The carbon overcoat layer on a magnetic hard disk acts as protective barrier for the underlying magnetic media. In order to achieve high recording densities, it is necessary that this overcoat layer be extremely thin (5 to 10 nm). In addition, the film must exhibit very low surface friction to minimize wear and extend disk lifetimes. Generally speaking, nitrogenated carbon (CNx) films typically last 3 to 4 times longer than their hydrogenated carbon (CHx) counterparts. In practice, it is critical to precisely control both the thickness and N content of these protective layers. In this paper, we will demonstrate the capability of measuring the thickness and the composition, simultaneously and nondestructively, for thin (approximately 10 nm) CNx overcoat layers on magnetic disks using broad band reflectance spectroscopy in conjunction with a novel and recently developed analysis approach. This new technique utilizes the well-established Forouhi-Bloomer dispersion equations¹ for the refractive index (n) and extinction coefficient (k) spectra in undertaking the analysis of the reflectance data. From the curve fitting, both the n and k spectra as well as the thickness of CNx layers are determined. A set of samples², with N content varying from 0% to 21%, is the focus of this study. A strong correlation is established between the measured k spectra and the N content (as determined by Auger electron spectroscopy²) in the films. It is concluded that this new approach enables the N content of CNx films on finished magnetic disks to be determined with an accuracy of about ±1%.

¹A.R. Forouhi and I. Bloomer; Phys. Rev. B, 34, 7018 (1986). A.R. Forouhi and I. Bloomer; Phys. Rev B, 38, 1865 (1988). ²The authors would like to thank and acknowledge IBM San Jose for preparing samples and providing Auger analysis for this work.

Introduction

DLC thin films are in common usage in the hard disk industry. These coatings provide good wear resistance and low friction. During the growth of DLC films, large amounts of hydrogen can be incorporated into the film. To keep hardness numbers high, it is desirable to limit hydrogen content to about 25 atomic %. Measurement of the hydrogen content can be done using Hydrogen Forward Scattering (HFS) but the accurate measurement of films thinner than 5.0nm is difficult. Secondary Ion Mass Spectrometry (SIMS) has well known capabilities in the semi-conductor industry for measuring low concentration depth profiles for various elements in silicon. SIMS can also be used with remarkable effectiveness for measuring stoichiometry in many different materials using a CsM+ analytical protocol. Combining this analytical method with a very low energy primary ion beam to enhance depth resolution results in an accurate method for measuring hydrogen in films as thin as 2.0 nm.

The CsM+ Analytical Protocol

During sputtering with a cesium primary ion beam, cesium can combine with material from the substrate forming a molecule that is then easily ionized to a positive charge state (M+) above the substrate surface. When ionized above the substrate surface, the ionization process is relatively free of matrix effects and the number of ions are representative of the stoichiometry. By using a well characterized standard, SIMS quantification can be accurate over a wide range of stoichiometries.

Depth Resolution

Depth resolution is fundamentally limited by primary ion beam mixing. Cascade and recoil mixing effects can be reduced by lowering the primary ion beam energy. Depth resolution can be further improved by increasing the angle of incidence relative to the sample normal.

Results

We present data from a number of thin DLC films showing hydrogen concentration as a function of depth. Some thicker film results are compared with HFS measurements.

In this research, titanium nitride thin films are deposited by reactive DC magnetron sputtering with two original constraints due to the expected industrial application as roll to roll decorative coating of steel: the samples should be connected to earth and could not be heated during deposition. Previous work has shown that low N₂ atmosphere should be maintained during sputtering process in order to obtain a colour range from grey to gold. In this study, the nitrogen content measured by Resonant Nuclear Reaction Analysis and the crystal structure revealed by Glanced angle X-Ray Diffraction were combined to reveal coating composition. The appearance presented in CIE Lab colour co-ordinates and the micro-hardness obtained with a Berkovitch nano-indentor were also evaluated for all the films.

Monocrystalline polished silicon wafers were firstly used as easy to handle substrates to start thin film physical properties investigations. Results obtained with these TiN_x layer deposited onto Si(100) are secondly compared to coating physical properties measured onto more realistic polycrystalline iron (α-Fe) substrates.

Only the coating produced under the lowest N₂ gas mass flow exhibits a different nitrogen content on both substrates resulting in a Ti₂N phase onto iron. Colour difference appears and are probably due to different substrate roughness. A relatively good hardness between 12 and 22 GPa is obtained on both substrates. In order to qualify this process for possible industrial application, corrosion testing have been also performed on TiN_x layer deposited onto iron substrate.