Nitriding of stainless steel and aluminium alloys by plasma immersion ion implantation

Nitriding of stainless steel and aluminium alloys by plasma immersion ion implantation

Surface and Coatings Technology 128᎐129 Ž2000. 21᎐27 Nitriding of stainless steel and aluminium alloys by plasma immersion ion implantation E. Richte...

415KB Sizes 0 Downloads 75 Views

Surface and Coatings Technology 128᎐129 Ž2000. 21᎐27

Nitriding of stainless steel and aluminium alloys by plasma immersion ion implantation E. Richter U , R. Gunzel, S. Parasacandola, T. Telbizova, O. Kruse, W. Moller ¨ ¨ Forschungszentrum Rossendorf e.V., Institute for Ion Beam Physics and Materials Research, P.O. Box 51 01 19, D-01314 Dresden, Germany

Abstract Stainless steels show excellent corrosion resistance, which is lost during conventional hardening processes at temperatures above 500⬚C, like gas or plasma nitriding, to improve the low hardness and to reduce the high wear rate. For aluminium alloys it is impossible to improve the mechanical properties by traditional nitriding. Plasma immersion ion implantation ŽPIII. of nitrogen is successfully used for hardening both austenitic stainless steels and aluminium alloys. Compared to the untreated materials the hardness and the wear resistance can be improved significantly. For austenitic stainless steels an ‘expanded austenite’ layer of up to several tens of micrometer thickness and for aluminium alloys an AlN layer of more than 10-␮m thickness were formed over a few hours. Corrosion tests showed no or only small changes in the corrosion behaviour for stainless steel. The AlN layer can be used as a supporting layer for the deposition of hard materials like CrN on aluminium alloys. It is shown that the partial pressure of oxygen in the residual gas in the vacuum chamber plays an important role for the nitriding of both stainless steel and aluminium alloys. 䊚 2000 Elsevier Science S.A. All rights reserved. Keywords: Plasma immersion ion implantation; Stainless steel; Aluminium alloys; Nitriding; Mechanical properties

1. Introduction Stainless steels are used for their excellent corrosion properties in a broad application range. This extends, i.e. from the food and chemical industry to medicine technology. On the other hand, aluminium and its alloys as lightweight construction materials with good corrosion resistance are widely used in the car, space and aircraft industries. However, the low hardness Ž200᎐300 kprmm2 for stainless steels and 50᎐100 kprmm2 for aluminium alloys. and large wear in abrasively stressed parts lead often to short lifetimes that could be prolonged with improved tribological properties. Diffuse processes using nitrogen are broadly applied to the surface hardening of metallic components, in particular steels. A number of different techniques U

Corresponding author. Tel.: q49-351-2603326; fax: q49-3512602-703. E-mail address: [email protected] ŽE. Richter..

are conventionally applied in industrial production, such as salt bath nitriding, nitriding from the gas phase, carburising w1x and plasma nitriding w2x. With these conventional nitriding processes, stainless steel-specific problems have been encountered. In order to avoid a transformation of the chromium oxide into chromium nitride at the surface, by which the anticorrosive properties would be lost, the process temperature is limited to 400᎐450⬚C w3x. Very long processing times are required. In the case of aluminium alloys the very compact oxide layer on the surface is a strong barrier for the diffusion of nitrogen into the bulk material. It is impossible to produce aluminium nitride surface layers thicker than 1᎐3 ␮m by conventional nitriding w4x. To enhance the tribological properties for both stainless steels and aluminium alloys the implantation of nitrogen ions behind the surface oxide barrier is an interesting possibility. Additionally, radiation-enhanced diffusion is observed in several systems. Therefore, both effects enable a lower process temperature below

0257-8972r00r$ - see front matter 䊚 2000 Elsevier Science S.A. All rights reserved. PII: S 0 2 5 7 - 8 9 7 2 Ž 0 0 . 0 0 6 3 8 - 1

