Deposition and characterization of (Nb,Cr)N thin films by unbalanced magnetron sputtering

Surface and Coatings Technology 167(2003) 154–160

0257-8972/03/$ - see front matter� 2002 Elsevier Science B.V. All rights reserved.doi:10.1016/S0257-8972(02)00907-6

Deposition and characterization of(Nb,Cr)N thin films by unbalancedmagnetron sputtering

J.N. Tan , J.H. Hsieh *a b,

Institute of Materials Research and Engineering, 3 Research Link, Singapore 117602, Singaporea

School of Mechanical and Production Engineering, Nanyang Technological University, Singapore 639798, Singaporeb

Abstract

Several(Nb,Cr)N thin films were deposited by an unbalanced magnetron sputtering system. The coating properties as afunction of NbyCr ratio were studied using glow discharge optical spectrometer(GDOS), SEM, XRD, AFM and nano-indentationmeasurement. The tribological properties were then carried out using a ball-on-disk setup with alumina balls. The wear resultsshow that coatings which are Cr-rich yield much lower values of friction coefficient and low wear rates than coatings which areNb-rich in compositions for the same tested distance and speed. The effects of NbyCr ratios on the coatings surface texture,structure morphology, hardness, and crystal structure are also discussed.� 2002 Elsevier Science B.V. All rights reserved.

Keywords: Glow discharge optical spectrometer; Unbalanced magnetron sputtering; Tribological properties

1. Introduction

Tool steels have long been used to manufacturemoulding tools and components due to their excellentstrength, toughness, and heat resistance. However,though the excellent mechanical properties, the toolsurface and near-surface regions are subjected to themost destructive forces for example wear, high stresses,elevated temperatures, and corrosive environments. Inorder to address such problems, an attempt was madeto improve the performance of the tool surfaces withthin film coatings in this study.

In the field of hard coatings, chromium nitride hasbeen a popular coating for industrial applications due toits excellent mechanical properties. It was found thatchromium nitride coatings on injection mouldingmachines not only have reduced the amount of wear ofthe moulds, moulding performance was also enhanced.However, physical vapour sputtering produced chromi-um nitrides are found to contain defects such as poresand pinholesw1x. In order to improve the protectivity ofthe surface coatings, a multi-component coatingw2x wasconsidered. Cr-based ternary nitride coatings such as

*Corresponding author. Tel.:q65-790-4330; fax:q65-791-1859.E-mail addresses: [email protected]

(J.H. Hsieh), [email protected](J.N. Tan).

(Ti,Cr)N w3–5x, (Cr,Al)N w6x have been quite widelystudied recently, however, few studies on(Nb,Cr)Nhave been reported in literature. Niobium alloys havebeen known for their high hardness, moreover, it pro-vides excellent corrosion protection. Thus, in this paper,a study was conducted on(Nb,Cr)N multi-componentcoating deposited on tool steel by physical vapourdeposition(PVD) using unbalance magnetron sputter-ing. Several coatings were deposited in this study andeach sample consists of a distinctive combination of Nb,Cr and N according to the target currents set during thedeposition.

The various combination of coatings was subsequent-ly characterized using analytical instruments such as theGDOS, Nanoindenter, Surface profiler, AFM, SEM,XRD, and the optical microscope. Tribological studywas performed using a pin-on-disc tribo-tester.

2. Experimental

2.1. Coating technique

The substrates used for the deposition were M2 steeldisks with diameter of 50 mm and thickness of 6 mm.The average roughness for the substrate is approximately0.0171mm. The sputtering system(Teer 550) w7x wasfirst pumped down to approximately 6.65=10 Pa andy4

155J.N. Tan, J.H. Hsieh / Surface and Coatings Technology 167 (2003) 154–160

Table 1

Sample Target current(A)

