Characterization of tribological behavior and wear mechanismsof novel oxynitride PVD coatings designed for applicationsat high temperatures

Jiri Nohava a,1, Pascal Dessarzin b,2,3, Pavla Karvankova b,4, Marcus Morstein b,n

a Anton Paar, Rue de la Gare 4, 2034 Peseux, Switzerlandb PLATIT AG, Eichholzstrasse 9, 2545 Selzach, Switzerland

a r t i c l e i n f o

Article history:Received 15 May 2014Received in revised form18 August 2014Accepted 21 August 2014Available online 30 August 2014

Keywords:High-temperature wearOxynitrideOxideRotating arc cathodes PVD coating

a b s t r a c t

This paper focuses on high-temperature tribotests of nanostructured Al–Cr-based oxynitride and oxidecoatings. Pin-on-disk tribological tests were performed at temperatures in between room temperatureand 800 1C, and the results compared to a conventional AlTiN coating. The AlCrN and AlCrON coatingsshowed good to acceptable wear resistance up to 600 1C while at 800 1C, both coating types failed. Thenew corundum-structured (Al,Cr)2O3 oxide showed excellent wear performance up to 800 1C. Imagingand chemical analysis of the high-temperature wear tracks allowed for explaining the differences in thewear mechanisms. The ability of a coating to resist oxidation and to delay substrate oxidation, togetherwith an excellent abrasional wear resistance, were identified as the major factors influencing wear ratesunder severe high-temperature conditions.

& 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Over the recent years, newly developed protective coatings forcutting tools have become highly resistant to wear if exposed to theextreme environments associated with modern machining processes.These coatings are generally deposited by various PVD methods andhave often been based on Ti1�xAlxN PVD coatings [1–5]. Such titaniumaluminum nitride coatings have been established as a standard formany years in a wide range of applications but they commonly fail athigh temperatures mainly due to oxidative wear [6–10]. In the recentyears, Ti was either completely replaced by Cr resulting in Cr1�xAlxNcoatings, or Cr was added resulting in Ti1�x�yCrxAlyN films [11–13].For these chromium-containing films, the wear resistance as well asthe oxidation resistance is higher than for titanium based coatings,thus offering better wear and oxidation protection to the tool[5,10,14,15]. As an alternative approach, Ti1�xAlxN/SiNx, Cr1�xAlxN/SiNx or Ti1�x�yCrxAlyN/SiNx [16–20] silicon-containing nanocomposite

coatings have also been introduced, mainly for cemented carbideshank tools up to intermediate cutting temperatures.

A new generation of different coatings has recently beendeveloped especially for the highest temperatures occurring inpractice, such as in metal cutting processes using indexablecutting inserts. These coatings use the addition of oxygen to formoxynitride coatings, that provide even better oxidation resistancethan nitride coatings [21]. However, substantial improvement ofcutting and wear performance could be achieved only by devel-oping an alpha-phase aluminum–chromium oxide structure in thecoating in order to prevent efficiently the oxidation at hightemperature [4,21–23]. These new α-(Al,Cr)2O3 (α-oxide) basedcoatings are known to withstand extremely high temperatures indry milling and turning of steels and other high-strength materialswhile exhibiting high wear resistance. While these new coatingsfeature a very promising cutting performance, their wear andfriction properties at high temperature have not yet been thor-oughly studied. The common pin-on-disk tribological tests onthese coatings have failed, resulting in practically no wear. Sincetribological properties including coefficient of friction (CoF) andwear rate are crucial in the development process of these coatings,a new testing procedure had to be established. This procedureneeded to be able to determine the wear resistance of the coatingsnot only at room temperature but at high temperatures (�800 1C)in particular. Evaluation of the pin-on-disk results at roomtemperature is usually quite straightforward using profilometermeasurements of the wear track cross-section for calculation of

Contents lists available at ScienceDirect

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

Tribology International

http://dx.doi.org/10.1016/j.triboint.2014.08.0160301-679X/& 2014 Elsevier Ltd. All rights reserved.

n Corresponding author. Tel.: þ41 32 544 6212.E-mail addresses: jiri.nohava@anton-paar.com (J. Nohava),

pascal.dessarzin@fhnw.ch (P. Dessarzin), p.karvankova@platit.com (P. Karvankova),m.morstein@platit.com (M. Morstein).

1 Tel.: þ41 32 557 5651.2 Tel.: þ41 56 202 7700.3 Current address: University of Applied Sciences Northwestern Switzerland,

Institute of Product and Production Engineering, Olten, Switzerland.4 Tel.: þ41 32 544 6238.

Tribology International 81 (2015) 231–239

the wear rate. However, analysis of high temperature pin-on-diskwear tracks requires more careful and complex evaluation, sincemany other phenomena such as oxidation can affect the testresults [6,12,25–27]. Therefore additional analyses have to beperformed in order to properly evaluate the wear performanceof the coatings of the various coatings.

