Wear, 89 (1983) 201 - 214
FRICTION AND WEAR PROPERTIES OF NITRIDED IMPLANTED 17-4 PH STAINLESS STEEL
G. M. ECER Sii Smith
Center, Irvine, CA 92714
SUSAN WOOD, D. BOES and J. SCHREURS Westinghouse
Center, Pittsburgh~ PA 15235 (U.S.A.)
(Received January 31,1983)
Summary N* implantation of thermally nitrided 17-4 PH stainless steel substantially improved friction and wear properties in lubricated sliding wear tests. Some improvement was also obtained by nitriding or by N+ implantation alone. Auger and surface topography analyses of specimens before wear testing indicated that the improved wear performance may be due in part to the implantation of carbon, and perhaps oxygen, when N+ was being implanted under a relatively high target chamber pressure and in part by the topographical alteration of the surface in thermal nitriding.
1. Introduction In recent years it has been demonstrated that surface modification by high energy ions (10 - 200 keV) can produce remarkable changes in the properties of metals. One area in which much improvement has been obtained is in friction and wear properties of metals. For example, under lubricated sliding .wear test conditions, the wear rates of types 416 and 304 stainless steels and M43 tool steel have been significantly reduced (by over two orders of magnitude) by implanting one of the surfaces with species such as nitrogen, carbon and titanium [l]. In other studies, the friction coefficient on steel surfaces was significantly reduced  by implanting Sn+, and the implantation of N”, C+ and Bf has been found to be effective  in reducing wear in carbon steel, mild steel, a nitriding alloy, a stainless steel and a hardenable tool steel. These results and others, in areas of fatigue [3 - 53, cavitation erosion  and rolling contact fatigue performance 171, indicate that ion implantation can become a viable surface treatment process, especially since the technique allows finished products to be treated at near room temperature without distortion. Elsevier Sequoia/Printed in The Netheriands
It has been suggested [Z, 7] that the improved tribological properties are a result of both the compressive stresses in the surface layers which contain the injected foreign atoms and the lattice defects formed in implantation. In nitrogen-implanted steels improved wear resistance persisted a long way beyond the region of ion implantation [ 8, 91. This remarkable finding was explained by a process similar to strain aging. Interstitial species such as nitrogen would form a Cottrell cloud decorating dislocations. Thus nitrogen (or carbon) atoms not only would inhibit the motion of dislocations but also could move inwards, with dislocations, as wear proceeds. Another experimental finding which bears practical importance is that reported by Dearnaley and Hartley  on the improved wear resistance obtained by Ni implantation of a gas-nitrided steel. The present investigation was conducted to explore further such synergistic effects on an alloy, 17-4 PH stainless steel, that is not fully responsive to thermal nitriding treatments but possesses good corrosion resistance which is desired in certain industrial applications.
2.1. Surface treutmen
Friction and wear test specimens were machined to the dimensions shown in Fig. 1 and were from solution-treated bar stock (nominal chemistry, Fe-17Cr-4Ni-4Cu-lMn-lSi-O.3Nb -t Ta). Some of the specimens were gas nitrided as follows: they were held for 1 h at 524 “C in dissociated ammonia, then held for 10 h at the same temperature in 30% dissociated ammonia and cooled in a cooling chamber in dissociated ammonia to below 150 “C. This yielded a calculated nitrided depth of about 25 pm. Some of the specimens in untreated or nitrided conditions were ion implanted using two different implanters. One, identified in the text as implanter A, was a modified model 200~CS5 by Varian-Extrion and the other, identified as implanter B, was a modified High Voltage Engineering LS-4 accelerator. The
2.4 mm Dia.ROTATING CYLINDER
Fig. 1. Friction and wear test arrangement and specimen dimensions.
