Influence of bias voltage on the tribological properties of titanium nitride films fabricated by dynamic plasma ion implantationydeposition

Transcription

1 Surface and Coatings Technology 161 (2002) Influence of bias voltage on the tribological properties of titanium nitride films fabricated by dynamic plasma ion implantationydeposition X.B. Tian, T. Zhang, R.K.Y. Fu, P.K. Chu* Department of Physics and Materials Science, City University of Hong Kong, Kowloon, Hong Kong, PR China Received 22 April 2002; accepted in revised form 2 July 2002 Abstract Dynamic plasma ion implantationydeposition (PIID) combining gas and metal plasmas has been proven to be an effective technique to fabricate titanium nitride films. Pulsed vacuum arc and simultaneous substrate biasing can provide the capability to optimize film properties through flexible matching of processing parameters. In this work, titanium nitride films were synthesized on AISI304 stainless steel samples using filtered titanium cathodic arc and hot filament glow discharge. The effects of the substrate bias (8, 16 and 23 kv) on the tribological properties were investigated. The bias voltage has a slight influence on the film thickness and friction coefficient. The pin-on-disk experimental results demonstrate that a higher bias voltage (e.g. 23 kv) leads to better tribological properties compared to a lower bias. Under our testing conditions, the slide time to result in rapid friction increase for the 23 kv sample is 1.25 times that of the 8 kv sample. The wear tracks on the 23 kv sample are more irregular than those on the untreated or 8 kv sample. The improvement is believed to be related to the higher adhesion induced by ion mixing and formation of the surface layer incorporating titanium, nitrogen, and oxygen Elsevier Science B.V. All rights reserved. Keywords: Ion implantation; Plasma processing and deposition; Tribology 1. Introduction Synthesis of a hardened surface layer with good adhesion properties is of practical importance in tribological applications. Ion beam assisted deposition (IBAD) provides such capabilities w1,2x and is an alternative method to conventional plasma nitriding w3x. The former is a low temperature technique although sample temperature can sometimes be raised in a controlled fashion to optimize the surface structures and properties. In contrast, plasma nitriding is typically performed at elevated temperatures, and the conditions may not be suitable for some metals with a low tempering temperature. In addition, plasma nitriding does not easily yield a big improvement for metals that do not readily form nitrides such as aluminum. Presented at ICMCTF *Corresponding author. Tel.: q ; Fax: q address: paul.chu@cityu.edu.hk (P.K. Chu). Compared to conventional IBAD, plasma-based IBAD is more interesting due to its capability to treat samples with an irregular shape. Combining metal arc plasma deposition and gas ion implantation, plasma ion implantation deposition (PIID) has been proven to be a versatile surface modification technique w4,5x. By carefully selecting experimental parameters such as the plasma density, pulse width, synchronization of the cathodic arc and sample bias pulses, and others, pure deposition, pure implantation, and IBAD can be accomplished using dual plasmas in the same machine and without breaking vacuum. Hence, it is intrinsically easier to optimize individual processes and materials properties such as internal stress, structure, and surface properties. As an example, TiN thin films with superior tribological properties and adequate corrosion resistance have been synthesized by PIID utilizing dual plasmas w6x. During the PIID experiments, the titanium plasma is generated using a vacuum arc whereas the nitrogen plasma is sustained by hot filament glow discharge. Our previous /02/$- see front matter 2002 Elsevier Science B.V. All rights reserved. PII: S Ž

2 X.B. Tian et al. / Surface and Coatings Technology 161 (2002) Table 1 Film thickness and ion-mixed zone thickness of the synthesized samples Treating voltage Film thickness (nm) Mixing thickness (nm) 8kV kv V investigation on the influence of the sample bias on the surface morphology and corrosion resistance shows that by increasing the bias voltage, the film grain size becomes smaller and the film is less corrosion resistant w6,7x. We believe that it is due to the broad energy distribution of the implanted ions and the ion penetration depths induced by the bias variations. In this work, we focus on the impact of the bias voltage on the tribological properties such as wear resistance. 