In order to understand the effects of plastic deformation and the interfacial fracture energy on the peel strength, thickness of the metal layer and the pretreatment conditions of polyimide were varied in the Cu/Cr/polyimide system. The work expenditure during the peel test was estimated using the stress strain curves of metal films, X-ray measurements of the plastic strain in the peeled films, and the elastoplastic beam analysis. Results indicate that the peel strength is strongly affected by the film thickness and the pretreatment condition in a synergistic way, and that the measured peel strength is more a measure of the plastic deformation during the peel test than a measure of the true interfacial energy.

Methods currently used for analyzing nanoindentation load-displacement data give good predictions of the contact area in the case of hard materials, but can underestimate the contact area by as much as 40% for soft materials which do not work harden. This underestimation results from the pile-up which forms around the hardness impression and leads to potentially significant errors in the measurement of hardness and elastic modulus. Finite element simulations of conical indentation for a wide range of elastic-plastic materials are presented which define the conditions under which pile-up is significant and determine the magnitude of the errors in hardness and modulus which may occur if pile-up is ignored. It is shown that the materials in which pile-up is not an important factor can be experimentally identified from the ratio of the final depth after unloading to the depth of the indentation at peak load, a parameter which also correlates with the hardness-to-modulus ratio.

The measurement of mechanical properties by nanoindentation methods is most often conducted using indenters with the Berkovich geometry (a triangular pyramid) or with a sphere. These indenters provide a wealth of information, but there are certain circumstances in which it would be useful to make measurements with indenters of other geometries. We have recently explored how the measurement of hardness and elastic modulus can be achieved using sharp indenters other than the Berkovich. Systematic studies in several materials were conducted with a Vickers indenter, a conical indenter with a half-included tip angle of 70.3°, and the standard Berkovich indenter. All three indenters are geometrically similar and have nominally the same area-to-depth relationship, but there are distinct differences in the behavior of each. Here, we report on the application of these indenters in the measurement of hardness and elastic modulus by nanoindentation methods and some of the difficulties that occur.

Small volumes of crystalline material can exhibit near theoretical lattice strengths due to the absence of pre-existing defects. We show here that careful control of experimental parameters and surface preparation are important in observing these increased stresses. In particular, tip impact kinetic energies at the initial contact of only a few eV have been found to influence events occurring at a much later stage in the loading cycle. This can be understood in terms of the energy required to deform a small number of atomic scale asperities and thus release sufficient surface energy to permit dislocation nucleation at impact. If the K.E. is not sufficient to overcome the asperities, dislocations must be nucleated at a later stage in the loading cycle, giving a apparently different mechanical property. We outline the implications for thin film systems where a hard surface layer may play a vital role, firstly, in preventing dislocation nucleation on impact and, secondly, in impeding the propagation of any such dislocations into the bulk below.

Nanoindentation studies often show an instantaneous displacement-excursion in the loaddisplacement curve. This anomaly is generally associated with a surface contamination effect, dislocation emission, a phase transition, or an oxide break-through event. The determination of which effect is operative is often difficult when investigating oxide covered surfaces. We have performed a detailed nanoindentation study on clean, flame annealed single-crystal Au thus eliminating the effects of a surface oxide or contamination layer. Multiple displacement excursions were observed exhibiting a new phenomenon of “staircase” yielding. Owing to the fact that our radius of contact is more than an order of magnitude smaller than the average dislocation spacing expected for well annealed Au, the excursions are explained in terms of multiple dislocation nucleation events on parallel slip bands. Indentation data were taken on Au (111), (110), and (100) single crystal surfaces.

The evolution of the contact area during an indentation process has been examined by an a.c. technique and also by finite element analysis on five different bulk materials. The consistency of the results obtained by the two approaches not only show the validity and advantage of the a.c. technique but also corroborate the applicability of the elastic contact equation.