OES % Cr:Nb:N Hardness(GPa) Friction Wear rate

Cr Nb

coefficient (=10# mm ym)5 3

1 2.5 2.0 Cr 70 5:1:4 22 0.4 0.722 2.5 3.0 Cr 70 4:1.8:4.2 21 0.52 1.273 2.5 4.0 Cr 70 3.7:2.8:3.5 19 0.81 3.084 2.5 6.5 Nb 55 0.5:4:5.5 26 0.8–0.9 2.685 2.0 6.5 Nb 55 2:3:5 24 0.9 3.86 1.5 6.5 Nb 55 1:3.5:5 26 0.9–1.1 4.547 0.0 6.5 Nb 55 0:1:1 28 1 2.16

Fig. 1. AFM micrograph of Sample 1, which had aR of 6.044 nm.a

argon gas was then introduced to fill the chamber up to0.44 Pa. It was followed by argon plasma cleaning ofthe substrate for approximately 30 min using an r.f.power of 250 W. Two pairs of targets, namely chromiumand niobium, were placed vertically opposite to eachother.

In the deposition process, the M2 steel substrate wasbiased with an r.f. power 80 W(y40 DC) to induceproper ion bombardment, hence achieving the desirablestructure, grain size and film density of the multi-layeredNbNyNbCrN. In order to get a high uniformity of thedeposited NbNyNbCrN film, the substrate holder wasmade to rotate at a constant speed about the axis(centre)where the substrate always faced the targets. Afterdepositing a thin layer of Nb on the substrate, pure

nitrogen gas was injected into the chamber to react withthe Nb to form NbN, followed by gradient layer andNbCrN layer.

The reactive deposition process was controlled by theoptical emission spectroscopy. Coatings with variouscomposition were developed by controlling the currentratio of the Nb and Cr targets. During the depositionprocess, the target current for the respective Nb and Crsources was varied as indicated in Table 1.

2.2. Coating characterization

The coating composition is determined by the glowdischarge optical spectroscopy(GDOS, LECO Spectru-

156 J.N. Tan, J.H. Hsieh / Surface and Coatings Technology 167 (2003) 154–160

Fig. 2. AFM micrograph of Sample 4, which had aR of 11.902 nm.a

Fig. 3. Fractograph of Sample 1. It can be seen that the coating showsa columnar structure.

mat GDS 750). The results are presented in Table 1.The coating hardness was measured using the UltraMicro Indentation System(UMIS 2000) load-controlledsubmicron indentation instrument. The load dwellingtime is 30 s. The resultant hardness results were pre-sented in Table 1.

As for the examination of the surface morphology,the AFM (Nanoscope Dimension 3100 SPM) was�

used. In the experiment, three different spots werechosen for each coating surface and a scan area of 5=5mm was measured. From the scanned area, surface2

topographies and surface roughness of the various sam-ples were determined and recorded. Next, for the exam-ination of the wear tracks, the optical microscopy wasused. Subsequently, a small piece of the coating wasfractured and examined using the SEM(JEOL 6700F)to see the fractographs of the coatings cross-sections.An XRD system(Bruker Axs) with a copper sourceemitting Ka X-ray in the line attachment was used toinvestigate the crystal structure of the multi-componentcoatings. In the theta-2 theta scan mode, the coatedsurfaces and the substrate surface were scanned from 20to 908. The scan rate used is 18 per minute.

2.3. Wear test

In the wear testing, a pin-on-disc apparatus was usedto evaluate wear. This apparatus consists of a pin with

an alumina ball at the end point. This pin is placedperpendicularly to the coated substrates. The disc wasmade to rotate beneath the alumina ball at a speed of10 cmys under a load of 2.5 N. The sliding distancewas 1000 m. After the test, the friction coefficient wasdetermined and the wear tracks of each coatings weremeasured using a surface profiler(KLA Tencor P-10).The wear loss of the sample was evaluated by measuring


Fig. 4. Fractograph of Sample 5. The coating here displays a finefibrous structure.

