Comparison of tribological properties of carbon and carbon nitride protective coatings over magnetic media Andrei Khurshudov,a) Koji Kato, and Sawada Daisuke Laboratory of Tribology, School of Mechanical Engineering, Tohoku University, Sendai 980-77, Japan

~Received 14 December 1995; accepted 7 June 1996! According to theoretical predictions made in 1989–1990, carbon nitride can be even harder than diamond. Carbon nitride thin films may become a good competitor for diamondlike carbon, which possesses a high hardness, high wear resistance, and low friction coefficient. At present, there are only a few studies on the tribology of CNx coatings. Tribological properties of the deposited by means of ion beam assisted deposition carbon nitride coating ~90%C–10%N! were compared with those of commercial carbon coatings at macro- and microscales. CNx coating on a magnetic rigid disk showed both lower macro- and microfriction compared with that of a carbon coating in unlubricated sliding against a silicon nitride pin. The life of the CNx coating was about three to ten times longer than the life of carbon coatings. © 1996 American Vacuum Society. I. INTRODUCTION In 1989–1990, Liu and Cohen1,2 theoretically predicted a hypothetical material, b-C3N4 , that has a structure similar to b-Si3N4 and may be even harder than diamond. One of the most attractive areas for CNx coatings may be the tribological area, but there is still only a limited number of works on the tribology of carbon nitride.3–9 Application of thin hard CNx coatings for magnetic media protection, for example, has not been well studied yet.3,9 Today, mainly diamondlike carbon ~DLC! coatings are used in the field of magnetic recording to protect magnetic disks and tapes. Carbon nitride is a material that may successfully compete with DLC coatings, which have high hardnesses, high wear resistance, and low friction. The purpose of this work is to describe the following. ~i! Deposition of thin ~1.5–30 nm! CNx coatings by means of ion beam assisted deposition ~IBAD! on magnetic media of a 1.8 in. magnetic rigid disk. ~ii! Tribological testing of the deposited CNx coating and comparison with the properties of commercial carbon coatings over magnetic media. II. EXPERIMENT

B. Coating characterization and testing

A. Coating deposition

The IBAD system used in the present study was developed by Hitachi, Ltd., Japan, and consisted of a cryogenically pumped chamber, a sputter deposition source, a low and high energy bucket type ion source, and a substrate holder. The diameter of the ion beam irradiation area was about 80 mm. The substrate holder consisted of a watercooled copper plate that can be rotated ~4 rpm! during the deposition. The CNx coatings were deposited on magnetic media of 1.8 in. rigid disks, held at room temperature, using simultaneous Ar1 ion beam sputtering of carbon and bombardment by nitrogen ions. Background and operating pressure in a vacuum chamber were better than 131026 and about a!

Electronic mail: [email protected]

2935

J. Vac. Sci. Technol. A 14(5), Sep/Oct 1996

1.431024 Torr, respectively. A 99.95% pure carbon target was used in these experiments. The ion source energy was 1 keV and ion current was 100 mA. At first, a 1.8 in. rigid disk with a 1.5-nm-thick carbon coating ~90%C–10%Si! was set into the chamber. Next, the disk surface was exposed for 5 min to bombardment by nitrogen ions ~1 keV, 100 mA/cm2!. This bombardment resulted in removal ~dry etching! of the surface layer that was about 10 nm thick. The rate of the etching was calibrated using surface masking with the following topography measurements by atomic force microscope ~AFM! ~Fig. 1!. The positive point @see cross section, Fig. 1~b!# is that the initial surface texture was not noticeably changed during etching. This process removed the initial carbon coating and exposed the magnetic layer. Carbon nitride coatings of 1.5–30 nm thickness were then deposited on the magnetic media of the disk. The deposition rate of the carbon nitride was monitored via a calibrated quartz crystal oscillator and was about 1.4 nm/min. The assisted N1 ions energy was equal to 1 keV and ion current density was 20 mA/cm2. The ion beam incident angle to the substrate was 45°.

