The influence of surface microcracks and temperature gradients on the rf plasma nitriding rate

The influence of surface microcracks and temperature gradients on the rf plasma nitriding rate

Surface and Coatings Technology 150 (2002) 277–281 The influence of surface microcracks and temperature gradients on the rf plasma nitriding rate F.M...

443KB Sizes 0 Downloads 17 Views

Surface and Coatings Technology 150 (2002) 277–281

The influence of surface microcracks and temperature gradients on the rf plasma nitriding rate F.M. El-Hossary* Physics Department, Science Faculty, South Valley University, Sohag branch, Sohag, Egypt Received 25 May 2001; accepted in revised form 11 September 2001

Abstract Rf plasma nitriding processes have been used recently for surface treatment. 304 Austenitic stainless steel has been nitrided using an inductively coupled rf plasma. High rates of nitriding and relatively high microhardness have been achieved. The effect of plasma power and plasma processing time on the rate of nitriding and the microstructure was investigated. A high nitriding rate of 3 mm2 ys and a microhardness value of 1800 HV0.1 were achieved under optimum plasma conditions. The interpretation is based on a new approach that the microcracks in the surface, at a certain ambient temperature and plasma, can create traps for the active nitrogen species. The nitrogen concentration and temperature gradient can also draw some of the nitrogen species towards the bulk side. The microhardness can be explained in terms of the high concentration of nitrided phases created under the surface of the treated samples. 䊚 2002 Elsevier Science B.V. All rights reserved. Keywords: Rf plasma; Nitriding process; 304 Stainless steel; Surface microcracks; Temperature gradient

1. Introduction The surface hardening of materials by nitriding has become an important commercial process. Through nitriding, very hard surface layers are obtained without substantial modification of the bulk material properties. For many decades, various methods based on plasma have been used for nitriding different materials w1–7x. These techniques improve the quality of nitrided surfaces with respect to traditional nitriding methods using gases containing ammonia w8,9x. Recently, rf plasmas have been used for nitriding different materials, including steels and stainless steels w10–13x. Rf plasma nitriding has many advantages over previous methods, including shorter treatment times (minutes instead of hours) and improved mechanical properties (surface hardness can be increased, in some cases, up to nine times) w10,11x. The mechanism of nitriding using plasma-based methods has been explained in contradictory ways. Ionic bombardment w14–16x and neutral atomic and molecular absorption w17x are the main models that have been explained for plasma nitriding mechanism. Related to this work, a very high rate of nitriding and a * Tel.: q20-25-437852; fax: q20-93-605745. E-mail address: [email protected] (F.M. El-Hossary).

hard compound layer were achieved for 304 austenitic stainless steel using an inductively coupled rf plasma w10,11,18x. Here, an approach is made in an attempt to clarify the reasons for the high microhardness values and the high rate of nitriding. 2. Experimental work All the samples of austenitic stainless steel nitrided conformed to the specification AISI 304. They were in the form of ;1-mm-thick rolled sheets, cut into coupons of different sizes. The plasma reactor was a quartz tube, 550 mm in length and 50 mm in diameter evacuated to a base pressure of order 10y3 torr. High purity oxygen free nitrogen was introduced, using a thermal mass flow controller and needle valve, at a rate of 1.2"0.1 ml miny1 adjusted to establish a gas pressure of order 0.06 torr as measured by a capacitance manometer. No special precaution was taken to remove all traces of hydrogen, which was present in the reactor tube. The discharge was generated by a copper induction coil energized from an rf (13.56 MHz) power supply through a tunable matching network. Samples were supported on a 45mm-diameter movable water-cooled platform. A thermocouple close to the sample holder provided us with a coarse value of the sample temperature.

0257-8972/02/$ - see front matter 䊚 2002 Elsevier Science B.V. All rights reserved. PII: S 0 2 5 7 - 8 9 7 2 Ž 0 1 . 0 1 5 2 4 - 9

278

F.M. El-Hossary / Surface and Coatings Technology 150 (2002) 277–281

Fig. 1. Square thickness of nitrided compound layer per time mm2ys vs. plasma nitriding time (s) for a plasma processing power of 520 W.

Two groups of austenitic stainless steel were nitrided. The first group was treated using a plasma processing power of 520 W for different plasma processing times 3, 4, 5, 6, 7, 10, 15, 18, 25 and 30 min. The other one was nitrided using different plasma powers 370, 400, 425, 450, 475, 500, 550, 600, 650 and 750 W for a fixed plasma processing time of 10 min. Cross sections of the above samples were made, mounted in bakelite, ground, polished and etched using 2% nital for 30 s. Cross section morphologies were recorded using an Epiphot 300 inverted metallurgical microscope. The thickness of the nitride layer was measured using the

microfilar eyepiece attached to the microscope. Microhardness measurements were made on the nitrided sections, as well as perpendicular to the nitrided surface, using a Letiz Durimet tester. X-Ray diffraction measurements were made using a Scintag Inc. unit with CuKa radiation. 3. Results Fig. 1 shows the relationship between the rate of nitriding in square micrometers per second and nitriding reaction time in minutes for 304 austenitic stainless steel

Fig. 2. Square thickness of nitrided compound layer per time mm2ys vs. plasma nitriding power (W) for a plasma processing time of 10 min.