Fig. 5. XRD spectra of the various composition of NbCrN coatings, NbN coatings, CrN coatings and M2 steel substrate.

the cross-sectional area of the wear track, A, and theworn volume is calculated with the following formula:

Vs2prA (1)

where r is the radius of the wear track. The wearcoefficient is later found by subtracting the worn volumeby the total wear distance.

3. Results and discussion

From the results of the GDOS, Table 1 shows theresults of the compositions of the seven samples. Thesamples are named as Samples 1–7 according to theirrespective target current ratios of Cr to Nb.

3.1. Surface coating morphology

From Figs. 1 and 2 of the AFM micrographs of thecorresponding Sample 1(Cr-rich) and Sample 4(Nb-rich), it is observed that the Cr-rich Sample 1 possesseslower roughness than the Nb-rich Sample 4. In fact,micrographs(which are not shown) of the samples alsoshow that the Cr-rich samples are smoother than theNb-rich samples.

3.2. Coating microstructure

For the microstructure of the samples, Figs. 3 and 4exemplify that both the Cr-rich sample and Nb-richsample have columnar crystal morphology. In this case,there is no significant difference among the sevensamples. This is most likely due to the use of the samesputtering process parameters such as the same sputter-ing temperature, the same voltage bias, and the samepressure throughout the seven samples. However, it wasnoticed that the Cr-rich samples are rather columnarwhile Nb-rich samples like Samples 5 and 7 have veryfine fibrous features and the most packed structures.


Fig. 6. Comparison of the wear rates of the various coatings.

From the XRD results(Fig. 5), we see that all thesamples are crystalline in structure. XRD patternsrevealed that the ternary coatings, Sample 4, 5, and 6showed great similarity to Sample 7(NbN) in terms ofstructure. This is shown in Fig. 5 by the resemblance ofthe shape of the three samples with the NbN diffractionpattern. As for Samples 1 and 2, they only showedsingle phase of CrN(2 0 0) while Sample 3 showed asingle phased NbCrN(2 0 0).

The XRD pattern also showed that the(2 0 0) peaksin Samples 1, 2 and 3 are all broadened and the intensityis not that high as compared with the other samples.This probably could be caused by the finer grain sizeof these three samples than that of other samples.

3.3. Tribological study

From Table 1, we can see that the hardest film isSample 7, NbN and the least hard film is Sample 3. In

fact, the samples with the higher Nb content, Samples4, 5, 6 and 7 are those with the higher hardness whereasthe samples with the higher Cr content, Samples 1, 2and 3 are those with the lower hardness.

Secondly, we see from Table 1, the results of the pin-on-disc test indicate that Sample 1 and 2 had the lowestcoefficient of friction. These results were hence corre-lated to the wear rates shown in Fig. 6, which alsoindicated that Sample 1 and 2 have the lowest wearrates. As understood from literatures, one of the pro-cesses responsible for wear of materials is friction suchthat a higher level of friction causes a greater amountof wear. Thus, for the samples with the higher Nbcontent, that is, Samples 4–7, the friction coefficientsare higher and hence the wear rates. From AFM study,it was also noted that the Cr-rich samples like Samples1, 2 and 3 which have a lower average roughnesscorrespond to lower wear rates as shown in Fig. 6. Theroughness reaches its minimum at 6.044 nm when the


Fig. 7. Optical micrograph of the wear track of Sample 2, at a mag-nification of 100=.

Fig. 8. Optical micrograph of the wear track of Sample 6, at a mag-nification of 100=.

ratio of CryNb is 5:1, where the chromium content isthe highest out of the seven samples.

According to the general wear equation developed byHolm w8x and the later work of Archardw9x. The wearvolume,V is directly proportional to the sliding distance,d and the applied normal force,F and inverselyn

proportional to the hardness or yield stress,H of thesofter surface as follows:

FnVsk (d) (2)H

Thus the wear volume of the coatings should be smallwith greater hardness. However, results shown in Fig. 6indicate that Sample 7 has a moderately high wear rate.