This study focuses on comparing the tribological behavior of anAlTiN-based reference, a nanostructured Al–Cr-based nitride and aseries of oxynitride AlCrOxN1�x and oxide (Al,Cr)2O3 coatingsdeposited on cemented carbide using an industrial rotatingcathodes arc PVD process [24]. The tribological properties areevaluated via room temperature and high temperature (600 1Cand 800 1C) pin-on-disk tests, followed by investigation of thewear tracks by various analytical methods, in order to understandthe high-temperature wear mechanisms and to demonstrate thedifferences in failure mechanisms of the various coatings.

2. Experimental details

2.1. Coating deposition and analysis

Four types of coatings were used in this study: AlTiN and AlCrNnitride coatings, an AlCrON oxynitride coating and a α-(Al,Cr)2O3

oxide layer. All coatings were deposited using a π311 (PLATIT AG,Switzerland) industrial arc PVD system equipped with three lateral(LARC) and one central (CERC) rotating arc cathodes. The processeswere carried out either in a pure nitrogen atmosphere (for thenitrides) or a N2/O2 mixture (N2/O2 flow ratio Z9:1 for theoxynitrides and E4:1 for the oxide coating) at 4 Pa pressure usinga medium-frequency unipolar pulsed bias voltage from �30 V to�100 V. During the deposition, the single side polished WC–Cosubstrates (Ø50� d¼10 mm, EMT 100, Extramet AG, Switzerland)were heated to 550 1C. The layer architecture of the coatings was(from the substrate to the top): TiN adhesion layer, AlTiN, AlTiN/SiNx nanocomposite layer and a functional top layer which wasvaried. The top layer was selected as either AlCrN, AlCrOxN1�x orα-(Al,Cr)2O3 (see Table 1). The AlTiN reference sample containedonly the TiN and AlTiN layers without any additional top layer. Theoverall coating thickness was �4 μm for all samples.

The chemical composition of the films was quantitativelyanalyzed by calibrated wavelength dispersive (WDX) ElectronProbe Micro Analysis. Annealing and oxidation tests were con-ducted at 900 1C in a gas-tight horizontal tube furnace under N2/H2 and dry air flows, respectively, using a heating ramp of 5 1C/min, a holding time of 120 min (annealing test) or 60 min (oxida-tion test), and natural oven cooling. XRD phase analysis wasperformed using Cu Kα radiation and Θ¼11 grazing incidenceon equivalent samples containing only the functional top layer.

Room temperature nanoindentation testing of the coatings wasperformed using a Nanoindentation Tester (NHT, CSM InstrumentsSA, Switzerland). This instrument was calibrated using a standardprocedure on fused silica reference sample (Efs 7372 GPa) imme-diately previous to testing. A minimum of 20 indentations were

performed at maximum loads of 30 mN, 50 mN and 70 mN atloading rates of 60 mN/min, 100 mN/min and 140 mN/min, respec-tively, and a holding time at peak load of 1 s. The values ofhardness and Young’s modulus were extracted from the load–displacement curves using the Oliver and Pharr method [28]assuming a Poisson’s ratio of ν¼0.3 for the coatings. Since noload dependence was found, the values for hardness and Young’smodulus in Table 1 were averaged from all indentation results.

2.2. Tribological experiments

Three series of wear tests were performed on a High-TemperaturePin-on-Disk Tribometer (CSM Instruments SA, Switzerland) at tem-peratures of 24 1C (room temperature, RT), 600 1C and 800 1C. Theconditions of the pin-on-disk tests varied according to test tempera-ture and partially also according to sample type (longer tests for morewear resistant coatings) in order to generate measurable wear. Allsamples were mechanically polished after deposition using Master-Prep 0.05 mm alumina suspension (Bühler, Switzerland) in order tominimize roughness to a comparable level. The same media was usedto recondition the surface in between the high-temperature tribotestruns. As static friction partner, ∅6mm polycrystalline Al2O3 balls wereused. Due to its high abrasive and oxidation resistance, this material isfrequently used in high temperature tribology [5,29,30]. Preliminarytests were performed also with a ∅6mm Si3N4 ball but the ball didnot withstand high temperatures as also found by other groups[31,32]. The main pin-on-disk testing conditions are summarized inTable 2.

The cross-sectional profile of the wear track on the samples wasmeasured by means of a Form TalySurf surface profilometer (TaylorHobson, Great Britain) in six areas equally distributed around the weartrack. The wear rate w of the coating was calculated as

W¼ Vd� P

ð1Þ

where V is the volume of the removed material from the sampledetermined by surface profilometry, d is total distance and P is normalload. The pile-up and hollow scar areas were calculated automaticallyusing a standard procedure in the MountainsMap software (DigitalSurf, Besancon, France) where the pile-up and hole is defined as area

Table 1Chemical composition, structure and mechanical properties of the tested coatings.

