretation, Springer-Verlag, New York, 1987. [37] M. Bartosik, R. Daniel, C. Mitterer, J. Keckes, Surf. Coat. Technol. 205 (2010) 1320. [38] H.E. Howard, T.L. Gall, Metals Handbook, American Society for Metals, Metals Park, OH, 1985. [39] Verein Deutscher Ingenieure Normen, VDI 3198, VDI-Verlag, Dusseldorf, 1991. [40] D.R. Lide, CRC Handbook of Chemistry and Physics, 72nd ed. CRC press, Boston, 1991. [41] P.Eh. Hovsepian, D.B. Lewis, W.-D. Münz, Surf. Coat. Technol. 133-134 (2000) 166. [42] P. Yashar, S.A. Barnett, J. Rechner, W.D. Sproul, J. Vac. Sci. Technol. A 16 (1998) 2913. [43] J. Lin, W.D. Sproul, J.J. Moore, S. Lee, S. Myers, Surf. Coat. Technol. 205 (2011) 3226. [44] J. Pelleg, L.Z. Zevin, S. Lungo, N. Croitour, Thin Solid Films 197 (1991) 117.

[45] [46] [47] [48] [49] [50] [51] [52] [53] [54]

159

C.V. Thompson, R. Carel, J. Mater. Sci. Eng. B 32 (1995) 211. I. Petrov, P. Barna, L. Hultman, J. Greene, J. Vac. Sci. Technol. A 21 (2003) 117. H. Jiménez, E. Restrepo, A. Devia, Surf. Coat. Technol. 201 (2006) 1594. D.B. Lewis, I. Wadsworth, W.D. Münz, R. Kuzel Jr., V. Valvoda, Surf. Coat. Technol. 116 (1999) 284. J.P. Zhao, X. Wang, T.S. Shi, X.H. Liu, J. Appl. Phys. 79 (1996) 9399. H. Holleck, V. Schier, Surf. Coat. Technol. 76 (1995) 328. J. Lin, J.J. Moore, B. Mishra, M. Pinkas, X. Zhang, W.D. Sproul, Thin Solid Films 517 (2009) 5798. E.O. Hall, Yield Point Phenomena in Metals and Alloys, Mac-Millan, 1970. J.S. Wang, Y. Sugimura, A.G. Evans, W.K. Tredway, Thin Solid Films 325 (1998) 163. D.F. Diao, K. Kato, K. Hokkirigawa, J. Tribol. Trans. ASME 116 (1994) 860.