Results of an investigation aimed at developing a technique by which the fracture toughness of a thin film or small volume can be determined in nanoindentation experiments are reported. The method is based on the radial cracking which occurs when brittle materials are deformed by a sharp indenter such as a Vickers or Berkovich diamond. In microindentation experiments, the lengths of radial cracks have been found to correlate reasonably well with fracture toughness, and a simple semi-empirical method has been developed to compute the toughness from the crack lengths. However, a problem is encountered in extending this method into the nanoindentation regime with the standard Berkovich indenter in that there are well defined loads, called cracking thresholds, below which indentation cracking does not occur in most brittle materials. We have recently found that the problems imposed by the cracking threshold can be largely overcome by using an indenter with the geometry of the corner of a cube. For the cube-corner indenter, cracking thresholds in most brittle materials are as small as 1 mN (∼ 0.1 grams). In addition, the simple, well-developed relationship between toughness and crack length used for the Vickers indenter in the microindentation regime can be used for the cube-corner indenter in the nanoindentation regime provided a different empirical constant is used.

Surface-film induced plasticity caused by the generation of dislocations at the film-substrate interface has been demonstrated previously for bcc metals and B2 ordered alloys deformed at low homologous temperatures. More recently, we have observed similar effects at elevated temperatures for ZrO2-coated MoSi2, which also demonstrates a film-induced reduction in hardness for microhardness tests over 25°C - 1300°C. In this investigation, 120 nm - 480 nm thick ZrO2 films were deposited on (001) single crystal MoSi2 by electron beam deposition. These composites were used to characterize the film-induced softening of MoSi2 at room temperature, where both MoSi2 and ZrO2 are brittle. Indents were made through the film into the substrate at 0.5 kgf loads using Vickers and Knoop indenters. Atomic Force Microscopy was employed to measure the geometry of Knoop indents after load removal, and to compare the results to the loaded indentation shapes. Increasing film thickness reduces the length of radial cracks in the substrate and lowers the composite hardness. The film also changes the relative amounts of deformation associated with material pileup at the surface near the edges of the indent, suggesting a change in deformation mechanism from pileup around the indenter edges to deformation by dislocation motion in the bulk.

Current methods for measuring hardness and modulus from nanoindentation load-displacement data are based on Sneddon's equations for the indentation of an elastic half-space by an axially symmetric rigid punch. Recent experiments have shown that nanoindentation unloading data are distinctly curved in a manner which is not consistent with either the flat punch or the conical indenter geometries frequently used in modeling, but are more closely approximated by a parabola of revolution. Finite element simulations for conical indentation of an elastic-plastic material are presented which corroborate the experimental observations, and from which a simple explanation for the shape of the unloading curve is derived. The explanation is based on the concept of an effective indenter shape whose geometry is determined by the shape of the plastic hardness impression formed during indentation.

Finite element simulation of the nanoindentation process has been carried out to investigate the nanomechanical response of materials and thin film systems. The influence of the geometry of the indenter, friction between the indenter and the surface, and pre-stress in the film, on the nanomechanical response have all been investigated. The effect of pile-up on the contact area calculation and the problems which occur with the commonly used methods in calculating the contact area have been studied. Results shows considerable error in calculating the contact area which depends on the indentation conditions used for the simulation. Simulation results suggest, that the influence of residual stress on hardness response is material sensitive.

A model calculates the compliance of elastic contact between a punch and bilayer, the latter consisting of a transversely isotropic film on top of isotropic substrate. Used in data analysis, the model helps estimate the Young’s modulus of amorphous silicon (100±20 GPa) layers within amorphous silicon/aluminum multilayer composites deposited on silicon.

The mechanics for plane strain geometry are presented for thin film fine lines by considering the case of microwedge indentation of a fine line subject to no residual stress. The analysis is separated into three parts depending on the absence or presence of buckling. During indentation, unbuckled or double-buckled configurations may exist, while single-buckling may occur when the indenter tip is removed. A microwedge tip was employed in the indentation testing of unstressed rf-sputtered tungsten fine lines to demonstrate some of the analysis.

The unloading of an indentation provides information about the shape of the indentation and the elastic properties of the materials. The assumptions of axisymmetry and material isotropy are critically examined, and a model for transversely isotropic materials is compared to measurements on single crystals. The methods used to infer the area of the indentation from the unloading curve are examined. The area is a fundamental value for the determination of hardness, modulus, and other mechanical properties in the so-called nano-indentor and other continuously monitored indentor techniques. The models of elastic recovery which are currently used are found to lack the flexibility to model the parameters which determine indentation depth. If the current self-consistent model is extended to cover the important aspects of the unloading, the area of the indentation is still not determined uniquely. Guidelines for further development of a unique model are suggested.