Coating O content[at% of (OþN)]

Al content[at% of metals]

XRD phasesas deposited

RT hardness [GPa] ΔH [GPa] and XRDphases after2 h/900 1C in N2/H2

RT indentationmodulus [GPa]

Applicationtemperature [1C]

AlTiN r0.5 56.6 fcc 33.970.8 –3.1; fcc* 550715 Up to 600AlCrN r0.5 50.3 fcc 33.571.2 –3.5; fcc 610716 Up to 700AlCrON 73 50.7 fcc 34.671.0 –0.4; fcc 467711 600–800(Al,Cr)2O3 Z99.8 54.0 α-(Al,Cr)2O3 26.270.9 þ0.2; α 446716 600–1000

n Spinodal decomposition into fcc-TiNþfcc-AlN.

Table 2Pin-on-disk parameters used in tribological testing.

Testing temperature

24 1C 600 1C 800 1C

Normal load [N] 10 7 7Distance [m] 2,000 2,000 1,200*

Revolutions 39,700 31,200 32,000*

Linear speed [m/s] 0.2 0.2 0.2Test duration �165 min �165 min �165 min*

n Parameters for the α-(Al,Cr)2O3 coating: distance 3000 m, 40,000 rotations,duration �255 min.

J. Nohava et al. / Tribology International 81 (2015) 231–239232

above/below a mean line of the surface profile. The mean line iscalculated from the surface profile not containing the wear trackinformation. The wear rate of the alumina ball was calculated aftereach test using the volume of the worn cap and the same formula (1)was used for calculation of wear rate of the ball.

Scanning electron microscopy (SEM) images and energy dispersiveX-ray spectroscopy (EDX) analysis were carried out at 20 keV in aZEISS EVO scanning electron microscope (ZEISS, Germany). Thefocused ion beam (FIB) cuts and the corresponding SEM images wererealized on a Lyra3 system (Tescan, Czech Republic).

After the full set of analysis had been completed, the coatingsurface was re-polished and the next test was done with differenttemperature and radius on the sample. Care was taken to beginwith the ambient temperature measurement in order to avoid thenecessity to polish away thick oxide layer that can form at hightemperature. The residual coating thickness was greater than thewear track depth in all cases.

3. Results

The main chemical, physical and structural properties of thetested coatings are summarized in Table 1. All tested coatingsshare an aluminum content of slightly above 50 at% of the metals.The two nitride and the oxynitride coatings all have a single-phasefcc cubic structure and a room temperature hardness well above30 GPa, which decreases about 10% after annealing at 900 1C ininert atmosphere except for the oxynitride, which nearly retainsits hardness. The cubic phase of these three nitrogen-containingcoatings is maintained during the annealing test, and no hexago-nal phase formation is observed. The test conditions represent anupper estimate for the temperature in the wear track duringtesting at the highest macroscopic temperature, 800 1C, given thata friction induced heating is expected to contribute less than 100 Kat this already high temperature. In comparison with the nitrideand oxynitride coatings, the (Al,Cr)2O3 oxide starts off with a lowerroom temperature hardness, which is however even slightlyincreased by the annealing. As expected, no phase transformationis observed here, too.

Another difference between the tested coatings is their oxida-tive behavior. As evident from Fig. 1, the oxide scale thickness ofthe AlTiN sample is rather high after annealing in air at 900 1C,while it is drastically thinner for the AlCrN counterpart. Both of theoxygen-containing AlCr-based coatings show hardly detectableimpact of the one-hour treatment on the coating. Based on theseresults, one would thus expect both the AlCrOxN1�x and the (Al,Cr)2O3 coatings to be valid candidates for high-temperatureapplications, from an oxidative point of view.

3.1. Coefficient of friction

The evolution of coefficient of friction (CoF) during the pin-on-disk tribological tests at room temperature, 600 1C and 800 1C issummarized in Fig. 2. The coefficient of friction of the AlTiN, AlCrNand AlCrON coatings was stable at room temperature while at600 1C and 800 1C, the CoF varied strongly during the measure-ment, indicating important wear damage. Optical images and SEMobservation of the wear track including EDX elemental mappingrevealed areas with severe wear of the AlTiN coating after the600 1C and 800 1C tests. The α-(Al,Cr)2O3 oxide coating, on theother hand, presented very stable CoF at all temperatures. Further-more, the average value of the CoF of this oxide coating decreased

Fig. 1. Oxide scale measured by calotest after oxidizing different coatings at 900 1C.

Fig. 2. Comparison of the coefficient of friction (CoF) against alumina ball fordifferent high-performance PVD coatings at (a) room temperature, (b) 600 1C and(c) 800 1C. Note the excellent stability and low value of the CoF of α-(Al,Cr)2O3.