22

E. Richter et al. r Surface and Coatings Technology 128᎐129 (2000) 21᎐27

400⬚C to be used for a given depth and treatment time. However, high costs and a sophisticated system for target manipulation limit the technological applications when three-dimensional workpieces are treated using the conventional beam line implantation. Plasma immersion ion implantation ŽPIII. was developed as an alternative technology to conventional ion implantation w5᎐11x. For PIII a plasma is generated in a sufficiently sized vacuum chamber, immersing the workpiece from all sides. By applying negative high voltage pulses to the sample, positively charged ions are extracted from the plasma through the plasma sheath and implanted into the whole surface at the same time, thereby reducing the process time for larger pieces and decreasing the costs. In principle, a dc voltage source may be used instead of a pulsed one. However, the pulsed mode is needed to avoid an undesired heating of the workpieces. In some PIII systems designed for commercial applications, the temperature during the treatment is even regulated by the repetition rate of the high voltage pulses w3,12x. This paper reports on increasing hardness, wear resistance at retained corrosion resistance after nitriding by PIII of the stainless steel X5CrNiMo17.12.2 and the aluminium alloys AlMg4.5Mn and AlMgSi1. Additionally, the depth distribution of the implanted nitrogen and the phase composition were analysed. The evolution of the near surface composition during the nitriding process has been analysed by real-time elastic recoil detection analysis. The interplay of sputtering and oxidation emerges as a key parameter for the successful nitriding of both stainless steel and aluminium underlining the importance of good vacuum conditions.

2. Experimental S a m p le s o f a u s te n itic s ta in le s s s te e l X5CrNiMo17.12.2 ŽDIN 1.4401; composition Žwt.%.: C - 0.07, Si 1, Mn 2, P 0.045, S 0.03, Cr 16.5᎐18.5, Mo 2᎐2.5, Ni 10.5᎐13.5., and the aluminium alloy AlMg4.5Mn and AlMgSi1, polished to a mirror finish and ultrasonically cleaned before use, were implanted with nitrogen using PIII. The implantation time was in the range from 2 to 6 h. The negative pulses Žy40 kV. with a rise time of 1 ␮s and a total pulse length of 5 ␮s were supplied by a switch using a hard tube tetrode system w13x. The temperature of the samples was measured with a pyrometer or a thermocouple. The repetition rate was accordingly adjusted to maintain the selected processing temperature. The depth profiles of the implanted nitrogen were measured with glow discharge optical spectroscopy ŽGDOS. which can be used for a depth up to several tens of micrometers. The microstructure of the modi-

fied surface region was analysed with XRD with different angles of incidence from 1 to 10⬚. Depth-dependent hardness data were obtained from dynamic Vickers microhardness measurements performed with a load up to 2 N on a Shimadzu DUH-202 device w14x. The sliding wear behaviour was determined from dry wear tests using an oscillating ball-on-disk type tribometre. A fixed ball Ž3 mm diameter, WC-Co. was pressed to the samples with a load of 3 N and moved with an average velocity of 1.5 cmrs. In order to demonstrate the role of oxygen for the nitriding process a model experiment was set up in a UHV chamber. The partial pressure of oxygen was systematically changed from 9 = 10y3 to 3 = 10y5 Pa. The nitriding treatment was performed using a Kaufman-type ion source at fixed ion energy, ion flux and substrate temperature for 10 min. For real-time in-situ compositional analysis the process chamber was connected to the Rossendorf 5 MV Tandem accelerator delivering a low flux of 35 MeV Cl7q ions for elastic recoil detection analysis ŽERDA.. Details of the experimental set-up for the ERDA-technique have been given elsewhere w15x. Finally, the corrosion behaviour was tested by potentiodynamic measurements in 0.1 N H 2 SO4 solution with a sweep rate of 10 mVrs. The potential was measured against a standard calomel electrode ŽSCE..

