Nitrogen plasma immersion ion implantation for surface treatment and wear protection of austenitic stainless steel X6CrNiTi1810

Nitrogen plasma immersion ion implantation for surface treatment and wear protection of austenitic stainless steel X6CrNiTi1810

Surface and Coatings Technology 116–119 (1999) 352–360 Nitrogen plasma immersion ion implantation for surface treatme...

821KB Sizes 0 Downloads 5 Views

Recommend Documents

No documents
Surface and Coatings Technology 116–119 (1999) 352–360

Nitrogen plasma immersion ion implantation for surface treatment and wear protection of austenitic stainless steel X6CrNiTi1810 C. Blawert *, B.L. Mordike Institut fu¨r Werkstoffkunde und Werkstofftechnik, TU Clausthal, Agricolastr. 6, 38678 Clausthal-Zellerfeld, Germany

Abstract Plasma immersion ion implantation is an effective surface treatment for stainless steels. The influence of treatment parameters (temperature, plasma density and pressure) on the sliding wear resistance are studied here. At moderate temperatures, nitrogen remains in solid solution without forming nitrides. This increases the surface hardness and the wear resistance without affecting the passivation of the steel. This may allow the use of such steels in applications where their poor wear resistance would normally prohibit their use. © 1999 Elsevier Science S.A. All rights reserved. Keywords: Austenitic stainless steels; Expanded austenite; Plasma immersion ion implantation; Wear

1. Introduction Austenitic stainless steels have a relatively low hardness and generally poor resistance to wear. Without surface treatment, their strength and hardness can only be increased by cold working or age-hardening. The increase in surface hardness obtained by these processes is generally much lower than that obtained by surface treatments and has only a minor effect on the wear behaviour. Thermochemical surface treatments, such as conventional nitriding or coatings, have limited application due to the high temperatures used during these processes and/or problems with adhesion to the substrate. Hardening by nitriding is quite effective but the precipitation of CrN becomes a problem above 673 K and the steels lose their good corrosion resistance. Plasma immersion ion implantion (PI 3) [1] can be considered as a suitable surface treatment because it offers nitriding treatment below this critical temperature. It has been shown previously that PI 3 treatment can increase the surface hardness and wear resistance of austenitic stainless steels while maintaining or even improving the corrosion resistance [2,3]. A nitrogenrich phase, variously called ‘expanded austenite’ or ‘S-phase’, has been found in the modified layer [2–5]. * Corresponding author. Tel.: +49-5323-93-8513/72-2120; fax: +49-5323-93-8515/72-3148. E-mail address: [email protected] (C. Blawert)

Neither the requirements for the formation of this phase nor its structure are certain, but the improvements in corrosion and wear behaviour can definitely be imputed to its presence in the modified layer. Expanded austenite appears to be a metastable phase with a supersaturation of nitrogen which remains in solid solution. The nitrogen atoms may form an ordered phase close to Me N (Me= 4 Fe,Cr,Ni). A wide range of process conditions can produce expanded austenite and its formation is not confined to PI 3 treatment. Related processes, such as ion implantation [6–9], plasma nitriding [5,10–14] and reactive sputter deposition [15] can produce this phase if the treatment temperature is around 400°C. A review of processes producing expanded austenite and a discussion of its properties has been provided by Williamson et al. [16 ]. The same authors have examined the wear behaviour of austenitic stainless steel in several studies. They found a large increase in wear resistance for surfaces containing expanded austenite produced by high current nitrogen implantation at temperatures around 400°C [17–19]. In this paper, we discuss the possibility of using PI 3 treatment to modify the surface of stainless steels to obtain improved wear resistance. Although the treatment of ferritic and austenitic–ferritic stainless steels is also possible, we will restrict ourselves to the presentation of results obtained with the austenitic stainless steel X6CrNiTi1810. Structural and mechanical properties were investigated and correlated to the wear behavior.

