SESSION #: ED-07

1

Raman Spectroscopy and Nano Hardness Investigation of Carbon Overcoats Used in Hard Disks and Sliders R. Brunner, A. Khurshudov, G. W. Tyndall, and F. E. Talke, Member, IEEE

Abstract—Structural and mechanical properties of a number of commercially available hard disks and sliders manufactured by various vendors were investigated using roughness measurement, Raman spectroscopy and nanoindentation. Raman spectra of sliders were found to be different than Raman spectra of disks. The variations in the spectra depend on the deposition technology that is used and identify the amorphisation stage of the carbon films. The carbon overcoats for the various disks investigated do not indicate substantial differences. Based on this result, it is concluded that manufacturers of the present day media use very similar manufacturing processes. Index Terms—Raman spectroscopy, hardness measurement, carbon, hard disks.

I. INTRODUCTION

T

HE INCREASE in storage density in hard disk drives requires a reduction in the magnetic spacing between slider and disk. A number of parameters contribute to the magnetic spacing, among them head/medium flying height, surface roughness on slider and disk, lubricant film thickness and the thickness of the carbon overcoat on slider and disk. Typical magnetic spacing values that need to be achieved at areal densities of 1 Tbit/inch2 are on the order of 5 to 6 nm [1,2]. Assuming that the flying height is 3 nm, a magnetic spacing of 6 nm would imply that the carbon film thickness on either the slider or the disk must be on the order of 1 nm. Film thicknesses of this order of magnitude are extremely challenging to evaluate, and great care has to be taken to deposit films of this small thickness with the required mechanical and materials properties. Diamond-like carbon (DLC) is presently used on slider surfaces and sputtered or CVD deposited carbon films are generally used for wear protection of hard disks. Hydrogenated or nitrogenated amorphous carbon films have been in common use [3]. As the films decrease in thickness, increasing care must be taken to deposit the films free of pin holes. This is especially true for ultra thin films on the order of 1 nm [4]. Coating technique such as cathodic arc deposition is mainly used in slider manufacturing and Manuscript received March,13th, 2006. R. Brunner and F.E. Talke are with the Center for Magnetic Recording Research, University of California, San Diego, 9500 Gilman Dr.-0401, La Jolla, CA, 92093-0401, USA (phone: 858-534-3175; e-mail: [email protected]; [email protected]). A. Khurshudov and G. W. Tyndall are with Samsung Information Systems America, San Jose, CA 95134 USA

develops super smooth and ultra thin tetrahedral amorphous carbon (ta-C) coatings [5]. RMS-roughness measurements of ultra thin carbon films for 1Tbit/inch2 storage density were reported by Casiraghi and Ferrari and are on the order of 0.1nm [5,6]. Raman spectroscopy was used by Scharf to characterize film thickness [7] and the structure of the thin carbon overcoat was intensively studied by Robertson [8,6]. Mechanical and tribological properties of tetrahedral amorphous carbon coatings were investigated by Liu [9]. This paper investigates structural and mechanical properties of present day carbon overcoats for a number of commercially available sliders and disks manufactured by various vendors. RMS-roughness measurements are first performed using atomic force microscopy to evaluate the smoothness of the coatings. Raman spectroscopy is then used to identify the morphology of the carbon overcoats and to evaluate to what extend the carbon films show differences for coatings manufactured by different vendors. Finally, characterization of the mechanical properties of the films is performed using nano-indentation measurements. II. RMS-SURFACE ROUGHNESS Contact mode AFM imaging was performed to measure the roughness of the sliders and the disks used in this study. Root-mean-square (RMS) roughness investigations were evaluated using a 0.75µm x 0.75µm sampling area. Figure 1 shows a comparison of the RMS roughness of the disks and of the sliders. The carbon film of the sliders shows RMS roughness between 0.25nm and 0.5nm, while the RMS-roughness of the amorphous carbon (a-C) for the investigated disks is between 0.25nm and 0.65nm. III. RAMAN SPECTROSCOPY Structural investigation of the carbon films was performed using Raman spectroscopy (Renishaw) at 514.5nm excitation wavelength. The disordered band (Dband) and graphitic band (G-band) configuration in the carbon films of the disks and sliders were investigated. The sp3-fraction in the spectra can shift the mode frequencies of the sp2-fraction. Using this phenomenon, one can derive the sp3-fraction from the shift in the graphite band [10]. The Raman spectra of the sliders indicate characteristics of tetrahedral amorphous carbon (ta-C) compared to the disks which show amorphous carbon (a-C) features. In the case