3. Results and discussion 3.1. Austenitic stainless steel In Fig. 1 the GDOS nitrogen depth profile is shown for the X5CrNiMo17.12.2 austenitic stainless steel after nitriding by PIII for 6 h at 380⬚C. The nitrogen signal is visible down to 17 ␮m, largely exceeding the mean projected range of approximately 20 nm for N2q ions in steel. This indicates a rapid diffusion during the PIII treatment. The nitrogen profile shows a diffusion-like fraction superimposed to a nearly rectangular shape profile extending to the full depth. After annealing at 350⬚C for 24 h under vacuum the diffusion-like part of the profile disappears. The nitrogen concentration in the surface region decreases from 15 to 9 wt.% and the depth of the nitrided layer increases only by only a few micrometers. Directly on the surface a small oxide layer Ž- 1 ␮m. was formed. It is known that the implanted nitrogen becomes trapped with a maximum concentration of approximately 20 at.% to form the so-called ‘S’ or ‘expanded austenite’ phase. XRD investigations show an austenitic structure with a lattice expansion of approximately 7% w16x. If the processing temperature on the steel surface during the PIII treatment is limited to 400⬚C, CrN

E. Richter et al. r Surface and Coatings Technology 128᎐129 (2000) 21᎐27

Fig. 1. GDOS depth profile for austenitic stainless steel implanted with nitrogen using PIII in the as-implanted state and after annealing under vacuum.

precipitates are not or only marginally observable in the XRD spectrum, and ␧-nitride was not found. Annealing experiments at 350⬚C for 24 h show no changes in the XRD spectrum. However, after tempering at temperatures higher than 450⬚C the ‘expanded austenite’ structure breaks down and CrN and ␧-nitride were formed w17x. Corrosion measurements demonstrate that the processing temperature during the PIII treatment plays an important role for the properties of the stainless steel after nitriding. If the temperature was carefully controlled corrosion potential and passivation current are not significantly changed after PIII nitriding below 400⬚C. The corrosion properties are nearly independent of the implanted dose of nitrogen. In contrast, for processing temperatures ) 400⬚C the corrosion current increases dramatically by a few orders of magnitude. The decrease of corrosion resistance is probably due to the formation of CrN precipitates and ␧-nitrides in a ␥⬘-nitride matrix as reported in other papers w9x. Fig. 2 shows the measured hardness vs. the penetration depth of the used Vickers indenter for the X5CrNiMo17.12.2 stainless steel for different nitriding times Ž1, 4 and 10 h. at 380⬚C. The original bulk hardness was 200 H V , increasing at the surface to approximately 400 H V . The measured hardness increases after nitriding very steeply to 1200 H V in the surface region. As expected with increasing treatment time the hardness of the nitrided stainless steel increases. The resulting hardness is in the order of other steel types, hardened in conventional ways, i.e. by gas nitriding or plasma nitriding at higher temperatures. The observed hardness increase is in good agreement with the sliding wear behaviour of the PIII-treated stainless steel as shown in Fig. 3 for an untreated sample and a nitrided sample for 6 h at 380⬚C. A 6-h PIII implantation results in a substantial reduction of

23

Fig. 2. Hardness᎐depth function for stainless steel nitrided by PIII at different treatment times.

wear by a few orders of magnitude. Annealing for 24 h at 350⬚C in a vacuum has no influence on the wear behaviour as also shown in Fig. 3. If the surface temperature is controlled carefully during the PIII process the corrosion behaviour does not change ŽFig. 4.. 3.2. Aluminium In the case of aluminium and aluminium alloys the first experiments show that processing temperatures ) 400⬚C are necessary to form nitride layers thicker than 5 ␮m during a time relevant for applications. In the following investigations of the alloy AlMg4.5Mn will be reported. Fig. 5 shows the GDOS depth profile of the implanted nitrogen by PIII for 6 h at a processing temperature of approximately 500⬚C. An AlN layer with a thickness of approximately 15 ␮m is formed under the given conditions. The evidence for the stoichiometric composition ŽAlrNs 1:1. was obtained by XRD measurements. Only reflexes of AlN were found. The hardness directly at the surface of the nitrided

Fig. 3. Wear behaviour of stainless steel without and after nitriding by PIII and after annealing under vacuum, measured by pin-on-disk technique.

