S

Ya

b

c

d

e

ARRAA

KSCMMM

1

waaffttcHpthmf

eoi

0h

Applied Surface Science 279 (2013) 189– 196

Contents lists available at SciVerse ScienceDirect

Applied Surface Science

j ourna l ho me page: www.elsev ier .com/ locate /apsusc

elf-lubricating CrAlN/VN multilayer coatings at room temperature

uexiu Qiua,b, Sam Zhangc,∗, Jyh-Wei Leed,e, Bo Lia, Yuxi Wangc, Dongliang Zhaoa

Central Iron and Steel Research Institute, 76 South Xueyuanlu Rd, Haidan District, Beijing 100081, ChinaAdvanced Technology & Materials Co., Ltd, 76 South Xueyuanlu Rd, Haidan District, Beijing 100081, ChinaSchool of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, SingaporeDepartment of Materials Engineering, Ming Chi University of Technology, 84 Gung Juan Road, Taishan, Taipei 24301, TaiwanCenter for Thin Film Technologies and Applications, Ming Chi University of Technology, 84 Gung Juan Road, Taishan, Taipei 24301, Taiwan

a r t i c l e i n f o

rticle history:eceived 14 January 2013eceived in revised form 12 March 2013ccepted 12 April 2013vailable online 25 April 2013

eywords:elf-lubricating

a b s t r a c t

Vanadium nitride (VN) is easily oxidized to form vanadium oxides and becomes lubricious under stress.CrAlN is hard thus CrAlN/VN multilayer coatings render both hardness and lubricious properties attrac-tive in dry machining of soft metals. This study investigates the effect of multi-layering on the coating’smechanical and tribological properties at room temperature. The CrAlN/VN multilayer coatings aredeposited on cemented tungsten carbide and Si wafer (1 0 0) substrates in an in-line magnetron sputter-ing system. A period contains one layer of CrAlN plus one adjacent layer of VN. The period thickness variesroughly from 3 nm to 30 nm; the total number of the periods varies from 30 to 300. X-ray diffractome-

rAlN/VNultilayer coatingsechanical performanceagnetron sputtering

try, scanning electron microscopy, transmission electron microscopy, X-ray photoelectron spectroscopyand electron probe micro-analyzer are employed to characterize the microstructures and chemistry.Nanoindentation and ball-on-disk wear test are used in mechanical and tribological studies. The CrAlN/VNmultilayer coatings have good lubricant property with lowest coefficient of friction of 0.26. At the periodthickness of 20 nm, the multilayer coatings obtained the best mechanical properties (hardness of 32.4 GPa,elastic modulus of 375 GPa, minimum wear rate of 1.1 × 10−7 mm3/Nm).

. Introduction

Dry machining is environmentally friendly and cost effectivehich becomes more and more popular. However, dry machining

lso poses great challenges to cutting tools [1] on soft metals suchs aluminum alloys. During machining, the soft metals exhibit highriction and strong adhesive interaction with tool materials there-ore a strong tendency of generating built-up edges (BUE). BUE ishe main reason for high cutting force, poor surface finish and shortool life [2]. To overcome BUE, the machine tools should be lubri-ious as well as hard and strong. Coating is an efficient way out.owever, transition metal nitride coatings like TiN, TiAlN and CrNerform poorly due to high coefficient of friction [2–4]; conven-ional solid lubricating coatings like diamond like carbon, MoS2, and-BN fail due to oxidization under extreme conditions like ambientoisture and elevated temperatures [5]; thus they are not suitable

or dry machining.Recently, vanadium nitride (VN) has attracted increasing inter-

sts as VN is easily oxidized to form Magnéli-phase vanadiumxides have easy slipping shear planes [5–7] and melt at approx-mately 685 ◦C [8] thus becomes ‘lubricious’ under stress and

∗ Corresponding author. Tel.: +65 67904400.E-mail address: msyzhang@ntu.edu.sg (S. Zhang).

