The influence of the sputtering process on the constitution of the compound layers obtained by plasma nitriding

The influence of the sputtering process on the constitution of the compound layers obtained by plasma nitriding

Surface and Coatings Technology 174 – 175 (2003) 1201–1205 The influence of the sputtering process on the constitution of the compound layers obtaine...

142KB Sizes 1 Downloads 20 Views

Surface and Coatings Technology 174 – 175 (2003) 1201–1205

The influence of the sputtering process on the constitution of the compound layers obtained by plasma nitriding C. Ruseta, S. Ciucab, E. Grigorea,* a

b

National Institute for Laser, Plasma and Radiation Physics, P.O. Box MG-36, Bucharest, Romania Polytechnic University of Bucharest, Materials Science and Engineering, Splaiul Independentei, No. 313, Bucharest, Romania

Abstract Using a relative simple experimental method, the sputtering rate has been measured under specific plasma nitriding conditions. The influence of gas composition and total pressure on the sputtering rate has been investigated. Depending on these factors, the sputtering rate varies in the range of 0.02–0.12 mmyh. The sputtered material has been analyzed as well. The results indicated that the phase constitution of this material is similar to that of the compound layer. 䊚 2003 Elsevier Science B.V. All rights reserved. Keywords: Plasma nitriding; Sputtering rate; Compound layer; X-ray diffraction analysis

1. Introduction

composition. The structure of the sputtered material during plasma nitriding was also investigated.

It is well known that the sputtering, as a specific process for plasma thermochemical treatments, plays an essential role in formation and developing of hard surface layers, particularly of compound layers w1–3x. This process has been extensively studied, either as a fundamental phenomenon or in connection with some particular applications. Based on a model of the plasma nitriding process, a computer program, able to predict the layer characteristics has been developed w4x. The chemical composition of the substrate material, process parameters and sputtering rate have been used as input data. In order to get a good agreement between the theoretical (predicted) and the experimental results, the real values of the sputtering rate must be introduced into the program. A lack of data appears in the literature on this subject. Therefore in the present paper, an investigation of sputtering under plasma nitriding conditions is presented. The aim of this investigation was to find out the influence of the process parameters on the sputtering rate and its correlation to the compound layer phase

2. Experimental

*Corresponding author. Tel.: q40-21-457-4550x1857; fax: q4021-457-4243. E-mail address: [email protected] (E. Grigore).

The experimental arrangement for the measuring of the sputtering rate is shown in Fig. 1a. The cathode (1), which is a rod (⭋20=380 mm) made of C45 plain carbon steel, is surrounded by five hollow cylinders (⭋80=70 mm) made of stainless steel plate with a thickness of 0.35 mm. These cylinders are acting like an anode and collect the sputtered material from the cathode. By weighting the cylinders before and after the treatment, the sputtering rate at the cathode surface was calculated. It was supposed that the collected material was g9-Fe4N, with a density of 6.8 gycm3. Approximately equal amounts of material have been collected by each of those five cylinders, proving the uniformity of the sputtering process along the cathode rod. The hypothesis concerning the nature of the sputtered material was confirmed by other set of experiments using the device shown in Fig. 1b. The material sputtered from the hemispherical cathode (6) is collected by the copper disc (7) and analyzed by XRD. Due to the low values of the sputtering rate, the process time was between 10 and 30 h in order to get a significant amount of material to be analyzed. The main process parameters, which have been changed, were total pressure, correlated

0257-8972/03/$ - see front matter 䊚 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0257-8972(03)00589-9

1202

C. Ruset et al. / Surface and Coatings Technology 174 – 175 (2003) 1201–1205

Fig. 2. The influence of the working pressure (a) and of the gas composition (b) on the sputtering rate.

Fig. 1. Experimental set-up for measuring the sputtering rates (a) and the nature of the sputtered material (b). (1) cathode; (2) cylindrical anodes; (3) cathode support; (4) anode support rods; (5) top lid of the treatment chamber; (6) hemispherical cathode; (7) copper disc (anode).

the hydrogen percentage in the treatment atmosphere decreases from 95 to 20%, the sputtering rate decreases from 0.13 to 0.05 mmyh. This is also a consequence of the discharge voltage reduction, which occurs when the nitrogen percentage in the processing atmosphere is increased.

with the discharge voltage in order to keep the same temperature of the cathode, and the gas composition. The nitrogen and hydrogen partial pressures were monitored by means of a quadrupole mass spectrometer. Chemical composition of the compound layers has been analyzed by glow discharge optical spectrometry method. 3. Results and discussion The influence of the processing pressure on the sputtering rate is shown in Fig. 2a. For low pressure values (p-5.2 mbar) the sputtering rate does not change significantly, while for high values (p)5.2 mbar), this rate drops drastically from 0.12 to 0.02 mmyh. This is due to a decrease of the ion and fast neutral energy, determined by the reduction of the discharge voltage and of the mean free path. A similar effect appears when the gas composition is changed (Fig. 2b). When

Fig. 3. The dual g9q´ structure of a compound layer produced on C45 steel by plasma nitriding at 520 8C, 10.5 mbar for 8 h in 25% N2q75% H2 gas mixture.