24

E. Richter et al. r Surface and Coatings Technology 128᎐129 (2000) 21᎐27

Fig. 4. Current density᎐potential function of untreated and PIII nitrided stainless steel, measured in 0.1 N H 2 SO4 against a standard calomel electrode ŽSCE..

AlMg4.5Mn alloy increases to 1200 H V, the value of compact AlN ceramics. The bulk hardness of the untreated alloy is approximately 100 H V , increasing at the surface to 300 H V . In Fig. 6 the sliding wear behaviour of the nitrided alloy is shown without any lubrication in relation to the untreated materials. Independent of the implanted dose Žtime of treatment . the wear resistance is substantially increased by a few orders of magnitude up to a sliding path of 250 m. After this sliding distance the AlN layer is penetrated and the bulk properties begin to dominate the wear behaviour, and the wear increases. For broader applications of aluminium alloys in the industry AlN surface layers with a thickness more than 5 ␮m will be interesting to support the deposition of typical hard materials like CrN or DLC top of the workpieces. Without any support the hard layer would break down because the alloy is too soft. In Fig. 7 an

example is given for the sliding wear behaviour of the untreated surface, a PIII nitrided AlN layer Ž6 ␮m. only, and additional TiNbN Ž2 ␮m. and CrN Ž4 ␮m. layers deposited after the PIII nitriding in the same processing chamber without any break of the vacuum using a standard PVD process. As shown in Fig. 7, this so called ‘duplex process’ leads to a long-time wear resistance for CrN. It is also shown that the adhesion plays an important role in such duplex processes for both the AlN on the basic materials and the hard layer on the AlN. The adhesion of the TiNbN top layer is poor. Although it shows a good wear resistance in the beginning of the sliding test, the wear increases rapidly, as it breaks down and delaminates. During the nitriding of AlMg4.5Mn by PIII the surface roughness is increased Ža mirror-like polished surface shows after nitriding for 6 h a roughness of RA f 2 ␮m.. Fig. 8 shows the typical surface topography ob-

Fig. 5. GDOS depth profile after nitriding by PIII.

Fig. 6. Wear behaviour of an aluminium alloy without and after nitriding by PIII at different treatment times.

E. Richter et al. r Surface and Coatings Technology 128᎐129 (2000) 21᎐27

Fig. 7. Comparison of the wear behaviour of an aluminium alloy untreated, nitrided by PIII and followed with a deposition of TiNbN or CrN hard layers by PVD technique.

tained from scanning electron microscopy ŽSEM.. This structure might enhance the adhesion of the suitable hard coatings being deposited on top of it. From the point of view of application it is important to compare the industrially employed plasma nitriding with the nitriding by PIII. Fig. 9 demonstrates the obtained results for the layer thickness and the resulting wear behaviour after nitriding of the alloy AlMgSi1 using both processes. Using the PIII for the nitriding at approximately 500⬚C the AlN layer is significantly

25

thicker Ž12 ␮m instead 1.5 ␮m for the plasma nitriding. and the resulting wear resistance is significantly better. Similar experiments with stainless steel ŽX5CrNi18.12. show the same trends. On the role of oxygen for the nitriding of stainless steel and aluminium alloys a dual beam experiment ŽFig. 10. has been set up that allows for time- and depth-resolved characterisation of the near surface composition during a well defined nitriding process. Nitriding is performed from a hot filament broad beam ion source in an UHV vacuum chamber with a base pressure of approximately 10y6 Pa equipped with a residual gas analyser and mass flow controlled gas inlets. This choice allows good control and a systematic variation of the important process parameters Že.g. substrate temperature, ion energy, ion flux, oxygen partial pressure.. The specific ion source and its performance is described elsewhere w18x. During the nitriding process compositional analysis is performed by means of Elastic recoil detection analysis ŽERDA.. ERDA is an ion beam analysis technique closely related to Rutherford Backscattering Spectroscopy ŽRBS.. It uses high-energy ions to generate recoil target atoms, the energy distributions of which can be converted into depth profiles. Opposed to RBS, ERDA is particularly suited for depth profiling of light elements. In the present study, ERDA is employed with a low flux Ž( 1