After deposition of the CNx coatings, the chemical composition was evaluated by x-ray photoelectron spectroscopy ~XPS! ~ULVAC PHI ESCA-5500!. The measurements were performed using Al K a ~1486.5 eV! x-ray radiation and recorded with an energy analyzer pass energy of 187 eV. The surface topography of the magnetic rigid disks with the carbon coatings and the deposited carbon nitride coatings was studied by AFM. Both macroscale and microscale tribological tests were carried out in this work. The following tribological parameters were evaluated: coefficient of friction, coating life ~number of cycles before failure!, and force of adhesion between coatings and the Si3N4 pin ~for the AFM test only!. All tribological tests were done in laboratory air, where the humidity was 40%–60% RH. Along with the testing of the carbon nitride coatings de-

0734-2101/96/14(5)/2935/5/$10.00

©1996 American Vacuum Society

2935

2936

Khurshudov, Kato, and Sawada: Tribological comparison of C and CNx coatings

2936

FIG. 1. Ion beam etched and unetched surface of a 1.8 in. carbon coated disk ~a!. Cross section ~b! shows the depth of the etching ~about 10 nm! and the roughness of the initial and treated surfaces.

posited on magnetic media, two types of commercially available carbon-coated rigid disks were tested using a pin-ondisk tester against a Si3N4 ball. The first was a 1.8 in. magnetic rigid disk with a protective 90%C–10%Si coating of 1.5 nm thickness on the same substrate as was used for the CNx coating deposition. The second was a 3.5 in. magnetic rigid disk with a sputtered amorphous carbon ~a-C! coating of 35 nm thickness. No lubricant was used. The radius of the Si3N4 ball was 2.0 mm, contact load was 0.005–0.02 N, Hertzian contact stress was about 131–209 MPa, and the sliding velocity was 0.5 m/s. The frictional force was continuously recorded during the test. Tests were stopped when even a faint wear track was observed by the naked eye on the disk surface. This correJ. Vac. Sci. Technol. A, Vol. 14, No. 5, Sep/Oct 1996

sponded to the first exposure of the magnetic layer. In all cases, damage ~wear! of the coating was confirmed by observation under the optical microscope. In order to compare microtribological properties of CNx and carbon coatings, such as friction and adhesion, the test was carried out using the atomic force microscope ~AFM! SPA300 series ~Seiko Instruments, Inc., Japan! operating in the contact mode.10 The frictional force was measured during sliding of the Si3N4 AFM tip ~10–20 nm radius! against the tested coating surface through the AFM cantilever torsion. Sliding speed was about 0.01 mm/s and contact load was equal to 0.09– 44.5 nN. Adhesion between the Si3N4 AFM tip and the tested coat-

2937

Khurshudov, Kato, and Sawada: Tribological comparison of C and CNx coatings

2937

FIG. 2. AFM image of the surface of 1.8 in. ~a! and 3.5 in. carbon coated disks.

ing was measured10 as a maximum attractive force at the point of tip–specimen separation. For this purpose, the specimen was moved into contact with the tip and pulled back after the contact. The force of the tip–specimen interaction was calculated from the AFM cantilever deformation. III. RESULTS AND DISCUSSION According to the x-ray photoelectron spectroscopy, the C–N composition contained about 90 at. % carbon and 10 JVST A - Vacuum, Surfaces, and Films

at. % nitrogen and could be referred as C0.9N0.1. The elemental composition of the film was determined using the relative sensitivity factors and the observed peak intensities. No oxygen was found in the bulk material of the film. The AFM study showed that deposited CNx coatings replicated the substrate microgeometry. No significant changes occurred in the surface topography of the 1.8 in. disk after CNx coating depositions. Figure 2~a! shows topography of the 1.8 in. disk with

2938

Khurshudov, Kato, and Sawada: Tribological comparison of C and CNx coatings

2938

FIG. 5. The life of CNx and carbon coatings sliding against a Si3N4 pin.

FIG. 3. Friction coefficient of CNx and carbon coatings sliding against a Si3N4 pin under a normal load of 0.02 N.

‘‘bump’’-type texture. Measured by AFM the center-line average roughness ~Ra!, peak-to-valley roughness ~P–V!, and root mean square roughness ~rms! were about 5.6, 43.2, and 7.6 nm, respectively. Figure 2~b! shows the topography of a 3.5 in. disk with circumferential texture. Corresponding roughness parameters were Ra53.3 nm, P–V521.1 nm, and rms54.0 nm. Figures 3 and 4 show friction coefficients for the carbon nitride coating ~1.5, 4, 8, and 30 nm thick! and carbon coatings. Figure 3, observed at normal load equal to 0.02 N, shows that both the initial ~0.12! and final ~0.14! friction coefficients for CNx were less than half that of carbon. Figure 4, observed at smaller normal load of 0.005 N, shows an increase in the friction coefficient. The friction coefficient of CNx is still smaller than that of carbon coatings, but the difference is not as large as in the case of Fig. 3.

