Diamond-like carbon film deposition on plasma nitrided steel substrates

Diamond-like carbon film deposition on plasma nitrided steel substrates

1112 Diamond and Related Materials, 3 (1994) 1112 1116 Diamond-like carbon film deposition on plasma nitrided steel substrates* F. Cellier a n d J.F...

497KB Sizes 2 Downloads 20 Views


Diamond and Related Materials, 3 (1994) 1112 1116

Diamond-like carbon film deposition on plasma nitrided steel substrates* F. Cellier a n d J.F. N o w a k 1RSID (USINOR-SACILOR), Centre de Recherche d'Unieux, Pont du Sauze, Z1 du Pare, BP 01-42490 Fraisses (France)

(Received September 24, 1993; accepted in final form November 23, 1993)

Abstract Diamond-like carbon (DLC) films are very hard and wear resistant, and are generally characterized by low friction coefficients. Nevertheless, their adhesion to steel substrates is limited due to the large difference in physical properties (i.e. hardness) of the materials. In contrast, plasma nitriding of steel gives hardened layers which are fatigue- and wear- resistant, but which often display rather poor friction properties. We describe in this paper a combined plasma nitriding treatment/d.c.-PACVD process giving multicomponent coatings with gradually varying mechanical properties and improved wear resistance and friction characteristics. The influence of the nitrided layers on the mechanical behaviour of deposited DLC films is studied by means of scratch test methods and hardness measurements. Ball-on-disc and ring-on-block experiments are used to evaluate the usefulness of such a duplex process with regard to tribological properties of the treated steel surfaces.

1. Introduction Surface hardening of steel by means of plasma nitriding gives treated layers which are fatigue-, wear- and plastic deformation-resistant but which are not corrosionresistant and which often display rather poor friction properties. In contrast, diamond-like carbon (DLC) films display desirable properties such as hardness, wear resistance and low friction coefficients, and are chemically inert [ 1-4]. Such films could find numerous applications in the fields of mechanical engineering and material forming tools. Nevertheless, their adhesion to steel substrates is limited due to the large difference in the physical properties (notably the hardness) of the materials. The trend is now clearly towards improving the performance of hard coatings like TiN [5-7] or CrN [8] using thermochemical treatments to strengthen the substrate materials. We show in this paper that a combined plasma nitriding treatment/d.c.-PACVD (direct current plasma assisted chemical vapour deposition) process gives a duplex structure with improved mechanical and tribological properties. The influence of the nitrided layers on the mechanical behaviour of deposited DLC films is studied by means of scratch test methods and hardness measurements. Ball-on-disc experiments are used to evaluate the usefulness of such a duplex process with regard to the tribological properties of the nitrided steel surfaces. Moreover, we compare the behaviour of DLC films with other tribological coatings by *Paper presented at Diamond Films '93, Albufeira, September 20-24, 1993.

0925-9635/94/$7.00 SSDI 0 9 2 5 - 9 6 3 5 ( 9 3 ) 0 0 1 6 9 - E

assessing the "seizure threshold" under lubricated conditions, using ring-on-block experiments.

2. Experimental details The substrate material used in the present investigation was a low alloyed, general-purpose construction steel, AISI 4135. The material was hardened and tempered to an initial hardness of about 42 HRC. The samples were in the form of discs 45 mm in diameter and 5 mm thick. The chemical analysis of the billet from which samples were machined is shown in Table 1. DLC films were deposited consecutively to the plasma nitriding treatment in the same cold-wall, primary pumped reactor by simply varying gas composition and plasma discharge parameters [-9]. Electrical power was TABLE 1. Chemical analysis of AISI 4135 steel Element

Quantity (wt.%)