J. Nohava et al. / Tribology International 81 (2015) 231–239 233

from �0.5 at room temperature to �0.3 at 800 1C indicatingexcellent stability of this coating even at very high temperatures.

3.2. Wear rate

The wear rate of the coatings was calculated from the weartrack profiles according to Eq. (1) for each of the tested coatings atroom temperature, 600 1C and 800 1C. Although the volume of thematerial removed during the tests was an adequate measure ofwear rate at room temperature, at high temperatures also the so-called ‘build-up’ had to be taken into account. Build-up [33] wasconsidered as the growth or re-deposition of material around or inthe wear track. The build-up rate was calculated analogically to thewear rate according to Eq. (1). The wear rates and build-up ratesfor all tested coatings are summarized in Fig. 3.

Room temperature: Comparison of the wear rates at roomtemperature revealed the highest wear for the AlTiN coatings.Wear rates of the AlCrN, AlCrON and α-(Al,Cr)2O3 coatings werealmost non-measurable. This is in agreement with very low wearrates for AlCrN and AlCrON coatings reported in the literature[5,12].

600 1C: At 600 1C the AlTiN-based coating showed the highestwear out of all tested coatings, although the wear rate wasremarkably lower than for the test of the same coating, performedat room temperature. The wear rate of both the AlCrON and theAlCrN coating increased in comparison to room temperature.However, the most significant change observed was the apparitionof build-up on these three coatings. The build-up was not presentat room temperature and can therefore be considered as atemperature-activated process. The build-up rate on the AlCrNcoating was significantly higher than that on the AlCrON coating.In contrast, the wear rate of the α-(Al,Cr)2O3 oxide coating wasvery low and no build-up on this coating was observed.

800 1C: Wear of nitride and oxynitride (AlTiN, AlCrN, AlCrON)coatings was characterized by high wear rates with importantbuild-up rates. Interestingly, the wear rate of the AlTiN coating at800 1C apparently decreased with respect to the wear rate of thesame coating at 600 1C and at room temperature. This ‘anomalous’behavior will be discussed in more detail later. The results of thepin-on-disk tests of the α-(Al,Cr)2O3 oxide coating at 800 1Crevealed very low wear rate with only minor build-up.

3.3. Wear track morphology and wear mechanisms

The typical morphology of the wear tracks on the nitride andoxynitride samples was characterized by two features: materialremoval and build-up. The hollowing out was created due toerosive material removal during the wear tests while build-up wasmainly due to re-deposition of the wear debris originating fromthe displaced material from both the wear track and the aluminaball. Both the hollowing out and the build-up morphologicalfeatures were observed on all nitride and oxynitride samples toa various extent, depending on the type of the coating andtemperature. At room temperature, the hollowing out was presentonly on the AlTiN and the AlCrON coating. At this temperature,build-up was observed on neither of the tested coatings. Fig. 4shows the typical wear track morphology on AlTiN and α-(Al,Cr)2O3 after room temperature pin-on-disk tests. The wear trackon the oxide coating is almost invisible.

The wear of the coatings at room temperature was governedmainly by an abrasive mechanism with potential contributions ofmicro-scale cohesive fracture, as observed also in Ref. [33]. The mainwear mechanisms at high temperature (see Table 3 and Figs. 7 and 8)were oxidative attack accompanied by gradual material removal[4,30]. On coatings with lower wear resistance also formation ofbuild-up was observed at high temperatures (600 1C and 800 1C). The

optical and SEM observation of the wear track obtained at 600 1C and800 1C on the α-(Al,Cr)2O3 oxide coating revealed only very shallowand uniform wear track. The build-up on this coating was non-measurable at 600 1C and only small build-up was observed afterthe 800 1C pin-on-disk tests. Figs. 4–5 show a comparison of the weartrack on the AlTiN and α-(Al,Cr)2O3 coating after the room tempera-ture and 800 1C tests. Note the large build-up on the AlTiN coatingafter the 800 1C tests and very low damage on the α-(Al,Cr)2O3

coating. Especially at 800 1C, point wise local oxidation of the coatedcemented carbide surface could be observed, in the form of cobalt-tungsten oxide features growing out of pinhole defects. However, thisdid not noticeably affect the overall wear behavior except for the AlTiNcoating, where the main part of the wear track was found to consist ofoxidized substrate material after the pin-on-disk test at 800 1C (Fig. 6).

Fig. 3. Wear rates of the four tested coatings at room temperature, 600 1C and800 1C. The filled columns represent the wear rate; the patterned columnsrepresent the build-up rate. The height of columns exceeding the Y-axis scale isgiven at the bottom of the corresponding column.