169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.apsusc.2013.04.068

© 2013 Elsevier B.V. All rights reserved.

temperature. Series of V-based hard coatings have been developed,such as (V, Ti) N [9,10], CrVN [11], CrAlVN [12,13] compos-ite coatings, AlN/VN [14,15] and TiAlN/VN [2,16,17] multilayercoatings with good mechanical and tribological properties. InTiAlN/VN multilayer coatings, melting of VOx at 700 ◦C gives riseto low friction coefficient of 0.18 [16]. In our previous research,CrN/VN multilayer coatings demonstrated low coefficient of fric-tion (COF) of 0.21 due to formation of VOx during tribo-test at roomtemperature [18]. CrAlN coatings have good oxidation resistance,anticorrosive and anti-adhesive properties [19–21], but inadequatetribological resistance hinders their applications in dry machining.Logically, the combination of CrAlN and VN should provide an alter-native for a hard and lubricant coating for soft metals. In this study,CrAlN/VN multilayer coatings are deposited via magnetron sput-tering to study the effect of the CrAlN/VN period thickness (�)on the mechanical performance and tribological behavior at roomtemperature. At a proper period thickness, the coatings are of highhardness and low COF rendering low wear rate.

2. Experimental

2.1. Deposition of coatings

Six CrAlN/VN multilayer coatings (labeled as MA 1 though MA6) with the same thickness ratios of CrAlN:VN = 1:2 were deposited

190 Y. Qiu et al. / Applied Surface Sci

VN

CrAlN

VN

CrAlN••• •••••••••••

VN

CrAlN

CrAl interl ayer

Substrate

oTruCbcptrnsadw

2

mJsdP�im

TD

Fig. 1. Schematic of CrAlN/VN multilayer structure.

n polished YT15 cemented tungsten carbide (WC + 15 wt.%iC + 6 wt.% Co) and silicon wafer (100) substrates in an in-lineeactive magnetron sputtering system (PSUVHL-200C, HOPE Vac-um Technology Co., Ltd., Taiwan). The sputtering targets wererAl alloy (Cr: 70 at.%, Al: 30 at.%) and pure V (99.99 wt.% purity),oth of 152.4 mm in diameter. N came from the Ar/N2 atmosphereontrolled at 20 sccm/40 sccm. The coatings were prepared by twoulsed DC powers simultaneously. Through on–off and time con-rol of target shutters, alternate deposition of VN or CrAlN layer wasealized to form sequential CrAlN/VN coating with required thick-ess, finishing off with VN at the top, the schematic of multilayertructure is shown in Fig. 1. CrAlN and VN monolayer coatings werelso prepared and tested as reference. The designation and typicaleposition conditions were listed in Table 1, multilayer coatingsith varying period thickness labeled as MA 1 to MA 6.

.2. Characterization

The chemical composition of the multilayer coatings was deter-ined by field emission electron probe micro-analyzer (FE-EPMA,

XA-8500 F, JEOL, Japan) at 12 kV, 1.5 × 10−8 A. The crystallinetructure analyses were conducted using grazing incidence X-rayiffraction (to avoid the influence from the substrate) (GIXRD,

ANalytical-X’Pert PRO MPD) with Cu K� radiation (40 kV, 30 mA,

= 0.1541 nm) at a step size of 0.05◦ (0.3 s per step). The coat-ng thickness was evaluated by field emission scanning electron

icroscopy (FE-SEM, JEOL JSM-6701F, Japan) at 15 kV. The period

able 1esignation, deposition parameter, lattice constant, hardness, elastic modulus and coeffic

Sample designationa CrAlN VN MA 1

Designed period thickness (nm) – – 3

Period numbers – – 300

CrAl target power density (W/cm2) 2.2V target power density (W/cm2) 5.5Base pressure (Pa) 4 × 10−5

Plasma etching Ar plasma for 15 min at 0.4 Pa under substrateInterlayer 70 nm thick CrAl interlayer for multilayer coatWorking pressure (Pa) 0.4Substrate temperature (◦C) 300Ar:N2 gas ratio 1:2Total gas flow (sccm) 60RF substrate bias (V) −200 (13.56 MHz)Coatings thickness (nm) 989 758 842

Calculated period thickness (nm) – – 2.8

Lattice constant (nm) 0.4177 0.4139 0.4139

Hardness (GPa) 20.2 ± 1.8 28.5 ± 2.5 25.1 ± 3.2

Elastic modulus (GPa) 298 ± 16 344 ± 23 309 ± 23

Coefficient of friction 0.32 ± 0.02 0.28 ± 0.02 0.27 ± 0.01

a MA 1–6 are CrAlN/VN multilayer coatings.