C Si Mn P S Cr Ni Mo AI Cu Ti

0.335 0.298 0.771 0.014 0.004 1.147 0.226 0.201 0.02 0.317 < 0.01

© 1994 - - Elsevier Sequoia. All rights reserved

F. Cellier, J.F. Nowak

Diamond-like carbon .film deposition on steel substrates

supplied to the plasma discharge in a diode mode by an 8 kW d.c. generator. The introduction and mixture of the gas were controlled by mass flow meters. Two series of samples were produced in the present work, the details of which are given in Table 2. Plasma nitriding conditions were adjusted to obtain, on AISI 4135 steel, either a diffusion nitrided case only (series I) or an inner diffusion nitrided case with an outer case mainly constituted of iron nitride compound F e 4 N , called the white layer (series II). A range of nitrided depths was obtained by varying the treatment times. DLC films were deposited from H z - C H 4 mixtures on the nitrided layers described above in a continuous way by lowering the total pressure and by substituting N 2 with C H 4. Nitrided and DLC coated samples display a homogeneous black appearance. The initial roughness of all samples was R a = 0.02 gm (mirror polishedt. The ionic nitriding treatment led to R a=0.04 lam which was not subsequently modified by the DLC film deposition process. The crystallographic structures of the nitrided top layers, as well as the interface structure after DLC film deposition, were verified by means of X-ray low-angle diffraction on a D-5000 Siemens diffractometer using Co Kct radiation. Microhardness Vickers measurements were performed on cross-sections of the samples in order to determine the nitrided depths. These measurements were compared with N 2 diffusion profiles obtained from G D O S (glow discharge optical spectroscopy) spectra. The nature of deposited DLC films was confirmed using Raman spectroscopy. Hydrogen concentrations were measured in some films using HFS (hydrogen forward scattering) spectrometry. Scratch test experiments were performed using the CSEM REVETEST apparatus with tangential force and acoustic emission monitoring. Dry sliding experiments were performed on a pin-ondisc type apparatus: a ball loaded with a normal force F N slides on the treated disc, which is rotated. The tangential force F v is monitored and the resultant friction


coefficient la = Fy/FT is plotted as a function of the number of disc revolutions. The wear rates of both disc and ball are measured using profilometry and weight loss methods. In the ring-on-block experiment, the loaded rotating ring rubs against an alternately moving plate or block. The system is immersed in an oil bath maintained at l l0~C. The dissipated power, which is the product of the normal load and the rotating speed of the ring, is progressively increased in order to reach the seizure threshold of the tested system. This experiment is used to study the various lubrication regimes and displays a good analogy with the rolling process in terms of the kinematics at the contact, but does not include the plastic deformation of the metal sheet.

3. Results and discussion As mentioned above, the transition between the ion nitriding phase of the surface treatment and the deposition of the DLC was carried out by reducing the pressure in the reaction chamber and replacing N2 with C H 4. The change of gas was conducted at a temperature of 400°C. Subsequent analyses of the samples using X-ray techniques confirmed that under such experimental conditions, cementite (Fe3C) formed systematically on the nitrided 35 CD 4 steel before the DLC film was formed (Fig. l(a)). Despite the presence of this Fe3C phase, the DLC films deposited in the course of the present work are adherent and exhibit good mechanical properties. The Raman spectra obtained are typically those of a hydrogenated amorphous DLC film (Fig. l(b)). The HFS analysis results gave a hydrogen content in the film of the order of 39 _+ 3 at.%. The duplex structures formed with different nitriding treatments were compared using a conventional scratch test during which the normal force applied to the sample increased from 0 to 50 N. The loading rate was set at

TABLE 2. Plasma treatments of 35 CD 4 steel samples and resultant structures Samples

Reference Series I

Series II

Plasma nitriding conditions

d.c.-PACVD of DLC films

Gas composition, pressure and temperature

Resultant structures

Nitrided depth

Gas H 2 N 2 T=480 C PN2 = 16 Pa Gas H 2 N 2 T = 5 2 0 ~C PN2=48 Pa

Diffusion nitrided case

10, 60, 100 and 150 p.m

Diffusion nitrided case+iron nitride (Fe4N) layer

10, 65, 100 and 160 gm; ~ 1 gm thick

Gas mixture and temperature

Growth rate

Film thickness

Gas H 2 CH 4 T~<400 °C

~,0.3 ~tm h t

1 1.5gm


F, Cellier, J.F. Nowak / Diamond-like carbon film deposition on steel substrates

series II 1




J r,'+




: 2theta M :



6 g. o100)

785+ Lin+lu



Rarnan o


series I






(cm "1)


J .j

!I! tl

] '

,,,+, +t r


x : 2thltl y :


/ i

350. L i n e #




Fig. 1. (a) X-ray spectra of plasma nitrided/DLC coated AISI 4135 steel samples--series I and series II typical nitrided cases. (b) Typical Raman spectra of amorphous hydrogenated DLC films.