specific details of the ~plan~tion parameters are given in Table 1. (It should be noted that the beam was rastered in implanter A.) TABLE 1 Nitrogen implantation parameters Implanter Fluence (N+ ions cmm2) Ion Beam current &A) Current density @A cm-2) Beam voltage (keV) Beam coverage (cm2) Implantation time (s) Target chamber pressure (Pa) Maximum target temperature (estimated) (“C)
1 x 101’ N+ 500 5.5 100 91 2919 (7 - 8) x 10-4 <50
14s 150 1.77 2268 (3 - 7) x10-s <35
8 x 10” NZ+ 25 14.1 150 1.77 9072 (3 - 7) x 10-s <35
Specimens utilizing implanter A were glued onto the aluminum target plate using a colloidal silver paint. The paint binder was evaporated by subjecting the ~angement to a partial vacuum as low as 1O-2 Pa for 16 h. Specimens utilizing impfanter I3 were mechanically clamped to achieve a close contact with the target plate. 2.2. Friction and wear tests Sliding friction and wear tests were conducted on untreated, thermally nitrided, N+-implanted and thermally nitrided plus N+-implanted 17-4 PH stainless steel specimens. The stationary disc, shown in Fig. 1, was fixed in a spring-floated solid copper base cylinder. Inside the copper cylinder a thermocouple could be placed to within 0.5 cm of the stationary disc. A chuck arrangement allowed the rotating cylinder specimen to be pressed down against the disc specimen (race) under a load FN which could be varied at will. A transducer attached to the copper base measured the frictional force FF which was continu~y recorded. All tests were run under lubricated conditions using a standard petroleum lubricant (Chevron BRB-2SRI: Saybolt Universal viscosity, 500 s) containing a polyurea thickener and additives for improved oxidation stability and antiwear characteristics. Each specimen couple was lubricated once, at the beginning of the test, which usually ran for 30 min. All tests were run by rotating the cylinder-shaped specimen at a speed of 1000 rev min-‘. Wear debris was collected after completion of the wear tests and analyzed by X-ray diffraction. The wear specimens were ultrasonically cleaned in acetone, before and after the tests, and mass changes recorded. In this type of wear testing arrangement, the load FN on the specimen is related to the pressure P applied on the race by
F, = Pn( r,2 - ri2) where r, and ri are the outer and inner radii of the rotating friction coefficient is
FN where F, is the friction force corrected for the apparatus and the specimens described. In addition to measuring the wear damage by the mass change, the specimens were examined by scanning electron microscopy (SEM) and a surface profilometer (Clevite Surfanalyzer 360). Surface profiles of a number of the stationary disc specimens were measured by traversing at least two diameters. 2.3. Transmission electron microscopy Samples for transmission electron microscopy (TEM) were prepared for ion implantation from the same starting material as that used in wear tests. Wafers, 0.6 mm in thickness, were spark cut from the 17-4 PH stainless steel bar stock, and discs 3 mm in diameter were punched out of the wafers. The discs were reduced to a thickness of 0.2 mm by grinding on silicon carbide papers and polishing through Linde B. All faces to be ion bombarded were dimpled using a solution of 80% acetic acid and 20% perchloric acid (PACE) at room temperature. Such a sequence is necessary to remove all traces of mechanical deformation in the sample surface. Irradiated specimens were further reduced to a thickness of 0.075 mm before jet thinning for TEM analysis. Jet polishing was carried out in PACE at room temperature. Microscopy was performed in a Philips EM-300 transmission electron microscope at 100 kV. 2.4. Auger analysis The chemical compositions of the near-surface regions of the 17-4 PH stainless steel specimens were determined by Auger electron spectroscopy (AES) in conjunction with sputtering. Both untreated surfaces and surfaces treated with nitrogen were analyzed. Areas inside and outside the wear track were compared with respect to the implanted surfaces. The analysis was performed with a scanning Auger microprobe with a 5 keV primary electron beam and a modulation voltage of 3 V for the cylindrical mirror analyzer. The primary beam current was 1 - 2 I.IA and the beam was rastered over an area 20 pm X 40 E.trn in size at television scanning rates. Survey spectra were recorded for Auger electrons with energies between 0 and 1000 eV in the conventional derivative mode. An argon ion beam was used to mill the surface at a rate of 8 nm min-’ (implanter A) or 5 nm min-’ (implanter B). To produce the 2 kV Ar+ beam, an ultrahigh vacuum chamber was backfilled with a high purity (4 X lop3 Pa (3 X lo-’ Torr)) argon atmosphere over a base pressure of less than 4 X lo-’ Pa (less
than 3 X lo-’ Torr). Spectra were taken before ion milling, after 30 s and 1 min of milling and at intervals of 1 min thereafter. A typical analysis of an area involved the removal of a layer 200 nm thick. The argon atmosphere was present during the AES analyses, but the ion gun was shut off. Concentrations were calculated using the method of normalized peak heights [lo]. This analysis procedure has been found to yield Auger concentration profiles that agree well with profiles predicted from theoretical calculations [ 111 for nitrogen implanted in a variety of steels.