ence 279 (2013) 189– 196

thicknesses of the multilayer coatings were obtained by the totalnumber of periods dividing the total coating thickness. Furthercharacterizations for the multilayer structure were performed bytransmission electron microscope (TEM, JEOL JEM-2100, Japan).Nanoindentation tests were carried out at a maximum load of5 mN with a nanoindenter (TI-900, Tribo-Indenter, Hysitron, USA)equipped with a Berkovich diamond tip angled at 142.3◦. Eightindentations were made on each coating. The maximum inden-tation depth was limited to around 50 nm, which was less thanone-tenth of the coating thickness. Ball-on-disk wear test was car-ried out to characterize the tribological performances. A tungstencarbide (WC) ball, 6 mm in diameter, was adopted as the counter-part. The tribo test was conducted at 24 ◦C and relative humidity of65% with a normal load of 2 N for a sliding distance of 1000 meters.The sliding linear speed was set as 84 mm/s at radius of 4 mm. TheCOF was determined after 50 m running-in period to minimize theinfluence from surface topology [22]. The wear scars were exam-ined by scanning electron microscopy (SEM, HITACHI S3400N) at15 kV. The wear rate of the coating was determined based on thefollowing equation [23]:

WR = t(3t2 + 4b2)2�r

6bFnS(1)

where t is the depth of the wear track determined by a surface pro-filometer (Kosaka Surfcorder ET3000), b is the width of the weartrack determined by SEM, r is the radius of the wear track, Fn

is the normal load and S is the sliding distance. The wear debriswas then examined with X-ray Photoelectron Spectroscopy (XPS,Kratos, Britain) for chemical bonding using Al K� (1486.71 eV) radi-ation (10 mA and 15 kV). Mixed Gauss/Lorentz peak deconvolutions(70% constrain) were used for peak fitting with Shirley-backgroundreduction (i.e., XPS Peak 4.1 program).

3. Results and discussion

3.1. Physics/chemistry and microstructure

The thickness of monolayer CrAlN, VN coatings were 989 nmand 758 nm, respectively. The thickness of multilayer coatings were842, 883, 928, 897, 903, 943 nm, respectively for MA 1 though MA 6;these values together with the respective period thickness are tabu-

lated in Table 1. The mean chemical compositions of the multilayercoatings are 12.6% Cr–4.8% Al–34.2% V–45.0% N–3.4% O (in at.%).The average atomic ratios of (Cr + Al):V of each coating is around1:1.97, very close to the designed 1:2.

ient of friction for CrAlN/VN multilayer and CrAlN, VN monolayer coatings.

MA 2 MA 3 MA 4 MA 5 MA 6

6 10 15 20 30150 90 60 45 30

bias − 500 Vings

883 928 897 903 9435.9 10.3 15.0 20.0 31.40.4140 0.4142 0.4142 0.4143 0.414024.8 ± 3.1 28.1 ± 2.5 27.7 ± 2.1 32.4 ± 3.6 29.4 ± 2.2300 ± 14 341 ± 24 339 ± 22 375 ± 26 345 ± 240.26 ± 0.01 0.32 ± 0.02 0.31 ± 0.02 0.31 ± 0.01 0.31 ± 0.02

Y. Qiu et al. / Applied Surface Sci

30 35 40 45 50 55 60 65 70 75 80

Diff racti on angle, 2θ (º )

CrAlN

VN

Inte

ns

ity

(a

rb.

un

its

)

MA 1, Λ = 2.8 nm

MA 2, Λ = 5.9 nm

MA 3, Λ = 10.3 nm

MA 4, Λ =15.0 nm

MA 5, Λ = 20.0 nm

MA 6, Λ = 31.4 nm

(200)(111) (220) (311)

MA 1- 6 are CrAlN/VN mu ltilayer coatings

Fn

Tsidsamsdfsa[edptsc

2

wa(c

d

ig. 2. GIXRD of CrAlN/VN multilayer coatings (MA 1–6) with different period thick-ess and CrAlN, VN monolayer coatings.