50 N min -1 and the table speed at 10mm min -1 to produce a scratch of 10 mm nominal length. The diamond stylus had a tip radius of 200 I~m. The results of these investigations are given in Fig. 2(a). They are expressed in terms of the scratch hardness term Hr which can be defined as Hr=2P/~a 2 where P is the applied normal force and a is the half-width of the indentation [ 10]. The parameter Hr can be considered as an indication of the resistance of the material to indentation in the plastic regime. For the two series of samples considered, the surface hardening induced by the ion nitriding treatment results in an increase in the measured value of Hr, this increase being proportional to the depth of the treated layers. Complementary Vickers hardness measurements under a 1 kg load gave the same type of correlation between the surface hardness values and the depth of treatment. The influence of the depth of the nitrided layer on the mechanical properties of the DLC film is shown in Fig. 2(b). For all specimens investigated the criterion which was used in order to compare and classify the different treatments was the load required to produce the first spall in the film (Fig. 3(a)). Careful analysis of the tested samples confirmed that this load corresponds to the initiation of systematic damage of the film (cracking, spalling, crushing and the complete removal of the film in the bottom of the scratch), which thereafter extends with increasing load. It is also noteworthy that the first indications of spalling coincide with an increase in the tangential force

~ 10000 ~ 9000. "~ 8ooo. *~ 7000 ~= =~ 6o00 ~ seoo ~ 4000

Series I







1 O0




depth {gin)


~ a0 _o ._==20

Series II

" ~ :



o i





1 O0

1 50



depth (p.m)

Fig. 2. (a) Scratch hardness Hr as a function of nitrided depth. (b) Spalling load as a function of nitrided depth.

F. Cellier, J.f. j~rOW~lk , Diamond-like carbon film deposition on steel suhstrates


(b) Fig. 3. SEM examination of a scratch on a plasma nitrided/DLC coated A1S1 4135 steel sample (applied load increasing from right to left): (a) spalling phenomena: (b) zone at the end of the scratch.

(friction coefficient), which in turn is associated with coating damage. In the case of samples which were not subjected to the nitriding treatment or only to mild nitriding treatment, a collapse of the substrate with associated spalling of the film was observed for low loads (4-5 N). The scratch resistance increased with the depth of the nitrided layer and the first indications of spalling are observed at higher load levels ( 15-20 N for nitrided depths greater than 50 pm). The presence of side spalling is, however, limited, and careful analysis of the zone at the end of the scratch revealed no significant lateral cracking at the edges of the indentation (Fig. 3(b)). For both series of samples investigated, no traces of the DLC films at the bottom of the scratch were detected once the normal load exceeded 30 50 N. The particular value of the load was, as described above, a function of the depth of the nitrided layer. In all cases, however, once this load level was attained the presence of fine debris was observed in the scratch. Analysis conducted on specimens which had the same depth of nitrided layers showed that those of series II required systematically higher loads to produce


the first spalls. The presence of the white F e 4 N phase and its higher hardness level (which in turn limits the amount of plastic deformation of the substrate) is considered to be the reason for the superior properties of this series. Measurements of the depth of the scratches showed that for all samples, the first spalls were observed when the depth of penetration was of the order of 3 ~tm. The friction properties of nitrided discs typical of both series 1 and I I were investigated under unlubricated conditions using a ball-on-disc test. The balls were made from AISI 52100 bearing steel hardened to 63 HRC. The normal load applied was 1 N, which resulted in an initial mean Hertzian contact pressure of 514 MPa. The friction coefficients measured in air at ambient conditions (T-25'C and relative humidity of 55%) were relatively high (tL>I after 100 revolutions) and the values were seen to vary for the two types of surface considered (Fig. 4). Significant wear of both the ball and the disc were observed for tests which were stopped after 2200 revolutions. However, the discs which had been nitrided and subsequently coated with DLC films exhibited stable friction coefficients which were relatively low for the case of dry friction (iI~0.25). Tests which were conducted up to 20 000 revolutions showed negligible wear of the disc and very little wear of the balls. These findings confirm that the DLC coating results in greatly improved wear and friction properties of nitrided surfaces. Finally, in order to assess the potential of these DLC films for metal forming (rolling) operations, some ringtype specimens (55 mm diameter × 25 ram) made from high alloyed steel (8% Cr, 3% Mo A F N O R Z 85 CDV 8-3) were nitrided under H,-N~ plasma. Under such conditions the depth of the diffused nitrogen layer was of the order of 100 IJm. The surface hardness reached of the order of 1200 HV 0.1 with R a ~ 0.05 ~Jm, A DLCcoated nitrided ring was subsequently tested using a ring-on-block type machine. The seizure threshold of the treated ring under lubricated conditions (pure rolling oil) was compared with that of an untreated ring, a nitrided ring and a ring which had been coated with an electrolytic Cr film 10 [am thick. The antagonistic plate =.