J. Nohava et al. / Tribology International 81 (2015) 231–239234

3.4. Wear of the alumina ball

The results of the wear of the alumina ball are summarized inTable 4. The wear of the ball reflected the wear resistance of thetested coatings: the ball wear rate was highest against the AlTiNcoating and lowest against the α-oxide coating. Even at 800 1C, thewear of the ball against the α-oxide coating was hardly measur-able. In general, the ball wear rate was increasing with increasingtemperature.

4. Discussion

4.1. Nitride and oxynitride coatings

The AlTiN, AlCrN and AlCrON showed increasing variation ofthe CoF with increasing temperature of the pin-on-disk tests. Thisbehavior is strongly related to the progressing degree of damageduring the wear tests: The nitride and oxynitride coatings exhib-ited more extensive damage (higher wear rate) at 600 1C and800 1C compared the test at room temperature. As observed also

by other researchers, increase of the surface roughness due tooxidation of the coating can lead to increase of variation of the CoF[7,12,26,30].

The wear rate of the nitride and oxynitride coatings wasexpected to increase even despite the possible formation of anoxide passivation layer [15,27,34,35]. However, an opposite trend,i.e. apparent decrease of wear rate (better wear resistance), wasobserved mainly on the AlTiN coating at high temperature. This isillustrated in Fig. 6 which shows typical profiles of the wear trackson the AlTiN sample after test at room temperature, 600 1C and

Table 3Wear resistance and the main wear mechanisms of the nitride, oxynitride and oxide coatings at 24 1C, 600 1C and 800 1C.

Wear resistance and wear mechanisms

24 1C 600 1C 800 1C

AlTiN Fair Good BadShallow abrasion Oxidationþabrasion Severe oxidation of coating & substrate

AlCrN Good Good FairShallow abrasion Oxidationþbuild-up Oxidationþbuild-up

AlCrON Excellent Excellent GoodVery shallow abrasion Shallow abrasion Oxidationþbuild-up

α-(Al,Cr)2O3 Excellent Excellent ExcellentVery shallow abrasion Very shallow abrasion Shallow abrasion

Fig. 4. SEM images (secondary electron contrast) of the wear tracks after room temperature experiments on the AlTiN (a) and α-(Al,Cr)2O3 sample (b). The insets showtypical surface profiles across the wear track.

Fig. 5. SEM image (secondary electron contrast) of the wear tracks after 800 1C temperature experiments on the AlTiN (a) and α-(Al,Cr)2O3 sample (b). The insets showtypical surface profiles of the wear track.

Table 4Wear of the alumina ball.

Ball wear rate [m3/N/m�10–17]

24 1C 600 1C 800 1C

AlTiN 54.2 108.7 263.5AlCrN 3.6 27.3 48.9AlCrON 0.1 25.8 12.6α-(Al,Cr)2O3 0.5 0.0 2.7

J. Nohava et al. / Tribology International 81 (2015) 231–239 235

Fig. 6. SEM images, EDX maps and wear profile on the AlTiN coating at room temperature, 600 1C and 800 1C. Note the increasing build-up in the track associated withtemperature increase.

J. Nohava et al. / Tribology International 81 (2015) 231–239236

800 1C. The material removal within the wear track graduallydecreased with increasing temperature, accompanied by increas-ing build-up. Since the wear rate calculation was based on theprofile of the wear track, the apparent wear rate was alsodecreasing with increasing temperature. To understand this‘anomalous’ behavior, EDX analysis of the 800 1C wear track inthe AlTiN coating was performed. The results of the EDX analysis(Figs. 6 and 7) revealed high content of oxygen not only in thewear track but also in the initially oxygen-free coating. Asfurthermore evident from Fig. 6, in the wear track center nocoating material could any longer be found and only tungsten andcobalt oxides were detected. From this observation, it can beconcluded that during the tribotest at the highest temperature, thecoating had been worn through in the center of the track and theWC/Co substrate was subsequently oxidized, which resulted in theobserved volume increase. Another part of the build-up observedon the AlTiN coating not only at 800 1C but also at 600 1C was dueto re-deposition of oxidized, worn particles from the substrateoriginating from local oxidized spots. Damaged and rough surfaceof the AlTiN coating, as well as the presence of wear particles, alsoresulted in higher wear of the alumina ball counterpart (seeTable 4) at higher temperatures.