F.M. El-Hossary / Surface and Coatings Technology 150 (2002) 277–281

Fig. 3. Microstructure cross-section of compound layer nitrided at 520 W for different plasma processing times: (a) 3, 5, 6 and 7 min and (b) 10, 15, 25 and 30 min.

samples nitrided at 520 W. The highest nitriding rate of 3 mm2 ys was achieved for the sample that was treated for 7 min. Even for the short plasma processing time of 3 min, the rate is still very high (0.3 mm2 ys) relative to most nitriding plasma methods. Some other experimental data concerning nitriding rate are given here for sake of comparison: a 0.6-mm2 ys rate reported by Lebrun et al. w19x and the 0.4-mm2 ys rate achieved by Billon and Hendry w9x. Metin and Inal, who nitrided pure iron using a d.c. discharge of mixture of nitrogen and hydrogen gases, reported a rate of only 0.028 mm2 ys w20x. Chung and Lim achieved a much higher rate of 0.7 mm2 ys, for 304 stainless steel w21x. The effect of rf plasma processing power on the rate of nitriding is shown in Fig. 2, where the nitriding plasma processing time was fixed for 10 min. At the medium nitriding plasma processing power of 500 W, a maximum nitriding rate of 2.67 mm2 ys was achieved. The variation of surface temperature and microhard-

279

ness as a result of nitriding using different rf plasma processing times or plasma powers have been reported elsewhere w18,22x. The samples nitrided for 4 min or up to 30 min at 520 W have microhardness value equal to 1750 HV0.1 in average. More or less the same value of microhardness was achieved for samples nitrided between 425 and 600 W for 10 min. Baldwin et al. w5x reported nearly 1500 HV for 316 stainless steel nitrided using rf plasma. Lei and Zhang w23x also achieved 1500 HV using microwave plasma source ion nitriding. We have achieved, under certain plasma parameters, a linear relationship between the surface temperature and the input plasma power w18x. A value of 3808C was recorded for the surface temperature when the plasma power was close to 375 W. As the plasma power increased to 750 W, the surface temperature increased to 6508C. It took up to 3–4 min for the plasma operation to reach maximum temperature. When the input plasma power was equal to 520 W, the surface temperature reached 5508C after 3 min. Fig. 3 shows the cross-section morphology of 304 stainless steel nitrided at 520 W for different plasma processing times. In Fig. 3a, the nitriding time was varied from 3 to 5, 6 and 7 min for each separate sample. Moreover, Fig. 3b shows four more samples, which were nitrided for longer plasma processing times: 10, 15, 25 and 30 min. In the first group of samples, one can find that the nitrogen penetrated through the surface and reacted with the bulk material at different depths and locations. With more careful inspection, one can also see that there are two different concentrations of nitride locations (more obvious in the sample which was nitrided for 3 min). The first has dark areas and is distributed randomly at different locations under the surface. The other, light gray and barely visible, begins just under the surface and ends at the interface with the bulk material; it has a regular distribution. When the nitriding plasma processing time was increased, the thickness and concentration of nitride locations increased. Up to 7 min, the light gray areas just under the surface can be seen, but at 10 min or longer nitriding time, they become narrow and dark. Fig. 4 shows the cross-section morphology of 304 stainless steel nitrided for 10 min at different plasma processing powers. At very low plasma power, 370 W, one can hardly see the narrow nitrided layer. However, at 400 W, the depth of the nitrided layer is clearly seen. One can also see, similar to the sample nitrided for 3 min shown in the previous figure, that the two locations of different concentrations of nitrides are present. When the applied plasma power was increased to 475 W, the thickness and concentration of nitrided layer increased sharply, as shown in Fig. 4. This trend continues for the sample that was nitrided at 500 W. With a further increase in the applied plasma power to 550 W, the light gray layer becomes very

280

F.M. El-Hossary / Surface and Coatings Technology 150 (2002) 277–281

Fig. 4. Microstructure cross-section of compound layer treated for 10 min at different plasma powers: 370, 400, 475, 500, 520, 600 and 750 W.

narrow and the thickness of the nitrided layer, as a whole, becomes less than in the sample, treated at 500 W. At 650 W, the light gray layer vanishes and the other nitride layer decreases. When the applied plasma power was increased to 750 W, the thickness of the nitride layer became very small, 1.9 mm. We have found, from X-ray structure data w18,22x, that the nitride layer formed in samples nitrided at low plasma power (370–400 W) or short plasma processing time (3–5 min) have FeNiN and ´-Fe2 – 3N phases. At medium plasma power (425–500 W) or plasma time (6–10 min) the former phase vanishes and the g9-Fe4N phase is clearly detectable. At higher plasma powers (520–750 W) or longer plasma times (15–30 min), ´Fe2 – 3N vanishes, g9-Fe4N is still detectable and very strong peaks which were identified to be CrN appeared. There is also an indication of g-austenite and a-ferrite phases present in the spectra. 4. Discussion The above results suggest the following interpretation of high rate of nitriding and high values of microhardness for treated 304 stainless steel.