2. Experimental The films were deposited on AISI 304 stainless steel samples 20 mm in diameter and 3 mm in thickness. The instrumental setup has been described elsewhere w6x. The titanium plasma was generated by pulsing a vacuum arc at a frequency of 200 Hz and pulse duration of 270 ms. The nitrogen plasma was produced by hot filament glow discharge. Three different sample voltages were used in our experiments, 8, 16 and 23 kv, and the pulse width was 30 ms. The high-voltage pulses were synchronized with the arc pulses at approximately 100 ms after the initiation of the vacuum arc. The total treatment time was 30 min. After PIID, the samples were depth profiled using Auger electron spectroscopy (AES) to disclose the elemental distributions. The film thickness was determined using the 390 ev Auger peak depth profile and the ion-mixed zone thickness was calculated based on the Fe signal (15 at.%) and N1.TiqN signal (15 at.%). The tribolgical experiments were conducted on a ball-on-disk tester. The ball was 5 mm in diameter and the material was WC Co 6%. The rotation diameter was 10 mm, speed was 30 rev.ymin, and applied force was 100=g. The test was conducted for 200 s without lubrication. After the tribological tests, the wear track area was assessed using a-step profilometry. phase to implantation phase is approximately 8 1. That is to say, the titanium on the substrate surface is mainly deposited. The 23 kv sample is slightly thicker probably because of more energetic ion bombardment and larger ion mixing. A fairly large amount of oxygen can be found in the deposited films. With increasing sample bias, the relative oxygen content increases slightly, and the trend is consistent with our previous experimental results in which more oxygen was incorporated in the top layer of AISI 304 samples treated by pure nitrogen plasma immersion ion implantation (PIII) w8x. The tribological properties were evaluated using a ball-on-disk tester and the variation of the friction coefficients with time is shown in Fig. 1. It is apparent that the tribological properties are enhanced by the treatment. There exists a critical sliding distance after which the friction coefficient becomes very large. The improvement in the critical distance depends on the sample bias although the friction coefficients measured from all the treated samples are similar in this work. At the initial stage of the measurement, the untreated sample exhibited a lower coefficient compared to the treated samples and it can be attributed to the change of the surface conditions and roughness of the treated samples due to the coating process. The wear track size depends on the sample bias as shown in Fig. 2. The track area decreases with higher sample bias. This may be due to the change in the materials and hardness of the contacting couples. The track edge and track width on the untreated sample are nearly uniform. There is evidence that delamination and tear occur during sliding and the groove is deep. In contrast, the wear tracks on the treated samples are more irregular and the track area is not uniform. As shown in Fig. 3, the 23 kv sample shows the smallest track size and longest sliding distance before demonstrating a rapid friction increase, implying that the 23 kv sample is more wear resistant. 3. Results As shown in Table 1, the variation in the bias only affects the layer thickness slightly. In our study, we used an average sputtering rate to calculate the film thickness. The sputtering rate is known to vary with compositional changes, and so the absolute values shown in Table 1 have some errors, estimated to be 5 10%, but comparison among them is more accurate. The results in Table 1 are not unexpected since the ratio of the deposition Fig. 1. Friction coefficients of untreated and treated samples acquired from ball-on-disk tests.

3 234 X.B. Tian et al. / Surface and Coatings Technology 161 (2002) Discussion We have synthesized TiN films on AISI 304 stainless steel using different bias voltages and found that the sample bias has a slight influence on the film thickness but substantially affects the tribological properties. This may be attributed to different dose and ion bombardment effects under various bias voltages. For the same implantation time, the implant dose varies with the bias voltage with relationship DoseAV 1y2 in planar plasma ion implantation w9x. The dose will be a little higher for a small sample due to spherical convergence of the plasma sheath. However, the plasma sheath expansion is slow and the sheath edge is very thin (near the samples) in our experiments as a result of the high plasma density and drift velocity of the plasma stream. According to the equation in Ref. w10x, Fig. 3. Effects of sample bias on tribological properties. Fig. 2. Surface morphology of wear tracks on the samples treated using different bias voltages: (a) 0 kv-untreated, (b) 8kV,(c) 16 kv, and (d) 23 kv.