Fig. 2 presents GIXRD patterns of the as-deposited coatings.he monolayer CrAlN coating presents an FCC structure with atrong (220) peak and a weak (1 1 1) peak; the monolayer VN coat-ng has a strong (2 0 0) peak and weak (2 2 0), (3 1 1) peaks. Theifference in main orientation (texturing) should come from thetress involved during deposition. Deposition of monolayer VN wast higher power density (5.5 W/cm2), thus higher temperature orore stress relaxation occurred; as a result, the VN grains are of less

train, more align in the [2 0 0] direction. For CrAlN, however, theeposition took place at lower power density (2.2 W/cm2), there-ore, deposition annealing effect was not as high, resulting moretrained grains, giving rise to more alignment in [2 2 0] directions (2 2 0) accommodates higher strain energy grains than (2 0 0)24]. On the other hand, varying period thicknesses have no obviousffect on the XRD results, all the multilayer coatings have similariffraction patterns, strong (2 0 0) and weak (1 1 1), (2 2 0), (3 1 1)eaks. As V:(Cr + Al) is about 2:1 (it can also be seen later, thehickness of VN is about double CrAlN sublayer thickness), under-tandably, VN dominates the growth direction of the multilayeroating, giving rise to (2 0 0) texture.

From the Bragg equation [25],

dh k l sin � = n� (2)

here � is wavelength of X-rays, � is diffraction angle, h, k, and l

re the Miller indices; the inter-planar spacing (dh k l) for the planeh k l) in CrAlN, VN monolayer and CrAlN/VN multilayer coatingsould be obtained.

The cubic lattice parameter (a) for the cubic phase structure isetermined by Eq. (3) [25]:

ence 279 (2013) 189– 196 191

a = dh k l(h2 + k2 + l2)

1/2(3)

The corrected values of lattice parameter (lattice constant: a0)could be estimated from the Nelson–Riley function [25]:

a = a0 + Kcos2 �

sin �(4)

Eq. (4) represents a straight line between cos2 �/sin � (X-axis)and a (Y-axis). The slope of the line gives the K. The intercept ofthis line on Y-axis gives the a0 and tabulated in Table 1. The latticeconstants of monolayer CrAlN and VN are 0.4177 and 0.4139 nm,respectively; and there is about 1% mismatch. The a0 of CrAlNis larger than that of CrN (PDF card # 65-2899, a0 CrN = 0.4149)signifying part of the doped Al acting as interstitial atom in CrN lat-tice structure. The lattice constants of multilayer coatings (0.4139,0.4140, 0.4142, 0.4142, 0.4143, 0.4140 nm, respectively for MA1 though MA 6) are very close to that of the monolayer VN(0.4139 nm). As the lattice constant of the multilayer coating isnothing but a weighted average of the constituents of the sublay-ers, dominating VN understandably results the multilayer coating’slattice aligning toward that of the VN.

The TEM images and selected-area electron diffraction (SAED)pattern of the multilayer coating with period thickness of 20 nm(i.e., sample MA 5) are presented in Fig. 3; multilayer structure isobserved from Fig. 3(a). The columnar grains are around 50 nm inwidth (cf., Fig. 3(b)). The thickness of one period measured fromthe image is about 20 nm, and the CrAlN/VN layer thickness ratio isabout 1:2, in perfect agreement with design. In addition, the diffrac-tion rings (cf., Fig. 3(b)) shown in SAED pattern have been identifiedas the B1 NaCl type.

3.2. Mechanical properties

Nanoindentation hardness (H) and elastic modulus (E) of thecoatings are also tabulated in Table 1. The hardness and elastic mod-ulus of multilayer coatings increase with period thickness varyingfrom 3 nm to 20 nm; and at 20 nm, peaked at 32.4 GPa and 375 GPa,respectively. The plastic deformation resistance characteristics, orH3/E*2 [26], of CrAlN/VN multilayer coatings are plotted in Fig. 4,where E* = E/(1 − �2), E is the Young’s modulus, and � is the Poissonratio and H is the hardness of the coating. The tendency between theplastic deformation resistance and period thickness is same withhardness; H3/E*2 peaked with 0.4 GPa at � = 20 nm. As comparison,the values of CrAlN and VN monolayer coatings are also shown inFig. 4.