0257-8972/99/$ – see front matter © 1999 Elsevier Science S.A. All rights reserved. PII: S0 2 5 7- 8 9 7 2 ( 9 9 ) 0 0 08 2 - 1


C. Blawert, B.L. Mordike / Surface and Coatings Technology 116–119 (1999) 352–360 Table 1 Treatment parameters of the samples used in the wear tests Treatment

#1 #2 #3 #4 #5a #6


Temperature (°C )

Pressure (mbar)

Time (h)

Pulse length (ms)

Pulse voltage (kV )

Frequency (Hz)

Dose (at/cm2)

300 400 500 400 400 400

1.5 1.5 1.5 1.7 1.7 4.7

3 3 3 3 3 3

100 100 100 100 100 100

40 40 40 40 40 40

72 149 285 150 150 150

1.8×1018 2.6×1018 5.7×1018 1.4×1018 2.3×1018 2.3×1018

a Additional 400 V bias on the plasma potential control electrode.

A number of different wear tests were performed to demonstrate the positive effect of PI 3 treatment.

2. Experimental details Five millimeter thick samples of X6CrNiTi1810 stainless steel were cut from bars of 25 mm diameter. They were polished to a 1 mm diamond finish and cleaned in an alcohol ultrasonic bath prior to PI 3 treatment which was carried out in the Mark I PI 3 devices at Technical University of Clausthal ( TUC ) and Australian Nuclear Science & Technology Organisation (ANSTO) [20]. The ANSTO device is equipped with a radiantly heated substrate holder and a positively biased electrode which can be used to raise the potential difference between the plasma and the chamber walls and so increase the plasma density [21]. Prior to treatment, the systems were pumped down to a base pressure below 10−6 mbar to minimise oxygen contamination. The chamber was then filled with nitrogen to the intended working pressure in the range from 1.5 to 4.0×10−3 mbar. An r.f. plasma was maintained by 300 W of r.f. power at 13.56 Mhz applied to an immersed inductive antenna. One series of samples was treated with 400 V applied to an electrode in the plasma, which raised the plasma potential and increased the plasma density. The samples were placed on an isolated flat working table surrounded on all sides by the plasma. Implantation was performed by applying negative voltage pulses with amplitude of −40 kV and length of 100 ms to the table. In the TUC system, treatment temperature was controlled by the repetition rate of the pulses. The heat loss was almost entirely due to radiation and, since there was no additional cooling, the treatment parameters of dose and time were not independent of each other. In the ANSTO system, the frequency was fixed at 150 Hz and the temperature was controlled by additional radiation heating. Whilst the time and temperature were the same for treatments in the two systems, nitrogen bombardment was used in the TUC system to bring the specimens up to temperature (15–30 min at

300 Hz depending on the chosen temperature) while radiation heating in vacuum was used in the ANSTO system. As a result, higher doses and thicker layers were achieved in the TUC treatments. The detailed treatment parameters for the specimens in our investigations are displayed in Table 1. From previous work [2], we know that after treatment at 400°C the implanted layer will be supported by a diffusion zone having higher load bearing capacity and therefore best wear resistance without forming chromium nitrides. The surface hardness was measured with a Leitz Durimet microhardness tester using a Vickers indenter at loads of 10, 25, 50, 100, 200 and 300 g and with an instrumented nanoindenter II at loads of 50 mN using a Berkovitch indenter. Structural changes in the modified layer and the layer thickness were investigated using cross-sections for optical and scanning electron microscopy. Glancing angle X-ray diffraction (Siemens D500) with CoKa radiation was used to determine the phases present in the modified layer. Since we are primarily interested in the sliding wear performance of the modified layers, two different model sliding wear tests were carried out. A secondary aim of this work was to determine if the modified scratch test (test 2 below) could be used for reliably assessing the increase in wear resistance obtained by PI 3 treatment. The two different wear tests were: 1. Rotating pin-on-disc wear test with a CSEM tribometer. Two different steel pins (treated and untreated X5CrNiMo1810 and X40Cr13) of 5 mm diameter were used as the counterpart. The load was increased up to 20 N. All tests were performed with ethanol lubrication. To obtain results which could be compared to the unlubricated tests, a low speed of 0.05 m/s was chosen so that metal–metal contact was still possible. 2. Oscillating pin-on-disc wear test with a Teer Coating Scratch Tester operating in the constant load mode of the apparatus. Instead of the diamond tip, 100Cr6 steel balls of 5 mm diameter were used as the counterpart. The sliding speed was 0.05 m/min. Only unlubricated tests were performed. Monitoring of the relative