4.2. Analysis of the wear track using focused ion beam tomographycombined with SEM

To further understand the high temperature oxidation and toexplain the decrease of the wear rate, a focused ion beam (FIB) cutwas prepared in the AlTiN wear track after the 800 1C test. TheEDX analysis in the FIB cut (Fig. 7) showed a high content ofoxygen in the center of the wear track (region with the mostsevere damage) and under the coating (peripheral part of the weartrack) as well as at the coating surface. Furthermore, presence oftungsten and cobalt (WC–Co substrate) was also detected in thecentral region of the wear track (Figs. 6 and 7) and none of thecoating elements Ti, Al and N could be detected anymore in thislocation. These results indicate that the coating was completelyremoved in the wear track resulting in a catastrophic failure andrapid oxidation of the exposed WC–Co substrate to such extentthat it filled completely the hollow area created during the weartests. Filling the removed material space with oxidized WC–Cosubstrate was reflected in a low apparent wear rate at hightemperatures; growth of the oxides above the surface lead toincreased build-up rate. Thus, for proper evaluation of tribologicalresults not only wear rate but also build-up rate has to be taken

into account and further analyses have to be conducted in order tounderstand properly the wear processes.

Similar FIB cuts with subsequent cross-sectional EDX analysiswere performed on the AlCrN and AlCrON coatings after the 800 1Ctests. As shown in Fig. 8 for the case of AlCrON, the top layer iseventually worn through and as soon as the Ti-based sublayers areexposed, oxidation becomes dominant and especially in the AlTiNbase layer, a distinct, granular oxide layer is formed as evident fromthe cross-sectional FIB cut of position 1. There is oxygen present atthe coating surface too, but the already high oxygen content in theoxynitride coating covers any tribooxidation phenomena. OxidizedWC–Co substrate was found in some regions on the wear track,however with less extent than on the AlTiN coating. Oxidation ofthe substrate in the wear track was less pronounced for the AlCrONcoating than for the AlCrN coating due to a better high-temperatureoxidation resistance of the oxynitride coating. The presence ofoxygen in the AlCrON coating lead to improvement of the wearresistance of this coating, which resulted in lower wear and build-up rates compared to AlCrN coating (see Fig. 3). Interestingly, itseems from the FIB cut into the AlCrO1�xNx wear track that that theoxidation proceeds much more mildly in the AlCr-based top layerthan it does in the TiAl-based underlying coatings, especially theAlTiN. However, the partial incorporation of oxygen into the nitridecoating was clearly not sufficient to create coating material capableto withstand tribological tests at 800 1C without severe damage, asevident from the fact that even the AlCrO1�xNx top layer is wornthrough in the center of the wear track (Fig. 8).

4.3. Oxide coating

The coefficient of friction of the novel α-(Al,Cr)2O3 oxidecoating was very stable at all temperatures and it even decreasedfrom mE0.5 at room temperature to about mE0.3 at 800 1C. Thisindicates very low damage of the coating at temperatures up to800 1C where other nitride and oxynitride coatings failed mainlydue high temperature oxidation and abrasion. Although somewear was observed at 800 1C, the wear track remained veryshallow and its morphology was very similar to the one obtainedfrom the room temperature tests on the same coating (compareFigs. 3 and 4). SEM observation of the FIB cut through the weartrack revealed only mild abrasive wear without any spallation ordelamination of the coating, thereby confirming an excellentprotection of the substrate even at 800 1C. This means that themain wear mechanism, high-temperature oxidation, was entirelyeliminated for this coating. In this context, it is worth noting thatthe hardness of this coating, even if found to be lower at room

Fig. 7. FIB cut into the AlTiN wear track (left and middle) and EDX elemental maps (right) after the 800 1C test. Note the extensive oxidation in the center of the wear track(coating completely removed) and under the coating. Brighter areas correspond to higher content of the given element.

J. Nohava et al. / Tribology International 81 (2015) 231–239 237

temperature compared to the other tested coatings (Table 1), maybe favorable at elevated temperatures. The magnitude of this effectwould need to be confirmed by high-temperature micromechani-cal testing. Furthermore, the underlying coating base layers andexcept for local pinhole-type defects, even the WC/Co substratehad efficiently been protected from high temperature oxidation bythe oxide coating. The EDX analysis in the FIB cut in the wear track(see Fig. 9) correspondingly showed only negligible substrateoxidation. The substrate remained fully intact, without any signsof cracks or other type of damage. Unlike the AlCrOxN1�x coating,where high temperature oxidation by further replacement of N byO is still possible, the alpha oxide coating remained inert; inaddition, it showed an excellent resistance to abrasive wear in theentire investigated temperature range, including the highest testtemperature (800 1C).

5. Conclusions

This work presents a study of the tribological behavior ofnitride, oxynitride and oxide wear-resistant coatings designedfor use at high temperatures. Cubic AlTiN, AlCrN and AlCrONnitride and oxynitride coatings deposited by a industrial rotatingarc cathodes PVD processes showed good wear resistance at roomtemperature and reasonable wear resistance at 600 1C. The repla-cement of titanium by chromium was found to be beneficial forthe wear resistance in a temperature range up to 600 1C. However,at 800 1C all nitride and oxynitride coatings exhibited severe wear.Only the new α-(Al,Cr)2O3 coating, deposited by rotating cathodesarc PVD in a nitrogen/oxygen atmosphere at 550 1C, maintained anexcellent wear resistance even at 800 1C. The superior wearresistance of this oxide coating was due to the nitrogen-free,

Fig. 8. FIB cut into the AlCrOxN1�x oxynitride wear track after the 800 1C test, showing different wear regions from the inside (1) to the outside (4). The EDX maps (right)show the wear through the different coating zones in positions 1 to 2. Brighter areas correspond to higher content of the given element.