4 X.B. Tian et al. / Surface and Coatings Technology 161 (2002) Table 2 Projected ion range and longitudinal straggling (unit: nm, error -"0.2 nm) of different ion species in TiN films Plasmas Ion species Percentage Projected range Longitudinal straggling 8 kv 16 kv 23 kv 8 kv 16 kv 23 kv Metal ion Ti q 11% Ti 2q 75% Ti 3q 14% Gaseous ion N q 20% N q 2 80% B 3 E1y V C 2 2 F 3 DmQen i i0vi0 G ds (1) where Q,m e i and ni0 denote the mean ion charge, ion mass, and plasma number density in the sheath, 0 is the permittivity of vacuum, and vi0 is the velocity of ions entering the sheath from the plasma sheath boundary. The sheath thickness is only approximately 10 mm for a typical titanium plasma w11,12x and the parameters used in this work. That is to say, the ion dose at higher energy is only affected slightly by the bias variation. However, the bias has a substantial influence on ion mixing. During the implantation phase, both titanium ions and nitrogen ions are simultaneously implanted into the surface. The ion range of each ion species is different depending on their mass and charge states (Table 2 w13x) but the net effect is favorable to atomic mixing. Comparing the three samples, the 23 kv sample possesses a wider mixing zone due to a higher implantation effect. This will increase the sticking force between the film and substrate surface, consequently leading to a better wear resistance. The sample surface is usually oxidized during PIIIrelated processes. The oxygen may originate from several sources: (1) metallic plasma arc sources w14x; (2) residual vacuum; (3) surface oxide on the samples, and (4) sputtering from the oxidized surface of the sample holder and chamber wall w15x. The hardness of the TiN films can be directly correlated with the oxygen content for titanium oxynitrides w16x. It has been reported that nitrogen ion implantation into titanium alloys not only produces nitride in the near surface region but also promotes the formation of a smooth oxide film on the surface w17,18x, and the oxide layer is stabilized by nitrogen. From this standpoint, oxygen incorporation into TiN films may not be detrimental although the exact mechanism has not been disclosed. Improvement of the wear resistance using TiN coatings has been widely investigated, and IBAD TiN films have attracted much attention due to the better adhesion between the substrate and film w19 23x. It may be attributed to ion bombardment effects: interface cleaning, thicker intermixed interface w19x, the amorphous structure, fine precipitates w20x, etc. In most previous experiments, the bombarding species are nitrogen ions w19 22x or argon ions in a nitrogen gas environment w23x, as opposed to our use of Ti and nitrogen ions simultaneously in this work. Although the nitrogen energy effect on the TiN films has been investigated, the focus has been on low energies usually less than 1 kev w22,24,25x. For IBAD processes using titanium ion mixing and a higher bombardment energy, the investigation on the energy effect has seldom been performed, although TiN has been synthesized using vacuum arc by several groups w26,27x. In our experiments, the bias variations have a slight influence on friction. In contrast, the friction coefficients of CrN synthesized by energetic ion mixing increase with energy w28,29x. The difference may be attributed to different ion species and plasma conditions. In spite of the small influence on friction, the sample bias in PIID can substantially increase the working lifetime and wear resistance of the films. Based on our experimental data, the critical time before a large increase in friction measured from the 23 kv sample increases by 50% and the wear track decreases by approximately 30% compared to 8 kv sample. However, even though a higher bombarding energy is favorable with regard to hardness improvement in nitrogen ion mixing experiments w28x, the hardness may exhibit a reverse trend if the voltage is too high w29x. 5. Conclusion Plasma ion implantationydeposition employing metal arc and gas plasmas can effectively synthesize wear resistant TiN thin films. The sample bias affects the plasma implantation dynamics, consequently impacting the tribological properties of the fabricated TiN films. The film thickness varies slightly with the sample bias due to the small plasma sheath and small implantation to deposition ratio. The sample bias affects the friction coefficient slightly but a higher bias voltage substantially increases the wear resistance. Acknowledgments The work described in this paper was supported by a grant from the City University of Hong Kong (SRG ).