C. Blawert, B.L. Mordike / Surface and Coatings Technology 116–119 (1999) 352–360

Table 2 Hertzian stresses induced in the wear specimens Disc


Load (N )

Compressive strain (N/mm2)

Tensile strain (N/mm2)

Shear strain (N/mm2)

X6CrNiTi1810 X6CrNiTi1810 X6CrNiTi1810 X6CrNiTi1810 X6CrNiTi1810 X6CrNiTi1810 X6CrNiTi1810 X6CrNiTi1810

X40Cr13 X5CrNiMo1810 X5CrNiMo1810 X5CrNiMo1810 100Cr6 (dry) 100Cr6 (dry) 100Cr6 (dry) 100Cr6 (dry)

10 5 10 20 5 10 15 20

1590 1224 1542 1942 1251 1576 1804 1986

212 163 206 259 167 210 241 265

477 367 463 583 375 473 541 596

Fig. 1. Glancing angle X-ray diffraction spectra (CoKa) for the untreated X6CrNiTi1810 stainless steel and after treatments #4 and #5 at 400°C. The X-ray incident angle was 5°.

humidity showed 15%±5% during all tests. Loads up to 20 N were used. In both of these tests, the friction force was monitored to obtain the coefficient of friction. During the wear tests, the appearence of the tracks was observed visually. The depth of the wear tracks was measured with a profilometer after each test and the tracks and wear scars on the balls were examined with an optical microscope. The Hertzian stresses related to the applied loads and various pin materials are listed in Table 2. Note that the different ball (pin) materials were selected to obtain the major wear on the specimen rather than on the ball.

3. Results and discussion 3.1. Microhardness, structure and layer thickness Typical X-ray diffraction patterns obtained with an incident angle of 5° from austenitic steel treated at

400°C are shown in Fig. 1. To demonstrate the effect of the treatment, the diffraction pattern obtained from the untreated material is displayed in the same graph. The austenitic steel in the untreated state has the typical austenite peaks except that there is an additional small peak which can be attributed to a martensitic/ferritic transformation of the very near surface that occurs during polishing of the sample. This peak disappeared at higher incident angles and was not observed on any of the treated samples. The treated samples revealed a shift of the austenite peaks towards smaller angles, indicating an expansion of the lattice. This shift has been observed previously [1–14] and has been associated with expanded austenite, formed by nitrogen remaining in solid solution in the f.c.c. lattice. Since it remains in solid solution, diffusion of nitrogen into the material can occur forming expanded austenite layers of several micrometer thickness in a few hours. By a variation of the treatment temperature and/or treatment time the layer thickness can be adapted to a wide range of loads. But however the amount of nitrogen

C. Blawert, B.L. Mordike / Surface and Coatings Technology 116–119 (1999) 352–360

is reduced as diffusion increases, except when certain combinations of temperature and supersaturation lead to nitride formation. Such nitride formation is not desirable in stainless steels, although it would increase the wear resistance, as the free chromium content may drop below its critical value (due to CrN formation) so that passivation in corrosive environments is no longer possible. Note that temperature and time are not the only important process parameters because plasma density, gas pressure, implantation voltage and the repetition rate of the HV pulses are also able to influence the nitrogen uptake. More detailed results showing the influence of process parameters on the microstructure of specimens used in the present wear tests are published elsewhere [3,23,24]. Cross-sections of the samples treated at 400°C show a clear separation between the modified surface region and the substrate — Fig. 2. The modified layer appears to be uniform (expanded austenite) and no precipitates are visible. The layer thicknesses in Table 3 were estimated from the cross-sections. With a constant treatment time of three hours, the layer thickness can be increased by performing treatments at higher temperatures, pressures or plasma densities. Note that the thickness of the modified layer is greater for samples treated in the TUC system (sample sets #1–#3) which were subject to extra ion bombardment during the heating segment of the treatment.