Nanohardness measurements have been carried out on barrier-type anodic aluminum oxide films grown on super-purity aluminum and on a 600 nm thick Al-1wt%Cu blanket film deposited on a silicon wafer. This work was performed using a custom-built displacement-controlled apparatus. The anodic oxide films range in thickness from 20 to 100 nm. The material systems investigated correspond respectively to hard-film/soft-substrate and soft-film/hard-substrate combinations. The experimental evidence suggests that indentation of anodic films begins when the penetration-depth is approximately 1/5th the film thickness. There is also evidence that the measured hardness for Al-1wt%Cu film stems largely from the thin film material when the probing depth does not exceed approximately 1/5th the film thickness. Beyond this depth the substrate influences the film deformation behaviour and the measured hardness gradually reaches the hardness of the substrate material. The oxide films on aluminum show unexpected deformation properties.

The deformation behavior of sol-gel derived zirconia films during ultra micro indentation has been investigated. The sol-gel films were deposited using a multiple stage coating procedure onto stainless steel substrates. The indentation process was carried out using both pointed indenters, such as a Berkovich, as well as spherical indenters. The use of different indenter types enables the determination of hardness, elastic modulus as well as the elastic-plastic or elastic-brittle transition of the films. The mechanical behavior of thin films is strongly influenced by the interaction between the substrate and the film. In brittle materials, the stress singularities induced by the indentation process cannot readily be reduced by long range plastic flow and may cause cracking. Therefore, a great deal of attention has been paid to post-indentation observations of the films, in particular examinations of the surface morphology for visible traces of the film behavior during indentation.

This paper describes the results of an analysis of the errors that can arise during nanoindentation, and their effect on the final results. The sources of errors that have been considered are calibration errors in load, displacement, instrument frame compliance and indenter shape assuming a Berkovich indenter with an error in the included angle, and operational errors such as thermal drift and specimen tilt. An assessment of calibration errors indicates that a simple summation of errors was not appropriate in this case, since compliance and area function are depended upon displacement. Thus the error analysis has been carried out using a cumulative approach. As expected it was found that the errors increased with decreasing load and displacement, and increasing hardness and Young’s modulus. However, minimum errors were observed at about 20 mN and were a function of the indented material

No appreciable error was found in the hardness or Young’s modulus calculated from load-displacement data at surface angles below 8°; above this value the indenter started moving sideways over the surface, resulting in significant errors.

Using ion beam assisted deposition, 1.5–2.2 μm thick molybdenum and nickel films were prepared on silicon substrates. Some films were found to be strongly textured. By changing the rate of Ar+ bombardment during the deposition, the resulting in-plane film stresses could be changed from being strongly tensile to strongly compressive. Using nanoindentation, the hardness and elastic modulus were measured for all films, but no major influence of the film stress or different textures could be found. The elastic modulus of the Ni films was found to be close to its polycrystalline bulk value, and that of Mo was found to be about 70% of its polycrystalline bulk value.

Platinum films of thickness 0.5 - 1.2 - 3.0 μm, deposited by magnetron sputtering onto glass and silicon, were investigated using ultra-microindentation. The tests were done with a pointed (Berkovich) indenter. The values of the hardness and elastic modulus of the composite “film + substrate” and the substrates alone were obtained for various loads and depths of penetration, and used to evaluate the hardness and elastic modulus of the films. The use of various film thicknesses and substrates has enabled us to investigate in more detail the influence of the contact radius/contact depth compared to the film thickness as well as the substrate properties on the total indentation behaviour. Various approaches to the data evaluation are discussed in the paper.

Thin film magnetic disks require protective overcoats, usually some form of carbon, to guard against corrosion and wear from interaction with the read/write transducer. In current products these films are less than 25 nm in thickness. This paper summarizes developments using scanning probe microscopes with sharp diamond tips (15 – 100 nm radius) to obtain indentation hardness tests with 5 nm deep indentations. We discuss an accelerated wear test that can measure wear at depths on the order of 1 nm. Finally material characterizations related to friction over sub-micron scans are discussed.

A novel observation has been made when studying the dependence of friction coefficient on normal load: below a critical load, which is material and tip dependent, no observable wear occurs, and the coefficient of friction is about 0.05. Above the critical load the coefficient is load dependent and increases to a value more usually associated with the materials being tested. A study of fatigue wear was made in the “no-wear” regime with three different results. For some materials, fatigue wear occurred with multiple passes, when none was apparent for a single pass. Other materials showed no fatigue wear, and one material, silicon, showed a build-up or “negative-wear” under multiple passes. Interpretations and implications of these results are discussed.

Molecular dynamics computer simulations of the growth processes and microstructural properties of amorphous carbon (a:C) and amorphous hydrogenated carbon (a:CH) ultra-thin films have been performed. Films 1 to 10 nm thick were grown on a diamond (100) surface using Brenner’s[1–2] bond-order potential for hydrocarbons. The stoichiometry, radial distribution function, chemical bonding (amount of sp2 and sp3 hybridization) and residual stress are presented.