In the case of “thick” layers: based on Chu and Barnett’s model[27], when the CrAlN/VN period thickness is 20 nm and higher(more than the interlayer width), dislocation gliding within individ-ual layers determines the overall strength (�tot) of the multilayercoating, given by the scale average of the layer yield stresses �CrAlNand �VN:

�tot = �CrAlNlCrAlN

�+ �VN

lVN

�(5)

where lCrAlN and lVN are the CrAlN and VN sublayer’s thickness.Thus, the lower-limit (required to move a pre-existing dislocationin a layer) and upper-limit (when there are not sufficient disloca-tions already exist, such that generation of dislocations is required)hardness enhancement for CrAlN/VN multilayer coatings can beachieved respectively as follows:

�HLower-limit = �tot − �0

= 2˛b cos �

m�

[GCrAlN ln

(lCrAlN

b cos �

)+ GVN ln

(lVN

b cos �

)]

= 0.258 × �−0.737(GCrAlN + GVN) cos � (6)

192 Y. Qiu et al. / Applied Surface Science 279 (2013) 189– 196

Fig. 3. TEM images and inserted SAED pattern of CrAlN/VN multilayer coating with a period thickness of 20 nm (MA 5).

�

wtd

HUpper-limit = �tot − �0 = 4˛b cos �

m�

×[

GCrAlN ln(

lCrAlN

2b cos �

)+ GVN ln

(lVN

2b cos �

)]

= 0.344 × �−0.672(G + G ) cos � (7)

CrAlN VN

here �0 is the average stress required for dislocation slip in mul-ilayer material, m is the Taylor factor, b is the magnitude of theislocation Burgers vector, ̨ is �/4 for screw dislocations and

(1 − v)�/4 for edge dislocations, � is the smallest angle betweenthe interface and the glide planes of crystal, set as 45◦ for multi-layers [28], GCrAlN and GVN are the shear moduli of CrAlN and VNlayers (G = E/2(1 + �), E is the elastic modulus and � is the Poisson’sratio, 0.25 for CrAlN [29] and 0.279 for VN [30]). As the hardnessof multilayer coating is inversely proportional to the layer thick-

ness (i.e., �p), MA 5 with � = 20 nm thus has higher hardness thanthat of MA 6 with � = 30 nm. Using Eqs. (6) and (7), for MA 5, thehardness enhancement is calculated as �Hlower-limit = 5.1 GPa and�Hupper-limit= 8.2 GPa. From the rule of mixture, the hardness of

Y. Qiu et al. / Applied Surface Science 279 (2013) 189– 196 193

0 5 10 15 20 25 30 35

0.1

0.2

0.3

0.4

0.5H

3/E

*2 (

GP

a)

Per iod thickne ss of CrAlN/VN multil ayer coat ings (nm)

VN

CrAlN

CrAlN/VN

Fig. 4. H3/E*2 ratio of CrAlN, VN monolayer coatings and CrAlN/VN multilayerc

tF6m

t3C

0 5 10 15 20 25 30 35

0

1

2

3

4

5

6

7

We

ar

rate

(×

10

-7 m

m3/

Nm

)

Per iod t hickness of CrAlN/VN multil ayer coat ings ( nm)

VN

CrAlN

CrAlN/ VN

Fig. 5. Wear rate of CrAlN, VN monolayer and CrAlN/VN multilayer coatings with

F2

oatings with various period thicknesses.

he multilayer coating should be 25.7 GPa (=(HCrAlN + 2 × HVN)/3).or MA 5, however, the hardness is 32.4 GPa, or enhancement of.7 GPa, falling right in the calculated range of hardness enhance-ent limit (5.1–8.2 GPa).In the case of “thinner” layers: as the period thickness decrease

o 10 nm and 15 nm (MA 3 and MA 4 in which the CrAlN layer is.3 nm and 5 nm respectively), taken away the interface width, therAlN layer is too thin to contain dislocations. As such, the hardness

ig. 6. Wear scars SEI and BEI images of multilayer coatings; (a) and (b) for the coating

0 nm.

different period thickness.

is determined by the dislocation motion in VN layer (6.7 nm and10 nm respectively), resulting similar hardness to VN coatings.