3. Results and discussion The mass loss results, summarized in Figs. 2 and 3 for rotating cylinder and stationary disc specimens respectively, and the wear track contours determined by profilometry, typical examples of which are shown in Fig. 4, reflect the individual and synergistic effects of the surface treatments given. In general, thermal nitriding and N+ implantation (to a fluence of 1 X 10” N+ ions cmm2 in implanter A) individually reduced the wear rate. The lowest wear rate occurred when the specimen couples had been thermally nitrided and subsequently implanted with N+ ions. Similar results were observed in
Fig. 2. Mass loss of 17-4 PH stainless steel (rotating cylinder specimens) after wear testing for 30 min (lubricant, Chevron BRB-2-SRI; fluence, 1 x 10” N+ ions cme2).
Fig. 3. Mass loss of 17-4 PH stainless steel (stationary disc specimens) after wear testing for 30 min (lubricant, Chevron BRB-2SRI; fluence, 1 x 10” N+ ions cm-‘).
Fig. 4. Typical profilometer traces of wear tracks in 17-4 PH stainless steel (stationary disc specimens) tested under a pressure of 2.87 MPa: trace A, thermally nitrided and ion implanted with I X 1Or8 N+ ions cmU2; trace B, thermally nitrided; trace C, ion implanted with 1 x 101’ N* ions cm-‘; trace D, untreated.
the friction behavior, which is summarized in Fig. 5 where the friction force variation is presented as a function of time for various specimens and load conditions. Each column represents one surface condition and each row represents a different load. Two specimens were utilized for some conditions (e.g. nitrided specimens with F, = 229 N), In general, the nitrided and N”-
FN = 273N
loo 0 0
a0 10 Time. mln
Fig. 5. Variation in friction force with test duration for 17-4 PH stainless steel specimen couples. In any given row the load is constant and in any given column the treatment is constant, The specimen designations are given for each load-treatment combination.
implanted condition shows a reduction in the frictional force compared with the untreated surface. Erratic variations in the frictional force are also largely absent for the former surface condition, and ~pro~emen~ are more pronounced at higher loads. Surfaces which were nitrogen implanted without being nitrided (third column) show some improved friction behavior at lower loads, whereas the nitrided surfaces (second column) behave very similarly to the untreated specimens. The initial friction coefficient of untreated specimens was reduced considerably by all surface treatments when tests were conducted under high specimen loads. Larger fluences of N+, implanted under a much lower beam current (Table 1) in implanter B, produced unexpectedly high wear as shown in Table 2. Although three of the implanted discs implanted at large Nf fluences (Table 2) were run against untreated cylinders, one was run against an implanted cylinder. These results indicated that the wear mechanism operating in N+-implanted 17-4 PH stainless steel specimens may have been influenced by factors other than the N+ dose.
Implanted with 8 x 1017 N’ ions cme2 11.0 16.3 32.7
17-4 PH stainless steel disc
Test durationa (min)
Mass losses b (mg) Cylinder Disc
with N+ in implanter
Implanted with 2 X 1Or7 N+ ions cmW2
All tests were conducted under a specimen load of 229 N using Chevron BRB-P-SRI petroleum aTests were stopped because of excessive chatter. bMass losses have been extrapolated to 30 min for easy comparison with other results.
steel disc specimens
17-4 PH stainless steel cylinder
Wear test results for the 17-4 PH stainless
B (Table 1)
Implanted with 2 x 10z7 N+ ions cmW2
Implanted with 1 x 1017 N+ ions crne2
3.1. Electron microscopy A brief TEM evaluation was undertaken to compare microstructures before and after the N+ implantation of 17-4 PH stainless steel foils in implanter B. It was apparent that the implantation of nitrogen produced some changes in the structure, but a complete understanding of these changes was not achieved. The microstructure before implantation was that of a typically complex 17-4 PH stainless steel showing a martensitic matrix with some possible retained austenite, carbides and &ferrite. Twinned martensite, not typically observed in 17-4 PH stainless steel, was found in one of the two foils implanted to a fluence of 2 X 10” N+ ions cm-?. The second foil and another bombarded to a fluence of 8 X lOi’ N+ cme2 were nearly free of martensite twinning. Thus the evidence linking twinning in martensitic platelets to N+ implantation remains inconclusive. While no nitrides of significant size (greater than or equal to 30 nm) were observed in the foils with a low N+ dose, the foil with a higher N+ dose contained triangularly shaped particles (Fig. 6) which were tentatively identified as nitrides. The absence of martensite twinning in the high dosage foil may have been a result of the development of a different stress state at the surface, a hypothesis which cannot, as yet, be confirmed since techniques for the measurement of stress in such thin surface layers are unavailable. An SEM survey of the surface topography of disc specimens before wear testing revealed that there was a clear difference between thermally nitrided specimens and all the others. The nitriding process had apparently broken up the distinct machining grooves found on other specimens, as shown in Fig. 7. Surface roughness traces, seen in Fig. 8, indicate that nitriding.has also increased the roughness. Ion implantation, by contrast, had little effect on the surface topography as observed by SEM. However, some surface smoothing may have occurred, as seen in Fig. 8, traces A and B. The nitrided specimens showed less wear than the untreated or N+implanted specimens. This may be explained by the difference in surface
Fig. 6. Particles tentatively identified as nitrides observed in 17-4 PH stainless steel after implantation with 8 X 10” N+ ions cm-’ (75 kV).