The trend is also noted for the other Nb-rich samples,being Sample 4, 5 and 6. This mismatch between thehardness and the wear rate could be due to the wearmechanism of ceramic materials as proposed by Ajaviet al. w10x. According to them, brittle fracture is theprime cause of wear in some ceramic materials. Fracturetoughness hence plays an important role in the wear ofthese ceramics. In another study by Evans et al.w11x,they have stated that the wear volume is inverselyproportional to square root of fracture toughness. Hence,this could imply that the Nb-rich coatings are morebrittle (lower fracture toughness) than the Cr-rich coat-ings which may have higher fracture toughness. Thisimplication is consistent with the results of AFM andXRD studies. In both studies, it shows that the Cr-richcoatings may have finer structure, and therefore highertoughness. In addition, it is also suspected that the brittleNb N phase found in Nb-rich coatings may have caused2

the loss of toughness in these coatings.

3.4. Coatings wear mechanism

In Figs. 7 and 8, we saw the optical micrographs ofthe wear tracks of Samples 2 and 6, respectively,whereby Sample 2 is rich in chromium while Sample 6is rich in niobium. As the pictures were of the samescale and magnification, the significant differences inwear behavior between these coatings is worth noting.On the wear tracks formed after the sliding tests,polishing and ploughing characteristics are found for theCr-rich Sample 2 and the Nb-rich Sample 6, respectively.

4. Conclusion

Seven PVD coatings were studied in this study. Itwas found that the ternary coating’s roughness, crystalstructure, and tribological properties depended stronglyon the Cr content. The coatings which are Cr-rich,generally have lower average roughness lower wearrates and their corresponding friction coefficients arealso lower.

Conversely, coatings with higher ratio of Nb in theircompositions have high average roughness, showed highwear rates, and their corresponding friction coefficientswere also high. These coatings have crystal structureswhich followed that of NbN. Moreover, the presence ofthe Nb N phase in the Nb-rich coatings could have2

caused them to have a lower toughness.Thus, maintaining a low roughness and high Cr

content of the coating has a significant meaning whenprotective coatings are applied onto moulding tools. Itcan help improving surface finishing and non-stickingbehaviors.


Acknowledgments

We would like to express sincere gratitude to Dr ZengXianting of the Singapore Institute of ManufacturingTechnology for their help during this study.

References

w1x M. Ohring, Materials Science of Thin films, Deposition andStructure, second ed., Academic Press, 2002, pp. 170–183.

w2x D.P. Monaghan, D.G. Teer, K.C. Laing, Efeoglu, R.D. Arnell,Surf. Coat. Technol. 59(1993) 21.

w3x P. Hones, R. Sanjines, F. Levy, Thin Solid Films 332(1998)240.

w4x J.J. Nainaparamil, J.S. Zabinski, A. Korenyi-Both, Thin SolidFilms 333(1998) 88.

w5x J. Vetter, H.J. Scholl, O. Knotek, Surf. Coat. Technol. 74y75(1995) 286.

w6x M. Pakala, R.Y. Lin, Surf. Coat. Technol. 81(1996) 233.w7x P.J. Kelly, R.D. Arnell, Vacuum 56(2000) 159.w8x R. Holm, Electrical Contacts, Stockholm: Almquist & Wiksells,

(1946) Sec 49.w9x J.F. Archard, Proc. Interdisciplinary Approach to Friction and

Wear, Sp 181, P.N. Ku,(Ed.) Texas: San Antonio,(1967)267–333.

w10x O.O. Ajavi, K.C. Ludema, Proc. International Conference onWear of Materials, 5–9 April 1989, Houston, TX: AmericanSociety of Mechanical Engineers, 349–362.

w11x A.G. Evans, D.B. Marshall, in: D.A. Rigney(Ed.), Fundamen-tals of Friction and Wear of Materials, American Society forMetals, 1981, p. 439.

Deposition and characterization of (Nb,Cr)N thin films by unbalanced magnetron sputtering

Documents