Fig. 9. FIB cut into the α-(Al,Cr)2O3 oxide wear track (left) and EDX elemental maps (right) after the 800 1C test. Brighter areas correspond to higher content of the givenelement.

J. Nohava et al. / Tribology International 81 (2015) 231–239238

purely oxidic composition and its stable alpha-alumina structure,which hindered high temperature oxidation and subsequentsevere wear. EDX and SEM analyses of FIB cuts into the weartracks showed that even if coating adhesive failure could be ruledout, the wear mechanisms especially at high temperatures are stillvery complex and characterization of wear resistance cannotmerely be based on the wear track profile and calculation of thevolume of material removed. Especially during high-temperaturetribological tests, a more complex evaluation needs to be applied,using complementary analytical methods in order to understandthe oxidative and diffusive wear mechanisms of coating andsubstrate. Using the FIB-SEM method, we could confirm thedecisive role of coating oxidation especially if combined withabrasion of the formed oxide, but also of the cemented carbidesubstrate oxidation during wear of PVD coatings at up to 800 1C.

Acknowledgements

The authors wish to acknowledge the help of Gerhard Bürki,Swiss Federal Laboratories for Materials Science and Technology,Thun, Switzerland, during the preparation of the FIB cut samples.

References

[1] Hogmark S, Jacobson S, Larsson M. Design and evaluation of tribologicalcoatings. Wear 2000;246:20–33.

[2] Panich N, Wangyao P, Hannongbua S, Sricharoenchai P, Sun Y. Tribologicalstudy of nano-multilayered ultra-hard coatings based on TiB2. Rev Adv MaterSci 2006;13:117–24.

[3] Hovsepian PEh Lewis DB, Luo Q, Münz W-D, Mayrhofer PH, Mitterer C, Zhou Z,et al. TiAlN based nanoscale multilayer coatings designed to adapt theirtribological properties at elevated temperatures. Thin Solid Films2005;485:160–8.

[4] Najafi H, Karimi A, Dessarzin P, Morstein M. Correlation between anionicsubstitution and structural properties in AlCr(OxN1�x) coatings deposited bylateral rotating cathode arc PVD. Thin Solid Films 2011;520:1597–602.

[5] Polcar T, Cavaleiro A. High-temperature tribological properties of CrAlN,CrAlSiN and AlCrSiN coatings. Surf Coat Technol 2011;206:1244–51.

[6] Woydt M, Habig K-H. High temperature tribology of ceramics. Tribol Int1989;22:75–88.

[7] Holmberg K, Matthews A, Ronkainen H. Coatings tribology—contact mechan-isms and surface design. Tribol Int 1998;31:107–20.

[8] Wang DY, Chang CL, Wong KW, Li YW, Ho WY. Improvement of the interfacialintegrity of (Ti,Al)N hard coatings deposited on high speed cutting tools. SurfCoat Technol 1999;120–121:388–94.

[9] Ohnuma H, Nihira N, Mitsuo A, Toyoda K, Kubota K, Aizawa T. Effect ofaluminum concentration on friction and wear properties of titanium nitridefilms. Surf Coat Technol 2004;177–178:623–6.

[10] Schramm B C, Scheerer H, Hoche H, Broszeit E, Abele E, Berger C. Tribologicalproperties and dry machining characteristics of PVD-coated carbide inserts.Surf Coat Technol 2004;188–189:623–9.

[11] Hasegawa H, Yamamoto T, Suzuki T, Yamamoto K. The effects of depositiontemperature and post-annealing on the crystal structure and mechanical propertyof TiCrAlN films with high Al content. Surf Coat Technol 2006;200:2864–9.

[12] Reiter AE, Mitterer C, Rebelo de Figueiredo M, Franz R. Abrasive and adhesivewear behavior of arc-evaporated Al1�xCrxN hard coatings. Tribol Lett2010;37:605–11.

[13] Birola Y, Islerb D. Abrasive wear performance of AlCrN-coated hot work tool steelat elevated temperatures under three-body regime. Wear 2011;270:281–6.

[14] Ding XZ, Zeng XT. Structural, mechanical and tribological properties of CrAlNcoatings deposited by reactive unbalanced magnetron sputtering. Surf CoatTechnol 2005;200:1372–6.