2,0 1,8 t,6



o~ ¢J






Plasma n~trided AISI 4135 steel samples Nitrided diffusion case (series I) 41/-'~k.__ _

Ball-on-disk experiments Ball AISIIO0 C 6 Load 1 N Sliding speed 0,1 m/s




diffusion ;ase + white layer (series II)

0,4 0,2 0,0


Pl~'sma ~tride~ a~drD'LC coaled AlSl 4135 steel sample ........ i ........ i ........ j ........ 1 O0 1000 10000 100000 Number of revolutions

Fig. 4. Ball-on-disc experiments: friction coefficients j~ as a function of the number of revolutions.


F. Cellier, J.F. Nowak / Diamond-like carbon film deposition on steel substrates



Q" film




triton 2000 1800 1600 1400 1200 1000 8O0 600 400 2OO

m ._m m



Fig. 5. Ring-on-block experiments: seizure threshold under lubricated conditions (pure rolling oil).

was made of an AISI 304 austenitic stainless steel (Ra~8 gm). The results obtained for these tests are shown in Fig. 5. The DLC- coated ring exhibited excellent properties compared with the other rings. Even under the most severe conditions that can be imposed by the test machine (600 N x 2000 tr min -1) the DLCcoated ring did not "stick". Detailed examinations did, however, show some minor localized damage of the DLC film.

4. Conclusions

In the present work, a duplex structure resulting from a treatment which consists of an ion nitriding phase followed by a DLC film deposition process using d.c. plasma has been realized. Using several different test methods, the excellent tribological and mechanical properties of such a structure have been shown. Under our particular surface treatment conditions, the cementite

Fe3Cphase is systematically formed at the DLC/nitrided case top layer interface. Further work is considered necessary in order to evaluate the influence of this phase on the adherence of the DLC films. Indeed, certain modifications to the operational parameters of the treatment sequence could eliminate the presence of this phase altogether. Owing to their high plastic limit, leading to good mechanical resistance, the nitrided steel surfaces proved themselves to be an excellent support for the DLC films. Using scratch test methods, we have shown that the mechanical behaviour of the DLC films depends on the depth and the hardness of the underlying nitrided layer. Tribometer test results have demonstrated the muchimproved dry friction properties of DLC coated nitrided surfaces compared with those which have been subjected to a nitriding treatment only (applications in the field of mechanical engineering, wear components and so on). The results obtained under lubricated conditions have also shown the potential use of such films in the field of metal forming process equipment. References 1 F.M. Kustas, M.S. Misra, D.F. Shepard and J.F. Froechtenigt, Surf. Coat. Technol., 48 (1991) 113. 2 D.S. Kim, T.E. Fischer and B. Gallois, Surf. Coat. Technol., 49 (1991) 537. 3 A. Grill and V. Patel, Diamond Relat. Mater., 2 (1993) 597. 4 S.D. Gorpinchenko, S.M. Klotsman, E.V. Kuzmina, S.A. Plotnikov and I.Sh. Trakhtenbreg, Diamond Relat. Mater., 1 (1992) 619. 5 H.-J. Spies, K. Hoeck and B. Larisch, Proc. 8th Int. Conf. on Heat Treatment of Materials, Kyoto, Japan, November, 1992, Japan Technical Information Service, Tokyo, 1993, p. 333. 6 S.-M. Chiu, C.-T. Wu and J.-S. Lin, Proc. 8th Int. Conf. on Heat Treatment of Materials, Kyoto, Japan, November, 1992, Japan Technical Information Service, Tokyo, 1993, p. 547. 7 A. Leyland, K.S. Fancey and A. Matthews, Surf. Eng., 7(3) (1991) 207. 8 Y. Sun and T. Bell, Proc. 8th Int. Conf. on Ion and Plasma Assisted Techniques, Brussels, May, 1991, p. 290. 9 F. Cellier and J.F. Nowak, French Patent F 91 12 322. 10 E. Felder, CEMEF Seminary, Vol. 1, BIII 34, May 22-26, 1989, Sophia Antipolis, France.