Fig. 8. SEM micrograph of an AlMg4.5Mn surface nitrided by PIII.

26

E. Richter et al. r Surface and Coatings Technology 128᎐129 (2000) 21᎐27

Fig. 9. Comparison of the wear behaviour of AlMgSi1 nitrided by plasma nitriding and PIII.

nArmm2 . of 35 MeV Cl7q and an angle resolving ionisation chamber that allows for fast data acquisition. The analysis can be considered to be non-destructive. During the nitriding process every 30 s meaningful depth profiles of nitrogen and oxygen are obtained. The depth resolution is systematically limited to approximately 10 nm. The maximum detection depth is approximately 200 nm. Details of the ERDA technique, the angle resolving ionisation chamber, and the set-up are described elsewhere w19,20x. The dual beam experiment is particularly useful to gain information on the role of the oxygen partial pressure p O 2 . Fig. 11 shows the evolution of the thickness of the surface oxide layer Žassuming Al 2 O 3 . and the nitride layer Žassuming AlN. for different p O 2 . during a 15-min nitriding treatment of pure polycrystalline at a sample temperature of 500⬚C, ion energy of 1 keV, and a current density of 200 ␮Arcm2 . Samples were held at the preset oxygen partial pressure for 10 min before the nitriding was started at time 0. At p O 2 s 3 = 10y3 Pa and p O 2 s 3 = 10y4 Pa the oxide is not removed during the nitriding treatment. Instead a

Fig. 10. The dual beam experiment set-up.

stationary oxide thickness of approximately 10 nm is established. In contrast at p O 2 - 3 = 10y5 Pa the oxide is essentially removed after 5 min. The corresponding nitride layer evolutions are strongly influenced by the presence of the oxide layer Žthe oxygen partial pressure.. When the surface oxide layer is removed at p O 2 - 3 = 10y5 Pa a constant nitride layer growth rate is observed indicating that the growth is not limited by diffusion. At the higher oxygen partial pressures the nitride layer growth is strongly hindered. The corresponding evolution of the depth profiles and a more detailed discussion on the kinetics are presented elsewhere w21x. The existence of a non-zero stationary surface oxide

Fig. 11. Oxide ŽAl 2 O 3 . and nitride ŽAlN. layer thickness vs. time before and during nitriding pure polycrystalline aluminium at 500⬚C, 1 keV, 200 ␮Arcm2 , and the oxygen partial pressures given in the figure.

E. Richter et al. r Surface and Coatings Technology 128᎐129 (2000) 21᎐27

layer is in agreement with semi-quantitative considerations on oxide growth from the residual gas and oxide removal due to sputtering. As shown by Parascandola et al. w22x the interplay of sputtering and oxidation can be considered as a key parameter for the successful nitriding of austenitic stainless steel and aluminium alloys.