SESSION # ED-07

of the sliders, the spectra show shifts of the G-band, while the spectra for the disks do not show large differences. Figure 2 shows Raman spectra for sliders from five different vendors. All spectra show well-defined G-band peaks with defined shifts of ±40cm-1 around the common G-peak position at 1560cm-1. However, a well-defined Dband can not be observed. The G-peak positions for the individual sliders are in the range from ~1530cm-1 to ~1600cm-1. Slider #2 and slider #5 show the G-peak at a wavenumber of ~1530cm-1 and for slider #3 we observe a shift to ~1600cm-1. The well defined G-peak in this range indicates a tetrahedral amorphous carbon character of the protective overcoat used for the sliders [6]. The relative thickness of the carbon overcoat may be estimated by the peak amplitude of the G-band [7]. For sliders #1, #2 and #4 the G-band height is around 1100 in Raman intensity indicating a thickness of the carbon overcoat for those three sliders in the same range. The height of the G-band peak for slider #3 is very low compared to that of the other sliders. The spectrum of slider #5 was much higher than that of the other sliders. To fit the spectrum of slider #5 in figure 2, we scaled the spectrum by 1/2. The height of the G-band for slider #5 is therefore ~2600 in Raman intensity rather than ~1300 as shown in figure 2. If the assumption is correct that the peak height of the Raman signal is related to the film thickness, then slider #5 should have the highest film thickness and slider #3 the lowest. Raman spectroscopy was also used to study the carbon overcoat in a series of six disks. As shown in Figure 3, all of the disk spectra are very similar. The dominant peaks in the spectra are identified as the disorder band (D-band) located at ~1389cm-1 and the graphite band at ~1556cm-1. In addition to these carbon peaks, we observe small scattering signals at 841cm-1 and 1295cm-1 that are attributed to the underlying magnetic film. Information regarding the type of graphitization present in the carbon film can be obtained from the relative intensities of the D-band and G-band. On the basis of the results shown in figure 3, the ratio I(D)/I(G) in all the films studied is about 0.75. This result indicates that all of the films comprise a relative high fraction of amorphous carbon [6]. With respect to film thickness prediction, it is assumed that the film thickness for disks #2, #3 and #4 is higher than that for disks #1, #5 and #6. Furthermore, the similarity of the spectra suggests that the various disks were manufactured by comparable deposition techniques, i.e., the difference in the overcoat characteristics are small.

2

Fig. 1. Comparison of RMS roughness for commercially available disks and sliders

Fig. 2. Raman spectra of different commercially available sliders

IV. NANO-HARDNESS MEASUREMENT In addition to RMS-roughness measurements and Raman spectroscopy, nano indentation of the films was performed to evaluate the hardness of the six disks. A nanoindentation system (Hysitron) with a cube corner tip was used for hardness measurements. Figure 4 shows a typical load-displacement profile for nano-indentation into the coating of a hard disk. The

Fig. 3. Raman spectra of different commercially available disks

SESSION # ED-07

maximum applied force was 70µN, which results in a contact depth of approximately 10nm. Hardness measurements of very thin films are affected by the substrate. During the indentation process the carbon overcoat is penetrated and substrate effects become apparent. A comparison of the disk samples for exactly the same conditions was performed. Figure 5 shows a bar diagram of the hardness for the six disks investigated. The relative hardness is on the same order for all the disks and was measured to be between 7GPa and 8GPa. Similar to the results from Raman spectroscopy there was no large difference found between the various samples. The similarity in the characteristics of the load-unload curves of the various media lead us to conclude that the overcoats of hard disks from different vendors show very similar features.

3

V. SUMMARY In this paper we investigated carbon overcoats of commercially available disks and sliders from different vendors. The characterization of the carbon overcoat with respect to graphitization, hardness and thickness was evaluated. RMS-roughness measurements show values between 0.25nm and 0.65nm for the disks and the sliders. Raman spectroscopy shows characteristics of higher order tetrahedral-amorphous carbon for the coating of the sliders. The carbon overcoats investigated from the disks show only minor differences in their morphology and indicate an highly amorphous character. Nano-indentation of the disks investigated show relative hardness between 7GPa and 8GPa. Raman spectroscopy as well as nano-hardness measurements do not indicate substantial differences in the nature of the carbon overcoats of hard disks. Based on the above results it is concluded that manufacturers of present day magnetic thin film disks use very similar manufacturing processes. REFERENCES [1]

Fig. 4. Load–unload curve for hardness measurement of disk coating

1 µm 0.5 µm b)

a)

Fig. 5. a) Nano-hardness measurements on six disks b, c) AFM images of the indent

c)

D.J. Li, Y.W. Chung, “Ultrasmooth CNx Overcoats for NextGeneration Hard Disks,” IEEE Trans. Magn., Vol. 39, No. 2, 765768 March 2003. [2] R.W. Wood, J. Miles, T. Olson, “Recording Technologies for Terabit per Square Inch Systems,” IEEE Trans. Magn., Vol. 38, No. 4, 17111718, July 2002. [3] L. Zhou, K. Kato, N. Umehara, Y. Miyake, “The effects of texture height and thikness of amorphous carbon nitride coating of a hard disk slider on the friction and wear of the slider against a disk,” Tribol. Int., Vol. 33, pp.665-672, 2000. [4] C.M. Mate, B.K. Yen, D.C. Miller, M.F. Toney, M. Scarpulla, J.E. Frommer, “New Methodologies for Measuring Film Thickness, Coverage, and Topography,” IEEE Trans. Magn., Vol. 36, No. 1, 110-114, January 2000. [5] C. Casiraghi, A. C. Ferrari, R Ohr, A. J. Flewitt, D. P. Chu, and J. Robertson, “Dynamic Roughening of Tetrahedral Amorphous Carbon,” Phys. Rev. Lett., Vol. 91, No. 22-6104, November 2003. [6] A. C. Ferrari, “Diamond-like carbon for magnetic storage disks,” Surf. And Coat. Tech., 180-181, 190-206, 2004. [7] T.W. Scharf, I.L. Singer, “Thickness of diamond-like carbon coatings quantified with Raman spectroscopy,” Thin Solid Films, 440, 138-144, 2003. [8] J. Robertson, “Diamond-like amorphous carbon” Materials Science and Engineering, R 37, 129-281, 2002. [9] D. Liu, G. Benstetter, E. Lodermeier, „Surface roughness, mechanical and tribological properties of ultrathin tetrahedral amorphous carbon coatings from atomic force measurements,” Thin Solid Films, 436, 244-249, 2003. [10] J. Robertson, “Ultrathin carbon coatings for magnetic storage technology,” Thin Solid Films, 383, 81-88, 2001.