In the case of “extremely thin” layers: as the period thicknessdrops to 3 and 6 nm (MA 1 and MA 2 respectively), CrAlN sublayer

is only 1 nm or 2 nm and the VN sublayer 2 or 4 nm, comparablewith interface width; in addition, as CrAlN and VN have similar dis-location slipping systems with good miscibility, dislocation gliding

with period thickness of 3 nm, (c) and (d) for the coating with period thickness of

194 Y. Qiu et al. / Applied Surface Science 279 (2013) 189– 196

F he balp

aR

3

tCTemhm

ig. 7. Optical images of wear scars for tungsten carbide balls after tribo-test; (a) teriod thickness of 20 nm.

cross interfaces thus determines the hardness, giving rise to theule of Mixture effect.

.3. Tribological properties

Coefficients of friction of the as-deposited multilayer coatingsogether with the monolayer counterparts are shown in Table 1.OF of the monolayer CrAlN and VN is 0.32 and 0.28 respectively.he COF of their multilayer counterparts are not much differ-

nt: 0.26–0.32. That is expected. Fig. 5 plots the wear rate of theultilayer coatings via ball-on-disk test. The multilayer coatings

ave wear rate from 1.1 to 6.7 × 10−7 mm3/Nm while that of theono layer CrAlN and VN coatings are 1.8 × 10−7 mm3/Nm and

l for the coating with period thickness of 3 nm and (b) the ball for the coating with

4.0 × 10−7 mm3/Nm, as inserted in Fig. 5. Fig. 6(a) and (b) is thewear scars of MA 1 (SEI and BEI images respectively). Fig. 6(c)and (d) is that of MA 5. Comparing a and c or b and d, MA 1’swear scars have much more wear debris than that of MA 5 inthe wear tracks. Obviously the wear rate was controlled by plasticdeformation resistance, the coating with higher H3/E*2 has lowerwear rate as well as less wear debris. The lowest wear rate of1.1 × 10−7 mm3/Nm occurs at � = 20 nm, consistent with the high-est H3/E*2 response. Fig. 7 (optical images) shows the wear scars of

the tungsten carbide balls. For MA 1, the worn-out diameter is about0.26 mm but for MA 5, the diameter is much larger or 0.29 mm. Thisagrees with the debris observation in the wear track: for softer coat-ing (MA 1), more debris in the track and less wear in the ball while

Y. Qiu et al. / Applied Surface Science 279 (2013) 189– 196 195

525 52 0 51 5 51 0

V 2p3/2

a. MA 5,CrAlN/VN, Λ= 20 nm

V 2p3/2

fro m coat ing

Inte

ns

ity

(a

rb.

un

its)

Bindi ng E nergy ( eV)

VN

V2O

3

V2O

5

525 52 0 51 5 51 0

b. MA 5,CrAlN/VN, Λ= 20 nm

V 2p3/2

fro m wear debr is

Inte

ns

ity

(a

rb.

un

its)

Bindi ng E nergy ( eV)

V 2p3/2

VNV

2O

3V

2O

5

75 70 65

Al 2p3/2

c. MA 5,CrAlN/VN, Λ= 20 nm

Al 2p3/2

from coat ing

Inte

ns

ity

(a

rb.

un

its

)

Bindi ng E nergy (eV)

AlN

75 70 65

d. MA 5,CrAlN/VN, Λ= 20 nm

Al 2p3/2

from wear debr is

Inte

ns

ity

(a

rb.

un

its

)

Bindi ng E nergy (eV)

Al 2p3/2

Al2O

3

AlN

F the w( 3/2 of w

fiicr

sto7FctsGtamarV(dowwFoG

ig. 8. XPS spectra of V 2p3/2 and Al 2p3/2 from coating and wear debris adhered ona) V 2p3/2 of coating, (b) V 2p3/2 of wear debris, (c) Al 2p3/2 of coating and (d) Al 2p

or harder coating (MA 5), less debris in the track and more wearn the ball, which mean the harder the coating is, more of the balls worn resulting in a larger area of contact, and accordingly theontact pressure will decrease. A decrease in pressure should alsoesult in lower wear rate of the coating.

