The interactions of hyperthermal particles with surfaces represent an important class of phenomena for both basic research and technology. These processes occur in ion beam and plasma technologies, and even during the interaction of atoms of high pressure gas with solid surfaces. The subplantation model recognized that the most important processes in low energy ion beam deposition (IBD) involve the penetration of ions and damage of the solid. This qualitative model has recently been developed into a semiquantitative subplantation (SQS) model that describes the effects of penetration, damage, and diffusion (including radiation enhanced diffusion) in the film growth from hyperthermal species¹.

The SQS model has been applied to such diverse phenomena as the deposition of diamond-like carbon (DLC) and carbon nitride CN_x films, ion induced damage of graphite, low temperature silicon epitaxy, and deposition of titanium silicide. The model provides interpretations of effects of ion kinetic energy, fluence, and substrate temperature. The model is quite general and leads to new conclusions about physical mechanisms. It provides estimates of important physical parameters and optima in IBD that exist due to a balance between beneficial and deleterious effects. Finding these optima may lead to significant technological improvements in a wide range of ion beam and ion assisted technologies.

¹ K. J. Boyd et al., J. Vac. Sci. Technol. A 16 (1998) 444.

Amorphous carbon or Diamond like carbon (DLC) is deposited on recording heads to improve tribological properties at the head-disk interface and also to protect the transducer against corrosion in the drive. With increasing areal densities, the thickness of the carbon overcoats are decreasing rapidly to 50 Å and less. With a reduction in the overcoat thickness, it is important to understand the nature of the thin film growth and the properties of the film as the thickness changes.

The present work provides results of the properties of a-carbon overcoats which were characterized in terms of residual stress, raman spectroscopy, surface roughness, electrical resistivity and surface energy as a function of film thickness. The DLC was deposited using an ion beam CVD process using CH₄precursor. The overcoat has an adhesion layer that compromises 1 to 2 nm of silicon that is sputter deposited. Overcoats of different thickness were deposited on the recording heads along with some silicon witness coupons and wafers. The overcoat thickness used in these experiment are 30Å to 300 Å.

The results indicate that there is considerable change in the film properties as the film grows. The residual stress which is about 3.4 Gpa for the 30 Å overcoat gradually decreases with an increase in film thickness up to 100 A where the stress is about 2.5 Gpa. Above 100 Å film thickness, the stress increases again. The Raman intensity ratio on these films follow a similar pattern. Data on surface roughness, surface finish, and electrical resistivity are also presented. Implications of these measurements on the thin film growth mechanism are discussed. The results of the above techniques of DLC characterization imply that the growth of thin film a-carbon changes as the film increases in thickness.

This article is in three parts. The first part covers the numerous forms of both crystalline and non-crystalline carbon. Included are also two carbon compounds, namely SiC and C₃N₄, because of their high technological importance and relevance to advanced microelectronics. These are collectively referred to as carbon-based hard materials (CBHM).

The second part discusses the various deposition processes used to form CBHM, with an emphasis on those methods which produce amorphous-carbon (a-C) and hydrogenated amorphous-carbon (a-C:H) films containing a significant fraction of sp³ bonding or the so-called diamond-like carbon (DLC) films. A common feature of all these methods is the exposure of the growing layer to ion bombardment at medium energies (20 - 500 eV), which is a prerequisit for promoting sp³ bonding. The resulting DLC films exhibit valuable properties such as high mechanical hardness, low friction, transparency and chemical inertness. A description is given of a novel controlled plasma reactor COPRA and its particular use in the deposition of polycrystalline diamond, DLC and C₃N₄ layers.

The last section provides an overview of the analytical techniques that have been most frequently applied in the CBHM research. The relative merits of the characterization methods have been summarized and discussed. Each technique has its strengths and none of them alone can provide answers to all relevant questions concerning the structural, chemical and physical characteristics of the films. However, Raman spectroscopy, spectroscopic ellipsometry and electron-energy-loss spectroscopy have been found to be most useful for the evaluation of the CBHM films. It has been argued that any two of the above-mentioned techniques in combination can provide a sufficient basis for obtaining meaningful information on the structure and properties of these interesting and technologically-important films.

Effect of W-C intermediate layers on adhesion of DLC coating was examined in comparison with the case of Ti-C intermediate layers. Analyses of microstructures and failure mechanism in scratch adhesion tests were carried out.

Stacking sequence of W/W-C/DLC and Ti/Ti-C/DLC were prepared using dc magnetron sputtering on a WC-Co substrate or a glass substrate with a SiO₂ overlayer. Unbalanced magnetron sputtering system was used for the carbon deposition. W-C and Ti-C layers were designated to have gradient chemical composition by changing sputtering power step by step. DC or RF bias in a range from -50 V to -200 V was applied to the substrates during the deposition. DLC hardness increased with increasing the bias voltage, and was in a range from 40 GPa to 100 GPa.

In cross-sectional TEM microstructure analyses using elongated probe electron diffraction techniques, the W-C layers were found to be amorphous through thickness whereas the Ti-C layers included fine crystals of TiC compounds. Although the W system multilayers contained evident discontinuity in W/W-C interface and even in the W-C layers, adhesion of W/W-C/DLC multilayer on WC-Co or glass/SiO₂ substrate reached nearly 100 N, independent of the DLC hardness. On the other hand, adhesion of Ti system multilayers showed about 40 N at most.

Starting points of adhesion failure in scratch tests exhibited a feature of delamination by shear stresses in W system multilayers, while Ti system multilayers were failed in the Ti-C layers with cracking alongside the scratch path.

It is presumed that the difference of adhesion in hard DLC coatings with W-C and Ti-C intermediate layers are attributed to the microstructural differences. Relationship between microstructures and adhesion failure mechanism of the multilayers will be discussed.