Fig. 2. Light microcope micrograph of a cross-section obtained from a X6CrNiTi1810 sample after PI 3 treatment for 3 h at 400°C (set #2). Table 3 Layer thickness estimated from SEM cross-sections Treatment

Temperature (°C )

Layer thickness (mm)

#1 #2 #3 #4 #5 #6

300 400 500 400 400 400

1 4.3 6.7 2.8 3.9 3.3


The surface hardness after treatment at various temperatures is displayed in Fig. 3. The maximum microhardness values of 1040 HV0.01 for the austenitic steel is comparable to results obtained by conventional nitriding. The hardness measured at 50 mN load for specimens from sample sets #4–#6 was 13.3, 19.1 and 17.8 GPa respectively in comparison to 4.6 GPa for the untreated sample. The hardness can easily be increased by a few hundred percent compared to the untreated stainless steel just by the lattice expansion due to the supersaturation. Our previous work [3] indicated that this hardness increase is the most important contribution to the improvement in wear resistance reducing both the abrasive and the adhesive wear. 3.2. Wear The sliding wear behaviour of the untreated and treated samples was assessed using two different pin-ondisc tests. In all tests the performance of the untreated steel was similar, suffering severe wear (Fig. 4(a)). A mixture of adhesion, abrasion and plastic deformation in combination with a strong tribo-chemical reaction was observed in and around the wear track. Soon after the start of the test, a non-protective black oxide layer formed in the wear track which was alternately removed and built up by every pass of the ball. These oxide flakes acted as a third body in the wear process. They cut into the soft substrate leading to abrasive wear. Observations of material loss from the balls showed that they also suffered abrasive wear, producing debris which also contributed to the abrasive wear of the steel. Large variations in the friction force suggested that the disc steel and ball material were sticking together leading to adhesive wear as a result of forming and breaking up of metal–metal contacts. No significant difference in the wear of the untreated discs was observed between the different ball materials and even lubrication only had a slight effect, increasing the time before oxide formation began. After PI 3 treatment, the wear regime changed completely ( Fig. 4(b)). The severe metallic wear regime was replaced by a much milder wear behaviour. Instead of the strong oxide formation in the untreated case, only a thin and therefore still adhesive, oxide layer formed. The wear track was no longer caused by the removal of material and the build-up of oxides, but was a result of plastic deformation in the substrate into which the modified layer was pressed. Although the conclusions of the two different sliding wear tests are similar, the results for each of the two tests are discussed separately. 3.2.1. Rotating pin-on-disc On all PI 3 treated wear specimens a thin oxide layer formed on the surface. In combination with the increase in surface hardness due to the expanded austenite, the


C. Blawert, B.L. Mordike / Surface and Coatings Technology 116–119 (1999) 352–360

Fig. 3. Vickers surface microhardness using various indentation loads on the X6CrNiTi1810 stainless steel treated at different temperatures.



Fig. 4. Comparison of wear tracks on (a) untreated and (b) treated X6CrNiTi1810 (set #4) stainless steel wear specimens (200x magn.). Wear conditions: lubricated rotating pin-on-disc test, 10 N load, 0.05 m/s, treated X5CrNiMo1810 pin (a) 500 and (b) 30 000 turns.

surface is effectively protected from wear. Consequently, there was also no evidence of adhesive wear or strong plastic deformation at the edges of the wear tracks. A

few scratches visible in the wear tracks indicate a very mild abrasive wear. The influence of the load on the wear depth is shown