Fig. 7. SEM micrographs showing typical surface topographies observed before the testing of (a) an untreated specimen, (b) a specimen treated by ion implantation (1 x 1Or’ N+ ions cmm2) and (c) a specimen treated by thermal nitriding and ion implantation (1 X 1017 N+ ions cme2 ).
Fig. 8. Typical surface roughness traces taken before testing of 17-4 PH stainless steel specimens: trace A, untreated; trace B, ion implanted with 1 X 10” N+ ions cm-*; trace C, thermally nitrided; trace D, thermally nitrided and ion implanted with 1 X 1017 N+ ions cmp2.
chemistries and by the possibility that a more compartmentalized surface produced by nitriding (Fig. 7(c)) may provide a better support for the lubricant.
3.2. Surface chemistry and composition X-ray diffraction analyses of the wear debris showed no evidence of oxide formation in any specimen tested except for one with an untreated surface which was tested at the highest load. The other diffraction patterns yielded only lines due to the martensitic matrix. N+ implantation of a nitrided specimen did not increase the nitrogen concentration near the specimen surface as efficiently as was expected (Fig. 9). This was partially attributed to the sputtering of the nitrogen-rich nitrided surface but mainly to the high concentrations of carbon and oxygen found in nitrided and nitrided plus N+-implanted specimens, as shown in Figs. 10 and 11. Both the nitriding process and the N” implantation in implanter A were identified as sources of additional carbon, some of which was recoil implanted into the metal by the nitrogen bombardment. The source of the excess oxygen was also attributed to the nitriding process which formed a thin oxide film on the specimens, presumably because of exposure to the atmosphere before cooling was complete. Subsequent implantation of nitrogen did, in fact, cause the oxygen to recoil into the metal matrix as shown by the Auger profiles in Fig. 11. In contrast, little carbon and oxygen were found in specimens implanted in implanter B where the target chamber pressure during nitrogen implantation was over an order of magnitude lower than that established in implanter A (see Table 1). Auger concentration profiles for nitrogen, carbon and oxygen for two specimens implanted in implanter B to fluences of 2 X 10” N+ ions cmW2 and 8 X lo7 N+ ions cme2 respectively are shown in Figs. 12 and 13. These findings suggest that the superior wear performance of the nitrided plus N+-implanted specimens could also result from the increase in
Fig. 9. Nitrogen concentration profiles determined by Auger electron spectroscopy (approximate sputtering rate, 8 nm min-‘): 0, untreated; A, ion implanted with 1 x 10” N+ ions cme2 (100 keV); 0, thermally nitrided; A, thermally nitrided and ion implanted with 1 X 10” N+ ions cm-2 (100 keV).
I 16 SPUTTERING
32 TIME, MIN
Fig. 10. Carbon concentration profiles determined by Auger electron (approximate sputtering rate, 8 nm min-‘): symbols as for Fig. 9.
Fig. 11. Oxygen concentration profiles determined by Auger electron (approximate sputtering rate, 8 nm min-I): symbols as for Fig. 9.
carbon (and oxygen) at or near the surface in addition to the topographic differences discussed earlier. The high nitrogen concentrations introduced into specimens implanted in implanter B (Figs. 12 and 13) seem to have made little difference to the wear performance since all these specimens wore at a higher rate (Table 2) than the lower dose specimens prepared in implanter A. The possibility that the high N+ doses of the order of (2 - 8) X 10” N+ ions cm-’ may have resulted in early fractures of the surface layers during the wear process, leading to an accelerated wear rate, should also be considered. Although this may explain the poor performance of the specimens with high doses of implanted nitrogen, it does not explain the improvement in wear resistance obtained by N+ implantation of nitrided specimens or the lower dose N+-implanted samples (see Fig. 2). For the latter the main effect of N+ implantation was, as can be seen in Figs. 10 and 11, an increase in the amounts of carbon and oxygen. Indeed, without an appreciable change in the maximum nitrogen concentration (Fig. 9), carbon was apparently recoil implanted into the steel. Auger analysis of the wear tracks showed very little nitrogen in implanted specimens after the wear removal of a layer of the order of 5000 nm in thickness.