4. Conclusions A substantial reduction of wear by more than two orders of magnitude and an increase of hardness was obtained for both austentic stainless steel and an aluminium alloy by nitrogen implantation using PIII. The austenitic stainless steel X5CrNiMo17.12.2 formed at 380⬚C the ‘expanded austenite’ phase in the surface region without loss of the corrosion resistance. For the aluminium alloy AlMg4.5Mn implanted by PIII at approximately 500⬚C, an AlN layer at the surface is obtained with a thickness up to 15 ␮m, depending on the treatment time. For aluminium alloys the nitriding by PIII to produce a supporting surface layer was combined with conventional PVD of hard coatings. This so-called duplex process is particularly attractive for technological applications. Comparing the well-established plasma nitriding at the same processing temperature Ž380⬚C for stainless steel and approximately 500⬚C for aluminium. with the PIII process, gave substantially thicker modified layers with substantially better wear properties for PIII. Using a dual beam experiment for nitriding with a Kaufman type ion source and real-time in-situ analysis by means of ERDA, it is shown that the oxygen partial pressure plays an important role for the formation of both the expanded austenite and the AlN surface layer. Oxygen is to be removed completely from the surface by sputtering for an effective nitriding, especially for the aluminium alloys.

27

References w1x R.H van der Jagdt, B.H. Kolster, M. Gillham, Mat. Design 12 Ž1991. 41. w2x G.A. Collins, D.J. Rei, MRS Bull. 21 Ž1996. 26. w3x M. Samandi, B.A. Sheddon, D.J. Smith, G.A. Collins, R. Hutchings, J. Tendys, Surf. Coat. Technol. 59 Ž1993. 26. w4x E. Menthe, K.T. Rie, J.W. Schultze, S. Simon, Surf. Coat. Technol. 74-75 Ž1995. 412. w5x J. Conrad, J. Appl. Phys. 62 Ž1987. 777. w6x J. Conrad, J.L. Radtke, R.A. Dodd, F.J. Worzala, N.C. Tran, J. Appl. Phys. 62 Ž1987. 4591. w7x J. Tendys, I.J. Donelly, M.J. Kenny, I.T.A. Pollock, Appl. Phys. Lett. 53 Ž1988. 261. w8x S. Mandl, J. Brutscher, R. Gunzel, W. Moller, J. Vac. Sci. ¨ ¨ ¨ Technol. B14 Ž1996. 2701. w9x W. Ensinger, Nucl. Instrum. Meth. B 120 Ž1996. 270. w10x S. Mandl, N.P. Barradas, J. Brutscher, R. Gunzel, W. Moller, ¨ ¨ ¨ Nucl. Instrum. Meth. B127r128 Ž1997. 996. w11x J.V. Mantese, I.G. Brown, N.W. Cheung, G.A. Collins, MRS Bull. 21 Ž1996. 52. w12x D.L. Williamson, O. Ozturk, R. Wei, P.J. Wilbur, Surf. Coat. Technol. 65 Ž1994. 15. w13x J. Brutscher, Rev. Sci. Instrum. 67 Ž1996. 2621. w14x T. Chudoba, PhD Thesis, Technical University of Dresden, Germany, 1996. w15x W. Assmann, H. Huber, C. Steinhausen, M. Dobler, H. Gluckler, A. Weidinger, Nucl. Instrum. Meth. B 89 Ž1994. 131. ¨ w16x S. Mandl, R. Gunzel, E. Richter, W. Moller, Surf. Coat. Tech¨ ¨ ¨ nol. 100-101 Ž1998. 372. w17x E. Menthe, PhD Thesis, Technical University of Braunschweig, Germany, 1999. w18x M. Zeuner, J. Meichsner, H. Neumann, F. Scholze, F. Bigl, J. Appl. Phys. 80 Ž1996. 611. w19x S. Parascandola, O. Kruse, E. Richter, W. Moller, J. Vac. Sci. ¨ Technol. B 17 Ž1999. 855. w20x O. Kruse, S. Parascandola, R. Grotzschel, W. Moller, accepted ¨ ¨ for publication in the Proceedings of the MRS Spring Meeting, Symposium U, San Francisco, April 5᎐9, 1999. w21x S. Parascandola, T. Telbizova, O. Kruse, W. Moller, accepted ¨ for publication in Nucl. Instrum. Meth. B. w22x S. Parascandola, O. Kruse, W. Moller, accepted for publication ¨ in Appl. Phys. Lett.