C. Blawert, B.L. Mordike / Surface and Coatings Technology 116–119 (1999) 352–360


Fig. 5. Influence of the load on the wear scar depth. The X6CrNiTi1810 discs were taken from set #4 and the X5CrNiMo1810 balls were treated twice as long as set #4 with all other parameters remaining unchanged. Wear conditions: ethanol lubrication, 30 000 turns at 0.05 m/s.

in Fig 5. With increasing load, the wear depth increases only slightly. In our previous work, it was shown that the wear track is a result of plastic deformation of the underlying substrate [3]. The modified layer of expanded austenite is not removed but pressed into the softer substrate under the applied load. Up to 10 N load, the load bearing capacity of this layer is sufficiently high to prevent much plastic deformation. Loads of around 20 N are necessary to press the layer into the substrate, as indicated in Fig. 5, by a large increase in wear depth. However, if the thickness of the expanded austenite layer can be increased, the load bearing capacity will be higher and the plastic deformation will be reduced. This is possible without increasing the treatment time and temperature, just by increasing the treatment pressure

(treatment #6) or the plasma density (treatment #5). The result is a thicker layer than that produced under the standard PI 3 treatment conditions (treatment #4) and improved wear behaviour as shown in Fig. 6. Although the adhesive wear is dramatically reduced by the PI 3 treatment, a small amount still occurs, especially at lower speeds. The wear depth on treated specimens is reduced by a factor of two when the speed is increased by a factor of three. The contact condition changes with increasing speed and the chance of forming metal–metal contacts is reduced. Measurement of the friction force supports this statement. The recorded trace shows a reduction in the average value with increasing speed. Increasing the speed from 0.05 to 0.15 m/s reduces the variation in the friction force as

Fig. 6. Influence of the PI 3 treatment conditions on the wear scar depth. Treatment conditions for the discs: #4 standard, #5 higher plasma density; #6 higher pressure. The X5CrNiMo1810 balls were treated for 6 h at 400°C with treatment conditions as in set #4. Wear conditions: ethanol lubrication, 30 000 turns at 0.05 m/s for the treated specimens and only 5000 turns for the untreated specimens.


C. Blawert, B.L. Mordike / Surface and Coatings Technology 116–119 (1999) 352–360

Fig. 7. Influence of the speed on the coefficient of friction. Data taken under the test conditions described in Fig. 6.

well, while a further increase has no further effect (Fig. 7). Part of the variation in the friction force can be attributed to the forming and breaking up of metal– metal contacts. 3.2.2. Oscillating pin-on-disc Generally, the wear is increased under dry wear conditions. Therefore a relatively smaller number of cycles is sufficient to reveal the effect of PI 3 treatment

( Fig. 8). Samples from treatments #1–#3 were used in this test to assess the influence of the treatment temperature on the wear performance. Increasing the treatment temperature results in thicker layers and higher load bearing capacities, but can also induce structural changes if the temperature becomes too high. Around 500°C, CrN precipitation occurs and, while this produces further improvement in the wear resistance, it also results in reduced corrosion resistance.

Fig. 8. Oscillating pin-on-disc wear depth as a function of the number of cycles. Wear partners were X6CrNiTi1810 discs treated at different temperatures (set #1– #3) and untreated 100Cr6 balls. Wear conditions: dry wear, 10 N load at 0.05 m/min.

C. Blawert, B.L. Mordike / Surface and Coatings Technology 116–119 (1999) 352–360

Fig. 9. SEM micrograph of the wear track after 1000 cycles of oscillating pin-on-disc test at 10 N and 0.05 m/min on an X6CrNiTi1810 wear specimen treated for 3 h at 500°C. The wear track is pressed into the substrate without significant removal of layer material.