FIG. 4. Friction coefficient of CNx and carbon coatings sliding against a Si3N4 pin under a normal load of 0.005 N. J. Vac. Sci. Technol. A, Vol. 14, No. 5, Sep/Oct 1996

Two explanations may be suggested for the friction coefficient increase with the load decrease. First, the most probable reason is the increasing effect of adhesive forces at the interface, which are not taken into account in the friction coefficient calculation process. The second reason is the equipment systematic error increase with the load decrease. A smaller difference in friction coefficients for a smaller nor-

FIG. 6. Adhesive force ~a! and friction coefficient versus normal load dependence ~b! in the contact of CNx and carbon coatings against a Si3N4 AFM tip.

2939

Khurshudov, Kato, and Sawada: Tribological comparison of C and CNx coatings

mal load leads to the conclusion that the adhesion between CNx and Si3N4 is stronger than that between carbon and Si3N4 . Figure 5 shows an effect of coating thickness on coating life. It can be seen that CNx provides from three to ten times longer life than carbon coatings of the same thickness. Figures 6~a! and 6~b! summarize the results of the AFM tests. Figure 6~a! shows values of the adhesive force for CNx and carbon coatings against the Si3N4 AFM tip. As was suggested earlier, adhesion between CNx and Si3N4 is stronger than that between the carbon coating and Si3N4 . In spite of higher adhesion at the interface, CNx showed a lower coefficient of friction than did carbon @Fig. 6~b!#. The Hertzian contact radius in the AFM tests was about 0.23–1.9 nm and almost equal for both carbon nitride and carbon coatings. Therefore, it is possible to conclude that the shear strength of CNx is lower than that of carbon. It partly explains how the CNx ~see Figs. 3 and 4! provided lower friction than the carbon coating did. IV. CONCLUSIONS Testing of tribological properties of thin ~1.5–30 nm! CNx coating ~90%C–10%N! made by ion beam assisted deposition on magnetic media of a 1.8 in. magnetic rigid disk and their comparison with properties of commercial carbon coatings over magnetic media have shown the following. ~i! The friction coefficient in the range of 0.12–0.14 was observed for the CNx coating sliding at normal load of 0.02 N against a silicon nitride pin. This friction coefficient was lower than that of a carbon coating ~about 0.28–0.3!.

JVST A - Vacuum, Surfaces, and Films

2939

~ii! The carbon nitride coating provided about three to ten times longer life than carbon coatings. ~iii! In spite of higher interfacial adhesion to the silicon nitride pin, carbon nitride showed lower friction than carbon at both macro- and microscales. ACKNOWLEDGMENTS The authors would like to thank Y. Miyake and S. Nagaike of the Head Disk Interface Design Department, Advanced Products and Technology Development Division, Data Storage and Retrieval Systems Division, Hitachi, Ltd., for their support of this work. They also would like to thank K. Goto from the Center of Surface Analysis, Center Laboratory, ALPS Electric Co., Ltd., for the assistance in ESCA examination. A. Y. Liu and M. L. Cohen, Science 245, 841 ~1989!. A. Y. Liu and M. L. Cohen, Phys. Rev. B 41, 10 727 ~1990!. 3 T.-An. Yen, C.-L. Lin, J. M. Sivertsen, and J. H. Judy, IEEE Trans. Magn. 27, 5163 ~1991!. 4 M. Y. Chen, X. Lin, V. P. Dravid, Y.-W. Chung, M.-S. Wong, and W. D. Sproul, Surf. Coat. Technol. 54/55, 360 ~1992!. 5 M. Y. Chen, X. Lin, V. P. Dravid, Y.-W. Chung, M.-S. Wong, and W. D. Sproul, Tribol. Trans. 36, 491 ~1993!. 6 E. H. A. Dekempeneer, J. Meneve, J. Smeets, S. Kuypers, L. Eersels, and R. Jacobs, Surf. Coat. Technol. 68/69, 621 ~1994!. 7 D. Li, E. Cutiongco, Y.-W. Chung, M.-S. Wong, and W. D. Sproul, Surf. Coat. Technol. 68/69, 611 ~1994!. 8 D. Li, Y.-W. Chung, M.-S. Wong, and W. D. Sproul, Tribol. Trans. 37, 479 ~1994!. 9 A. Khurshudov, K. Kato, and D. Sawada, Proceedings of the International Tribology Conference, Yokohama, Japan, 1995 ~in press!. 10 A. Khurshudov and K. Kato, J. Vac. Sci. Technol. B 13, 1938 ~1995!. 1 2