By combining the micro-wear and microscratch techniques in a microindenter, a new methodology to investigate the mechanical properties of ultrathin coatings has been developed. With a 0.18 μm/sec scratch speed along a Y-direction while oscillating the indenter tip along the X-direction at a 6.55 Hz frequency, uniform wear tracks of approximately 20 μm × 90 μm in dimension can be created under either a constant or a gradually increased applied normal load. The wear-scan methodology has rendered several immediate advantages over the conventional micro-scratch technique, offering features such as a much larger testing area thus facilitating the failure and morphology analyses afterwards, the possibility of studying the wear debris generation and the flexibility in testing characteristics. Two thin film disks with different hydrogenated carbon overcoats have been tested by using the micro-wear scan technique. The results in the mechanical measurements, the wear processes and the morphology evolution are illustrated and discussed.

The ability of a hard coating to resist delamination in wear applications is found to depend on the substrate crystal orientation. A variety of alloys with and without a carbon coating were tested by microscratching with a diamond stylus. The hardness of the carbon coating was characterized by nanoindentation techniques before the microscratch tests. As the diamond passed from one crystal to another differing amounts of damage were observed for each crystal orientation. The width of the delaminated region was found to be dependent on the extent of the slip line fields in the substrate generated by the scratching. To eliminate the possibility of differing adhesion on different crystal faces, a nickel aluminide intermetallic with the two crystals rotated about the same pole was tested. In the uncoated state when scratched across the grain boundary the slip line fields were extensive for one crystal but limited for the other crystal. Similar fixed-load scratches on the carbon coating produced delamination on the easy slip orientation but only a slight deformation of the coating on the other crystal. The observed failure appears to be due to the deformation of the substrate and not inferior adhesion of the coating to the crystal face. This finding should have important ramifications in the design and failure analysis of systems, such as magnetic recording media, medical implants, etc.

Nanoindentation and nanoscratching experiments have been performed to assess the mechanical and tribological behavior of three thin film materials with potential application as wear resistant coatings for magnetic disk storage: (1) hydrogenated-carbon (CHx); (2) nitrogenated-carbon (CNx); and (3) boron suboxide (BOx). The hardness and elastic modulus were measured using nanoindentation. Ultra-low load nanoscratching tests were performed to assess the relative scratch resistance of the films and measure their friction coefficients. The mechanical and tribological performance of the three materials are discussed and compared.

Mechanical and tribological properties of soft magnetic thin films with high permeability and low coercivity are very important for the application of these films in high-density recording heads. This paper reports our experimental observations on these important properties of FeTaN thin film head materials. Hardness(H) and Young’s modulus(E) for FeTaN sputtered films were determined by nanoindentation. Wear resistance of these films against commercial magnetic tapes was measured with a sphere-on-flat wear tester. The experimental results indicate that the FeTa films can be hardened when nitrogen is introduced. It was found in this study that the thermal stability of the mechanical properties such as hardness of thin films containing nitrogen is better than that of the film without nitrogen. However, our experiments also revealed that the wear resistance of FeTaN films decreases when the concentration of nitrogen increases and the hardness of the worn surface at a wear scar is lower than that of the unworn surface.

In order to characterize friction and wear behavior between cylindrical specimens and flexible counterface materials such as magnetic tape, a tribometer is suggested. The tribometer is used to investigate the behavior between video cassette recorder (VCR) head drums and VCR tapes. Friction coefficients of uncoated and diamond-like carbon (DLC) film coated head drums of Al alloy are compared before and after a wear test. Before the wear test, an uncoated head drum shows friction coefficients ranging from 0.22 to 0.3 as the applied tension increases from 200 to 600 gf at the sliding speed of 0.16 m/s. The friction coefficients of DLC coated head drum are 14 to 30 % lower than those of uncoated head drums. At the sliding speed of 4.9 m/s, the friction coefficients of both uncoated and DLC coated drums are reduced to less than 0.1. After the wear test, however, the friction coefficients of uncoated head drums increase to the range from 0.3 and 0.47 depending on the applied tensions. Furthermore, the friction coefficients are not reduced even at the high sliding speed of 4.9 m/s. DLC coated head drums, on the other hand, exhibit similar friction coefficients to those before the wear test. Degradation of frictional property in the uncoated head drum is discussed in terms of the surface damages mainly due to magnetic metal oxide particles transferred from VCR tapes. This result shows that the DLC coating of VCR head drum can improve the friction and wear properties. It is also observed that the DLC coating can reduce the damage of VCR tapes.

Continuous microscratch tests on an uncoated and a diamond-like carbon coated CoCrMo alloy were performed in air to evaluate the micro-friction and micro-tribology by utilizing a load and depth sensing instrument. By using a diamond-like carbon coating, the friction coefficient for a small applied load was found to decrease compared to the CoCrMo substrate. The hard diamond-like carbon coating prevents the local plastic yielding of the micro-contacts and the adhesion between the metal/metal surfaces during the microscratch tests. However, the effect of the diamond-like carbon coating on the friction was not seen when the applied load is high.