Overall, the wear mechanisms observed in the unlubricated oscillating tests are identical to those observed in the lubricated rotating pin-on-disc tests but much more pronounced. For this reason we will concentrate in what follows on the appearance of the wear tracks and the differences between untreated and treated specimens. In contrast to the untreated specimens, a thin adhesive protective oxide layer forms by a mild tribochemical reaction on the hardened surface of the treated specimens. In consequence, there is a strong reduction of both abrasive and adhesive wear so that the original


surface structure is preserved after PI 3 treatment and can be seen under the oxide layer. The surface layer is pressed into the softer substrate, as can be seen in Fig. 9. This plastic deformation leads to material build-up at the edges of the wear track (Fig.10). Almost no material is lost and the volume of the build-up material is the same as the volume that is ‘lost’ in the track. With increasing temperature, the surface hardness and layer thickness grow and the plastic deformation is reduced. Note that the vertical scale of the profile measured on the untreated specimen in Fig. 10 is five times greater than for the untreated specimens. The very rough profile of the wear track on the untreated specimen can be attributed to the formation of the thick oxide layers observed in the micrographs and to the high amount of abrasive wear which results in deep scars within the track. In stark contrast, wear tracks on the treated specimens are smooth. The similarity between the lubricated rotating and unlubricated oscillating tests, which was previously noted for other material combinations using the same types of model wear test [22], is remarkable. A low speed was chosen in the lubricated test so that metal– metal contact could still occur. However under dry conditions it was necessary to replace the austenitic stainless steel ball ( X5CrNiMo1810) by a harder ball bearing steel (100Cr6) to obtain remarkable wear on the disc and not on the ball. After that it was not possible to observe any change in the dominant wear mechanisms in the two tests. The only difference found was in the time needed to reach a certain state of wear. Under dry conditions, the wear speed is enhanced. For example, observing the wear depth as a function of the applied load in the dry test results in the same critical load of 20 N that was observed in the lubricated test.

4. Conclusions Controlled surface treatment of stainless steel is possible using PI 3, keeping the nitrogen in solid solution

Fig. 10. Profiles of the wear tracks after 1000 cycles of oscillating pin-on-disc test at 10 N load and 0.05 m/min on X6CrNiTi1810 discs with different treatment temperatures.


C. Blawert, B.L. Mordike / Surface and Coatings Technology 116–119 (1999) 352–360

with only a minor effect on the passivation ability of the chromium, maintaining overall good corrosion resistance and increasing wear resistance. The formation of expanded austenite results in an increase in surface hardness which leads to a change in the wear mechanisms, even though the change in the coefficient of friction is only small. For untreated stainless steel, severe adhesive and abrasive wear was observed while treated samples suffered only very mild abrasive wear, a small amount of tribo-chemical reaction and a kind of plastic deformation. The wear can be influenced by a variety of treatment parameters because the layer thickness determines how well the substrate is able to support the modified layer. The dominant wear mechanism for treated samples is that the modified layer is pressed into the substrate under the applied load. So the hardness, the thickness of the layer and the strength of the substrate are important for the wear behaviour of the layer. However, short treatment times of three hours are able to achieve good wear resistance at low and medium loads. At elevated temperatures or after longer treatments, even thicker layers can be produced if required for higher loads. Besides temperature and time, layer thickness can also be influenced by pressure and plasma density. It is preferable to use the latter two parameters to influence the thickness of the layer, since longer treatment times incur higher treatment costs and higher treatment temperatures may transform expanded austenite into ferrite and CrN with subsequent reduction in corrosion resistance. To summarise, it appears that PI 3 treatment has the potential to become a very useful surface treatment for stainless steels, extending their use to a wider range of applications because the surface can be hardened without loss of corrosion resistance.

Acknowledgements This work has been supported by the Bundesministerium fu¨r Bildung, Wissenschaft, Forschung und Technologie (BMBF ) under contract numbers FKZ 13N6345 and AUS-028-96. We thank ANSTO for the use of their facilities for PI 3 treatment and evaluation of the samples whilst C. Blawert was a visiting scientist at ANSTO. In particular,

we appreciate the assistance of, and many fruitful discussions with, George Collins, John Tendys and Ken Short.