[15] Brizuela M, Garcia-Luis A, Braceras I, Onate JI, Sanchez-Lopez JC, Martinez-Martinez D, et al. Magnetron sputtering of Cr(Al)N coatings: mechanical andtribological study. Surf Coat Technol 2005;200:192–7.

[16] Jilek M, Cselle T, Holubar P, Morstein M, Veprek-Heijman MGJ, Veprek S.Development of novel coating technology by vacuum arc with rotatingcathodes for industrial production of nc-(Al1�xTix)N/a-Si3N4 superhard nano-composite coatings for dry, hard machining. Plasma Chem Plasma Process2004;24:493–510.

[17] Parlinska-Wojtan M, Karimi A, Coddet O, Cselle T, Morstein M. Characteriza-tion of thermally treated TiAlSiN coatings by TEM and nanoindentation. SurfCoat Technol 2004;188–189:344–50.

[18] Yau BS, Huang JL, Lii DF, Sajgalik P. Investigation of nanocrystal-(Ti,Al)Nx/amorphous-SiNy composite films by co-deposition process. Surf Coat Technol2004;177–178:209–14.

[19] Karvankova P, Karimi A, Coddet O, Cselle T, Morstein M. Thermal stability ofCr1�xAlxSiyN coatings with medium to high aluminium content prepared byarc evaporation. In: Bull SJ, Chalker PR, Chen SC, Meng WJ, Maboudian R,editors. Materials research society symposium proceedings, vol. 890; 2006.p. 85–90.

[20] Chen L, Yang B, Xu Y, Pei F, Zhou L, Du Y. Improved thermal stability andoxidation resistance of Al–Ti–N coating by Si addition. Thin Solid Films2014;556:369–75.

[21] Pilkington A, Dowey SJ, Toton JT, Doyle ED. Machining with AlCr-oxinitridecoated cutting tools. Tribol Int 2013;65:303–13.

[22] Khatibi A, Genvad A, Göthelid E, Jensen J, Eklund P, Hultman L. Structural andmechanical properties of corundum and cubic (AlxCr1�x)2þyO3�y coatingsgrown by reactive cathodic arc evaporation in as-deposited and annealedstates. Acta Mater 2013;61:4811–22.

[23] Edlmayr V, Pohler M, Letofsky-Papst I, Mitterer C. Microstructure and thermalstability of corundum-type (Al0.5Cr0.5)2O3 solid solution coatings grown bycathodic arc evaporation. Thin Solid Films 2013;534:373–9.

[24] Najafi H, Karimi A, Dessarzin P, Morstein M. Formation of cubic structured(Al1�xCrx)2þdO3 and its dynamic transition to corundum phase duringcathodic arc evaporation. Surf Coat Technol 2013;214:46–52.

[25] Roberts WH. Some current trends in tribology in the UK and Europe. Tribol Int1986;19:295–311.

[26] Semenov AP. Tribology at high temperatures. Tribol Int 1995;28:45–50.[27] Polcar T, Martinez R, Vítů T, Kopecký L, Rodriguez R, Cavaleiro A. High

temperature tribology of CrN and multilayered Cr/CrN coatings. Surf CoatTechnol 2009;203:3254–9.

[28] Oliver WC, Pharr GM. An improved technique for determining hardness andelastic modulus using load and displacement sensing indentation experi-ments. J Mater Res 1992;7:1564–83.

[29] Kutschej K, Mitterer C, Mulligan C, Gall D. High-temperature tribologicalbehavior of CrN–Ag self-lubricating coatings. Adv Eng Mater 2006;8:1125–9.

[30] Wagner J, Hochauer D, Mitterer C, Penoy M, Michotte C, Wallgram W, et al.The influence of boron content on the tribological performance of Ti–N–Bcoatings prepared by thermal CVD. Surf Coat Technol 2006;201:4247–52.

[31] Carrapichano JM, Gomes JR, Silva RF. Tribological behaviour of Si3N4–BNceramic materials for dry sliding applications. Wear 2002;253:1070–6.

[32] Themelin L, Desmaison-Brut M, Billy M. Oxidation behavior of a hot iso-statically pressed silicon nitride material. J Phys 1993;IV(3):881–8.

[33] Bolelli G, Cannillo V, Lusvarghi L, Manfredini T. Wear behavior of thermallysprayed ceramic oxide coatings. Wear 2006;261:1298–315.

[34] Philippon D, Godinho V, Nagy PM, Deplancke-Ogletree MP, Fernandez A.Endurance of TiAlSiN coatings: effect of Si and bias on wear and adhesion.Wear 2011;270:541–9.

[35] Chim YC, Ding XZ, Zeng XT, Zhang S. Oxidation resistance of TiN, CrN, TiAlNand CrAlN coatings deposited by lateral rotating cathode arc. Thin Solid Films2009;517:4845–9.

J. Nohava et al. / Tribology International 81 (2015) 231–239 239