The reasons of the self-lubricity lie in the formation of VOx,een clearly in Fig. 8 that illustrates the XPS V 2p3/2 spectra ofhe coatings and wear debris of � = 20 nm. The binding energiesf V2O3, V2O5, AlN and Al2O3 are 515.9 eV, 517.1 eV, 73.8 eV, and4.3 eV, respectively [31]; and that of VN is 514.3 eV [32]. Fromig. 8(a) and (b), vanadium exists as VN, V2O3 and V2O5 in bothoating and wear debris. Estimated from the fitted XPS peaks area,he content of VOx (V2O5+ V2O3) in unworn coatings is about 28%;imilar phenomena can be found in the work of Franz et al. [12] andlaser et al. [33]. As the multilayering ends with the VN sublayer on

he top, and at room temperature, VN is typically oxidized severaltomic layers following the so-called “low-temperature oxidation”echanism [34]. Since XPS technique is only sensitive to a depth of

bout 10 nm [35], the VOx percentage result thus only reflects theatio within the XPS detectable depth (unworn or worn), not theOx bulk content of the coating. After tribo-test, the content of VOx

V2O5+ V2O3) increases from 28% (coating) to 35% (wear debris)ue to tribo-oxidized VN. For the coatings with period thicknessf 3 nm, it has lower plastic deformation resistance thus higherear rate, more wear debris found in the wear track (see Fig. 6)

hich result in more VOx Magnéli-phase slip planes and lower COF.

ig. 8(c) and (d) indicates that part of the aluminum element wasxidized from AlN to Al2O3 after ball-on-disk wear test. Since theibbs free energy of Al2O3 is −1582.0 kJ/mol which is lower than

orn surface of CrAlN/VN multilayer coating with period thickness of 20 nm (MA 5);ear debris.

that of V2O5 (−1419.6 kJ/mol) and V2O3 (−1139.3 kJ/mol) [36], dur-ing tribo-test, Al can be oxidized preferentially thus partially retardthe formation of vanadium oxides. This actually benefits retainingof hardness. In coatings with higher H3/E*2 thus lower wear rate,less debris formed in the wear tracks (see Fig. 6), thus, result in lessVOx and higher COF.

4. Conclusions

Magnetron sputter formation of CrAlN/VN multilayer coatingsfrom CrAlN and VN generates about 1% lattice mismatch in betweenthe sublayers. The multilayer coatings are found self-lubricatingwith a coefficient of friction of 0.26 at room temperature because offormation of VOx. At period thickness of 20 nm, the coating exhibitnanoindentation hardness of 32.4 GPa, elastic modulus of 375 GPaand low wear rate of 1.1 × 10−7 mm3/Nm against cemented tung-sten carbide.

Acknowledgements

The authors gratefully acknowledge financial support ofthe National Natural Science Foundation of China (Grant No.51202035), National Science Council, Taiwan (NSC 100-2221-E-131-008-MY3), and this work is also supported by CISRI grant:11020990A.

References

[1] B. Podgornik, S. Hogmark, O. Sandberg, Wear 261 (2006) 15–21.

1 ce Sci

[[

[

[

[[[

[[

[[

[

[

[

[[

[

[[

[

[[

[

[

96 Y. Qiu et al. / Applied Surfa

[2] P. Eh Hovsepian, Q. Luo, G. Robinson, M. Pittman, M. Howarth, D. Doerwald, R.Tietema, W.M. Sim, A. Deeming, T. Zeus, Surface and Coatings Technology 201(2006) 265–272.

[3] M. Lahres, P. Müller-Hummel, O. Doerfel, Surface and Coatings Technology 91(1997) 116–121.

[4] M. Nouari, G. List, F. Girot, D. Coupard, Wear 255 (2003) 1359–1368.[5] P.H. Mayrhofer, C. Mitterer, L. Hultman, H. Clemens, Progress in Materials Sci-

ence 51 (2006) 1032–1114.[6] W.-D. Münz, MRS Bulletin 28 (2003) 173–179.[7] O. Storz, H. Gasthuber, M. Woydt, Surface and Coatings Technology 140 (2001)