References [1] G.A. Collins, R. Hutchings, K.T. Short, J. Tendys, Heat Treatment of Metals 4 (1995) 91. [2] M. Samandi, B.A. Shedden, D.I. Smith, G.A. Collins, R. Hutchings, J. Tendys, Surf. Coat. Technol. 59 (1993) 261. [3] C. Blawert, F.M. Knoop, A. Weisheit, B.L. Mordike, Surf. Coat. Technol. 85 (1996) 15. [4] B.A. Shedden, M. Samandi, G.A. Collins, Materials Science Forum 189 (1995) 435. [5] M. Samandi, B.A. Shedden, T. Bell, G.A. Collins, R. Hutchings, J. Tendys, J. Vac. Sci. Technol. B 12 (1994) 935. [6 ] R. Leutenecker, G. Wagner, T. Louis, U. Gonser, L. Guzman, A. Molinari, Mater. Sci. Eng. A115 (1989) 229. [7] D.L. Williamson, L. Wang, R. Wei, P.J. Wilbur, Mater. Lett. 9 (1990) 302. [8] D.L. Williamson, O. Ozturk, S. Glick, R. Wei, P.J. Wilbur, Nucl. Instr. Meth. B 59/60 (1991) 737. [9] M.E. Chabica, D.L. Williamson, R. Wei, P.J. Wilbur, Surf. Coat. Technol. 51 (1992) 24. [10] Z.L. Zhang, T. Bell, Surf. Eng. 1 (1985) 131. [11] P.A. Dearnley, A. Namvar, G.G.A. Hibberds, T. Bell, E. Broszeit, in: W.D. Munz, H. Oechsner, K.-T. Rie, G.K. Wolf ( Eds.), Proc. 1st Int. Conf. on Plasma Surface Engineering, Garmisch-Partenkirchen, Sept., 1988 vol. 1, Deutsche Gesellschaft fu¨r Metallkunde Informationgellschaft, Oberursel, 1989, p. 219. Vol. 1. [12] S.P. Hannula, P. Nenonen, J.P. Hirvonen, Thin Solid Films 181 (1989) 343. [13] K. Ichii, K. Fujimura, T. Takase, Technol. Rep. Kansai Univ. 27 (1986) 135. [14] E. Menthe, K.-T. Rie, J.W. Schultze, S. Simson, Surf. Coat. Technol. 74/75 (1995) 412. [15] A. Saker, Ch. Leroy, H. Michel, C. Frantz, Mater. Sci. Eng. A140 (1991) 702. [16 ] D.L. Williamson, O. Ozturk, R. Wei, P.J. Wilbur, Surf. Coat. Technol. 65 (1994) 15. [17] D.L. Williamson, R. Wei, P.J. Wilbur, Nucl. Instr. Meth. B 56/ 57 (1991) 625. [18] R. Wei, P.J. Wilbur, W.S. Sampath, D.L. Williamson, L. Wang, Lubrication Engineering 47 (4) (1991) 326. [19] D.L. Williamson, Nucl. Instr. Meth. B 76 (1993) 262. [20] G.A. Collins, R. Hutchings, K.T. Short, J. Tendys, C.H. van der Valk, Surf. Coat. Technol. 84 (1996) 537. [21] M.J. Baldwin, G.A. Collins, M.P. Fewell, S.C. Haydon, S. Kumar, K.T. Short, J. Tendys, Jap. J. Appl. Phys. 36 (1997) 4941. [22] C. Blawert, B.L. Mordike, G.A. Collins, R. Hutchings, K.T. Short, J. Tendys, Surf. Coat. Technol. 83 (1996) 228. [23] C. Blawert, B.L. Mordike, G.A. Collins, K.T. Short, J. Tendys, Surf. Coat. Technol. 103 (1998) 240. [24] C. Blawert, B.L. Mordike, Y. Jira´skova´, O. Schneeweiss, paper presented at the 4th Int. Workshop on Plasma Based Ion Implantation, Detroit, USA, 2nd–4th June, 1998, Surface Engineering, submitted for publication.