5 236 X.B. Tian et al. / Surface and Coatings Technology 161 (2002) References w1x C.M. Cotell, J.K. Hirvonen, Surf. Coat. Technol. 81 (1996) 118. w2x W. Ensinger, A. Schroer, G.K. Wolf, Nucl. Instrum. Meth. B 80y81 (1993) 445. w3x T. Czerwiec, N. Renevier, H. Michel, Surf. Coat. Technol. 131 (2000) 267. w4x J.R. Conrad, J.I. Radke, R.A. Dodd, F.J. Warzala, N.C. Tran, J. Appl. Phys. 62 (1987) w5x A. Anders, Handbook of plasma immersion ion implantation and deposition, John Wiley and Sons, Inc, New York, w6x X.B. Tian, R.K.Y. Fu, P.K. Chu, J. Vac. Sci. Technol. A20 (2002) 160. w7x X.B. Tian, L.P. Wang, R.K.Y. Fu, P.K. Chu, Mater. Sci. Eng. A (in press). w8x X.B. Tian, P.K. Chu, J. Vac. Sci. Technol. A 19 (2001) w9x M.A. Lieberman, J. Appl. Phys. 66 (1989) w10x A. Anders, Appl. Phys. Lett. 76 (2000) 28. w11x I.G. Brown, Rev. Sci. Instrum. 65 (1994) w12x G.Y. Yushkov, A. Anders, E.M. Oks, I.G. Brown, J. Appl. Phys. 88 (2000) w13x J.F. Ziegler, J.P. Biersack, U. Littmark, The Stopping Power and Range of Ions in Solids, Pergamon, New York, w14x J.M. Schneider, A. Anders, B. Hjorvarsson, L. Hultman, Appl. Phys. Lett. 76 (2000) w15x P.K. Chu, Surf. Coat. Technol. 156 (2002) 244. w16x A. Koniger, J.W. Gerlach, H. Wengenmair, C. Hammerl, J. Hartmann, B. Rauschenbach, Surf. Coat. Technol. 84 (1996) 439. w17x R. Hutchings, W.C. Oliver, Wear 92 (1983) 143. w18x W.C. Oliver, R. Hutchings, J.B. Pethica, Metal. Trans. A 15 (1984) w19x F.J. Koerber, H. Petersein, H. Ranke, Thin Solid Films 181 (1989) 505. w20x S. Raud, J. Delage, J.P. Villain, P. Moine, Thin Solid Films 181 (1989) 333. w21x E. Vera, G.K. Wolf, Nucl. Instrum. Meth. B 148 (1999) 917. w22x T. Sikola, J. Spousta, et al., Surf. Coat. Technol. 109 (1998) 284. w23x A. Koniger, J.W. Gerlach, H. Wengenmair, C. Hammerl, J. Hartmann, B. Rauschenbach, Surf. Coat. Technol. 84 (1996) 439. w24x M.L. Klingenberg, J.D. Demaree, Surf. Coat. Technol. 146 (2001) 243. w25x R. Hubler, E.K. Tentardini, E. Blando, et al., Surf. Coat. Technol 119 (1999) 969. w26x M. Sano, T. Teramoto, K. Yukimura, T. Maruyama, Surf. Coat. Technol 128y129 (2000) 245. w27x R. Guenzel, N. Shevshenko, W. Matz, A. Mucklich, J.P. Celis, Surf. Coat. Technol 142 (2001) 978. w28x Y.Q. Fu, X.D. Zhu, B. Tang, et al., Wear 217 (1998) 159. w29x K. Sugiyama, H. Hayashi, J. Sasaki, O. Ichiko, Y. Hashiguchi, Nucl. Instrum. Meth. B 80y81 (1993) 1367.