C. Ruset et al. / Surface and Coatings Technology 174 – 175 (2003) 1201–1205

1203

Fig. 4. Influence of the composition of hydrogen–nitrogen gas mixture on the carbon and nitrogen profiles (A) and constitution (B) within the compound layers for C45 steel. (a) 100% H2; (b) 75% H2q25% N2; (c) 20% H2q80% N2.

It is well established in Refs. w3,5x that the sputtering is directly responsible for the decarburisation process occurring during the plasma nitriding. Consequently, when the treatment parameters are set for a high sputtering rate, the decarburisation is so intense that a very low carbon concentration is reached at the cathode

surface and according to the Fe–N–C phase diagram w6x, a g9-Fe4N mono-phase compound layer is formed. If the plasma nitriding is performed at high pressure ()5.2 mbar) or with a high nitrogen percentage in the treatment atmosphere, the sputtering rate is too low to produce the necessary level of decarburisation and con-

1204

C. Ruset et al. / Surface and Coatings Technology 174 – 175 (2003) 1201–1205

Fig. 5. XRD pattern of the sputtered material, collected on the copper anode. The experiment was carried out at 520 8C, 2.0 mbar for 30 h in 25% N2q75% H2 gas mixture.

sequently, a significant amount of ´-Fe2–3(N,C) phase is produced along with g9. The XRD pattern of a C45 sample treated by plasma nitriding at 520 8C, for 8 h, at a pressure of 10.5 mbar, shows clearly a dual g9q´ structure of the compound layer (Fig. 3). The working gas was 25% N2q75% H2. The influence of the gas composition on the structure of the compound layer, together with the depth profile of nitrogen and carbon concentrations is shown in Fig. 4. A direct correlation between the carbon profile within the compound layer and its structure can be seen. A strong decarburising effect appears when the treatment atmosphere contains only hydrogen (Fig. 4Aa). When 25% nitrogen is introduced in the chamber, the decarburisation is attenuated, but it remains at a relative high level. As a consequence, the carbon concentration at the surface is low enough (Fig. 4Ab) and it is associated with the g9 mono-phase structure as it can be seen in Fig. 4Bb. With further increase of the nitrogen to 80%, the sputtering rate and decarburisation drop even more (Fig. 4Ac), the carbon atoms cannot be removed from the surface and they are bounded, together with nitrogen atoms, in the ´-Fe2–3(N,C) structure (Fig. 4Bc). The influence of the pressure on the sputtering rates seems to be stronger than that of the nitrogen. This can be seen in Fig. 2 and it is confirmed by XRD patterns shown in Figs. 3 and 4Bc. In the first case, the ´-phase is dominant in the dual structure, while in the second case the g9 is dominant. Similar results have been obtained with low alloy steels. Using the experimental set-up shown in Fig. 1b, the material, sputtered under plasma nitriding conditions, has been investigated as well. The XRD pattern shown in Fig. 5 indicates that the phase constitution of the material collected on the copper anode is mono-phase g9. The deposited layer is very thin, so the copper substrate can be seen as well. No a-Fe could be detected.

On the other hand, the experiment has been performed at low pressure (2.0 mbar), high voltage (850 V), with 25% N2q75% H2 gas mixture. Under these conditions it is sure that the compound layer has a g9 mono-phase structure. This means that the material from the cathode surface (g9-Fe4N) has been transferred to the anode with the same structure by sputtering. Similar results have been obtained at high pressure where a dual g9q´ structure has been transferred. 4. Conclusions The compound layers produced by plasma nitriding have not always a g9 mono-phase structure, but a dual g9q´-phase composition can appear depending on the treatment parameters. The sputtering phenomenon seems to be responsible for the structure of the compound layer. Depending on the current density at the surface of the cathode and on the binding energy of the compound formed at that surface, the effect of the sputtering process might be different. For low current densities, which can be found in plasma nitriding process (0.5–2 mAycm2), the material is transferred from cathode to anode (or to other substrate) without changing the structure. Similar phenomena seem to occur for intermediate current densities (10–40 mAycm2) like PVD-magnetron sputtering, while at high current densities (;200 mAycm2) such as in glow discharge optical spectrometry, the structure is destroyed and the material is sputtered as chemical constituents. References w1x K. Keller, Hart. Tech. Mitte. 26 (1971) 120–130. w2x B. Edenhofer, Heat Treat. Metals 2 (1974) 59–67.

C. Ruset et al. / Surface and Coatings Technology 174 – 175 (2003) 1201–1205 w3x T. Lampe, S. Eisenberg, G. Landien, Surf. Eng. 9 (1) (1993) 69–76. w4x Y. Sun, T. Bell, Modelling of the plasma nitriding of low alloy steels, Conference on Developments in the Nitriding of Iron and Titanium Based Alloys, The Royal Society, London, March 1, 1995, pp. 2–5.

1205

w5x C. Ruset, A. Bloyce, T. Bell, Towards an epsilon mono-phased compound layer on plain carbon steels by plasma nitrocarburising, Proceedings of the Tenth Congress of the IFHT, Brighton, England, September 1–5, 1996, pp. 150–168. w6x J. Slycke, et al., Scand. J. Metall. 17 (1988) 122–126.