76–81.[8] T. Kitagawa, A. Kubo, K. Maekawa, Wear 202 (1997) 142–148.[9] J.H. Ouyang, S. Sasaki, Surface and Coatings Technology 187 (2004) 343–357.10] J.H. Ouyang, T. Murakami, S. Sasaki, Wear 263 (2007) 1347–1353.11] M. Uchida, N. Nihira, A. Mitsuo, K. Toyoda, K. Kubota, T. Aizawa, Surface and

Coatings Technology 177–178 (2004) 627–630.12] R. Franz, J. Neidhardt, B. Sartory, R. Tessadri, C. Mitterer, Thin Solid Films 516

(2008) 6151–6157.13] R. Franz, J. Neidhardt, R. Kaindl, B. Sartory, R. Tessadri, M. Lechthaler, P. Polcik,

C. Mitterer, Surface and Coatings Technology 203 (2009) 1101–1105.14] J.-K. Park, Y.-J. Baik, Materials Letters 62 (2008) 2528–2530.15] G. Li, J. Lao, J. Tian, Z. Han, M. Gu, Journal of Applied Physics 95 (2004) 92–96.16] P.H. Mayrhofer, P. Eh Hovsepian, C. Mitterer, W.-D. Münz, Surface and Coatings

Technology 177–178 (2004) 341–347.17] Q. Luo, Wear 271 (2011) 2058–2066.

18] Y. Qiu, S. Zhang, B. Li, Y. Wang, J.W. Lee, F. Li, D. Zhao, Surf. Coat. Technol. (2012),

http://dx.doi.org/10.1016/j.surfcoat.2012.03.010.19] Y.C. Chim, X.Z. Ding, X.T. Zeng, S. Zhang, Thin Solid Films 517 (2009) 4845–4849.20] X.Z. Ding, A.L.K. Tan, X.T. Zeng, C. Wang, T. Yue, C.Q. Sun, Thin Solid Films 516

(2008) 5716–5720.

[[

[

ence 279 (2013) 189– 196

21] J.L. Mo, M.H. Zhu, B. Lei, Y.X. Leng, N. Huang, Wear 263 (2007)1423–1429.

22] V. Podgursky, R. Nisumaa, E. Adoberg, A. Surzhenkov, A. Sivitski, P. Kulu, Wear268 (2010) 751–755.

23] C.L. Li, F.B. Wu, J.W. Lee, Y.Z. Tsai, L.C. Chang, Surface and Coatings Technology204 (2009) 941–946.

24] J.-M. Zhang, K.-W. Xu, V. Ji, Applied Surface Science 185 (2002) 177–182.25] B.D. Cullity, Elements of X-ray Diffraction, 2nd ed., Addison-Wesley Publishing

Company, Inc, Reading, MA, 1978.26] J. Musil, F. Kunc, H. Zeman, H. Poláková, Surface and Coatings Technology 154

(2002) 304–313.27] X. Chu, S.A. Barnett, Journal of Applied Physics 77 (1995) 4403–4411.28] Chien-Ming Kao, Jyh-Wei Lee, Hsien-Wei Chen, Yu-Chen Chan, Jenq-

Gong Duh, Shin-Pei Chen, Surface and Coatings Technology 205 (2010)1438–1443.

29] K. Zheng, P. Liu, W. Li, F. Ma, X. Liu, Applied Surface Science 257 (2011)9583–9586.

30] P. Lazar, J. Redinger, R. Podloucky, Physical Review B 76 (2007) 1–9.31] B.V. Crist, Handbook of Monochromatic XPS Spectra, vol. 1: The Elements and

Native Oxides, XPS International Inc., Mountain View, California, USA, 1999.32] Binding Energies for Relevant Bonds obtained from the XPS Database of the

National Institute of Standards and Technology, http://srdata.nist.gov/xps/(10/2012).

33] A. Glaser, S. Surnev, F.P. Netzer, N. Fateh, G.A. Fontalvo, C. Mitterer, SurfaceScience 601 (2007) 1153–1159.

34] N. Cabrera, N.F. Mott, Reports on Progress in Physics 12 (1948–1949) 163–184.35] S. Zhang, L. Li, A. Kumar, Materials Characterization Techniques, CRC Press

Taylor & Francis Group, 2008, pp. 27.36] J.A. Dean, Lange’s Handbook of Chemistry, vol. 9, 12th ed., McGraw-Hill Book

Company, New York, NY, 1979, pp. 4–94.