At a certain level of temperature, 500–6008C and under the plasma ambience, the 304 austenitic stainless steel tends to have a large volume of microcracks. Active nitrogen atoms, molecules and ions penetrate through these microcracks and react with surrounding constituents of the sample to form nitride phases. The interfaces between regions of nitride and bulk materials have a relatively large area compared with that of the interface between the plasma and the sample surface. Nitrogen, therefore, diffuses at a high rate in all directions between the regions of nitride and the bulk material. At the same time, the active nitrogen reacts with the overall sample surface forming a uniform light gray nitrided layer in the top surface. As a result of the concentration gradient, the nitrogen diffuses uniformly towards the bulk. When the plasma time increases to 7 min or longer, the non-uniform nitrided layer forms a continuous band over the sample. However, there is still a gap, a light uniform nitrided layer between the top surface and a heavily nitrided band (compound layer) exists. This gap decreases as the plasma time increases. The above statement also applies to the other group which was nitrided for 10 min at different plasma

F.M. El-Hossary / Surface and Coatings Technology 150 (2002) 277–281

processing powers up to 500 W. Above this value, the temperature of the sample increases to reach more than 6008C in which the surface is nearly free of microcracks and so the nitrogen diffuses at a normal rate. Another factor has to be considered: the high difference in temperature through the sample. From the thickness of the sample, 1 mm, and difference in the temperature of approximately 500 K or more, the temperature gradient could be 5=105 Kym. According to the Thomson effect, this will lead to a potential difference between the two levels of the sample. This potential, apart from the floating potential, might attract more ions from the surface towards the base side. The X-ray spectra show wide peaks due to enormous stresses and the morphology of the cross-section demonstrates a highly-condensed nitrided layer in which a relatively high microhardness layer was formed. 5. Conclusion Nitriding using an inductively coupled rf plasma is a powerful technique to achieve a very efficient nitriding process for austenitic stainless steel. The volume microcracks, temperature gradient and nitrogen concentration gradient could be the main parameters that affect the rate of nitriding and the microhardness value. References w1x E.J. Miola, S.D. De Souza, M. Olzon-Dionysio, et al., Mater. Sci. Enf. A256 (1998) 60.

281

w2x B. Edenhofer, Metall. Mater. Technol. 7 (1976) 421. w3x S. Parascandola, O. Kruse, W. Moller, Appl. Phys. Lett. 75 (13) (1999) 1851. w4x M.J. Baldwin, M.P. Fewell, S.C. Haydon, et al., Surf. Coat. Technol. 98 (1998) 1187. w5x M.J. Baldwin, G.A. Collinss, M.P. Fewell, et al., Jpn. J. Appl. Phys. 36 (1997) 4941. w6x D.L. Williamson, J.A. Davis, P.J. Wilbur, Surf. Coat. Technol. 103 – 104 (1 – 3) (1998) 178. w7x T. Czerwiec, N. Renevier, H. Michel, Surf. Coat. Technol. 131 (1 – 3) (2000) 267. w8x B. Billon, A. Hendry, Surf. Eng. 1 (1985) 114. w9x B. Billon, A. Hendry, Surf. Eng. 1 (1985) 125. w10x F. El-Hossary, F. Mohammed, A. Hendry, D.J. Fabian, Z. Szaszne-csih, Surf. Eng. 4 (1988) 150. w11x F. El-Hossary, J. Mater. Sci. Lett. 11 (1992) 1375. w12x J.M. Priest, M.J. Baldwin, M.P. Fewell, et al., Thin Solid Films 345 (1999) 113. w13x M.J. Baldwin, M.P. Fewell, S.C. Haydon, et al., Surf. Coat. Technol. 98 (1998) 1187. w14x M. Hudis, J. Appl. Phys. 44 (1973) 1489. w15x S. Parascandola, R. Gunzel, R. Grotzschel, W. Moller, Nucl. Instrum. Meth. Phys. Res. B 136 – 138 (1998) 1281. w16x A. Brokman, F.R. Tuler, J. Appl. Phys. 21 (1981) 468. w17x G. Tibbetts, J. Appl. Phys. 45 (1974) 5072. w18x F.M. El-Hossary, Surf. Eng. 16 (6) (2000) 491. w19x J.P. Lebrun, H. Michel, M. Gantois, Mem. Etud. Sci. Rev. Metall. 69 (1972) 727. w20x E. Metin, O.T. Inal, J. Mater. Sci. 22 (1987) 2783. w21x M.F. Chung, Y.K. Lim, Scripta Metall. 20 (1988) 807. w22x M.M. Ibrahim, F.M. El-Hossary, N.Z. Negm, M. Abed, R.E. Ricker, Appl. Surf. Sci. 59 (1992) 253. w23x M.K. Lei, Z.L. Zhang, J. Vac. Sci. Technol. A 13 (6) (1995) 2986.