15 24 32 SPUTTERING TIME, MIN
I I I 32 16 24 SPUTTERING TIME. MIN
Fig. 12. Depth distributions of nitrogen (@), carbon (A) and oxygen (m) in a 17-4 PH stainless steel specimen ion implanted to a fluence of 2 x 10” N+ ions cm-* (approximate sputtering rate, 5 nm min-‘). Fig. 13. Depth distributions of nitrogen (@), carbon (A) and oxygen (m) in a 17-4 PH stainless steel specimen after ion implantation to a fluence of 8 x 10” N+ ions cm-* (approximate sputtering rate, 5 nm min-‘).
4. Conclusions Lubricated sliding wear tests were conducted using untreated, thermally nitrided, N+-implanted and nitrided plus N+-implanted 17-4 PH stainless steel specimens. The mass loss due to wear was reduced both by low dose N+ implantation and by thermal nitriding. However, the best results were obtained when the thermally nitrided specimens were N+ implanted. In these specimens the wear damage, as measured by profilometry and by mass loss, was reduced substantially (by up to two orders of magnitude). Auger analyses conducted in conjunction with ion sputtering indicated that the apparent synergistic effects of nitriding and N+ implantation were, at least in part, due to the high degree of surface contamination with carbon and possibly oxygen during implantation. Carbon and oxygen partially originated from the thermal nitriding process. Some carbon was also deposited and recoil implanted during the N+ implantation when the target chamber pres-
sure was relatively high. Changes in surface topography on thermal nitriding also may have contributed to the improved friction and wear performance of nitrided specimens.
Acknowledgments The authors would like to thank the following individuals for their contributions in the areas indicated: R. Kelecave (friction and wear testing), W. J. Choyke and N. J. Doyle (implantation), J. J. Haugh (microscopy), T. Mullen (SEM) and D. Detar (AES).
References 1 J. K. Hirvonen, C. A. Carosella, R. A. Kant, I. Singer, R. Vardiman and B. B. Rath, Improvement of metal properties by ion implantation, Thin Solid Films, 63 (1979) 5 - 10. 2 N. E. W. Hartley, Ion implantation and surface modification in tribology, Wear, 34 (1975) 427 - 438. 3 G. Dearnaley and N. E. W. Hartley, Ion implantation into metals and carbides, Thin Solid Films, 54 (1978) 215 - 232, 4 G. Dearnaley, Ion implantation for improved resistance to wear and corrosion, Mater. Eng. A&., 1 (1978) 28. 5 W. W. Hu, C. R. Clayton, H. Herman and J. K. Hirvonen, Fatigue life enhancement by ion implantation, Ser. Metall., 12 (1978) 697. 6 H. Herman, W. W. Hu, C. R. Clayton, J. K. Hirvonen, R. Kant and R. K. MacCrone, Modification of mechanical properties through ion implantation, Proc. Conf. on Ion Plating and Allied Techniques, London, CEP Consultants, Edinburgh, 1979, p. 255. on rolling contact 7 G. White and G. Dearnaley, The influence of Nz + ion implantation fatigue performance, Wear, 64 (1980) 327 - 332. 8 G. Dearnaley, Practical applications of ion implantation. In C. M. Preece and J. K. Hirvonen (eds.), Ion Implantation Metallurgy, Metallurgical Society of AIME, New York, 1980, p. 1. 9 W. W. Hu, H. Herman, C. R. Clayton, J. Kozubowski, R. A. Kant, J. K. Hirvonen and R. K. MacCrone, Surface related mechanical properties of nitrogen-implanted 1018 steel. In C. M. Preece and J. K. Hirvonen (eds.), Ion Implantation Metallurgy, Metallurgical Society of AIME, New York, 1980, p. 92. 10 L. E. Davis (ed.), Handbook of Auger Electron Spectroscopy, Physical Electronics Industries, Eden Prairie, MN, 2nd edn., 1976. 11 M. Baron, A. L. Chang, J. Schreurs and R. Kossowsky, Nitrogen distribution and nitride precipitation in nitrogen implanted 304 and 316 steels, Nucl. Znstrum. Methods, 182 - 183 (1981) 531 - 538.