One of the critical aspects of assessing the properties of coated systems is the prospect of being able to measure the properties of the coating in isolation of the substrate. This has led to an increase in the use of continuously-recording indentation techniques (or nanoindentation techniques) for assessing the mechanical properties of coated systems since they can be used to measure the materials response to contact deformation at a scale relevant to the coating thickness. This paper presents the results of a new method for analysing the nanoindentation loading curves for coated systems. The analysis of the loading curve uses the relationship P = Kmδ2 which describes the indenter displacement, δ, in terms of the applied load P. For each material, Km can be predicted from its modulus E and hardness H. One result is that if either of E or H is known, then the other may be calculated from the experimental loading curve. Further, the method has the potential to differentiate between the responses of the coating only, the coating and substrate in combination, and that dominated by the substrate once the load has become sufficient for cracking of the coating such that it no longer plays a significant role in supporting the applied load. In many cases, careful analysis of the loading curves allows the displacement ranges associated with these regimes to be identified. This is shown to be a powerful experimental tool for the interpretation of the mechanical properties of coated systems.

A nanoindentation study of a calcium-aluminosilicate/α-alumina system is described. Thinfilms of CaAl2Si2O8 glass were deposited on α-alumina single crystals of {1120} and {1102} orientations by pulsed-laser deposition. These samples were then subjected to a heat treatment causing the glass to crystallize and to dewet the sapphire surface. A nanoindentation device retrofitted on an atomic force microscope was employed to probe the mechanical properties of the samples. It uniquely enables the characterization of sub-micron features such as the dewet crystallites which necessitate small displacements and exact positioning of the indenter. Strong dependencies on the crystallographic anisotropy were observed in the morphology, but less so on the mechanical properties. In addition, displacement-excursions were observed and are explained in terms of yielding in the substrate.

The applications of super hard coatings of diamond-like carbon (DLC) films are often limited by adhesion problems caused by intrinsic stress in the deposited layer due to the PACVD growth process. One approach to reduce the stress are less optimal process parameters. However, they also affect the desired high hardness of the films. Our approach in this study is to add B2H6 to the gas mixture, which shifts the composition towards another very hard compound (B4C). A sequence of amorphous BxC1-x:H films with a thickness of 0.5μm and a boron content between 0 and 50 at.-% were deposited on silicon substrates using a capacitively coupled r.f PACVD reactor. The tribological film properties and the internal stress were determined by depth-sensing indentation and laser reflection measurements. For an addition of x ≥5 at.-% boron the decrease of the internal film stress is found to be clearly larger than the effect on the hardness value. For boron contents x > 18% the internal stress is reduced by a factor of 5 while the reduction of hardness is only a factor of 2.3. For microscopic structure analysis Raman and infrared spectroscopy are applied. They reveal an increasing hydrogenation of the carbon network. Therefore, the softening is attributed to a boron induced modification to a more polymeric-like material.

Methods currently used for analyzing nanoindentation load-displacement data to determine a material's hardness and elastic modulus are based on Sneddon's solution for the indentation of an elastic half-space by a rigid axisymmetric indenter. Although this solution is widely used, no attempts have been made to determine how well it works for conditions of finite deformation, as is the case in most nanoindentation experiments with sharp indenters. Analytical and finite element results are presented which show that corrections to Sneddon's solution are needed if it is to be accurately applied to the case of deformation by a rigid cone. Failure to make the corrections results in an underestimation of the load and contact stiffness and an overestimation of the elastic modulus, with the magnitude of the errors depending on the angle of the indenter and Poisson's ratio of the half-space. For a rigid conical indenter with a half-included tip angle of 70.3°, i.e., the angle giving the same area-to-depth ratio as the Berkovich indenter used commonly in nanoindentation experiments, the underestimation of the load and contact stiffness and overestimation of the elastic modulus may be as large as 13%. It is shown that a simple first order correction can be achieved by redefining the effective angle of the indenter in terms of the elastic constants. Implications for the interpretation of nanoindentation data are discussed.

This paper reports nanoindentation studies of the effect of Ta and N content on the mechanical properties of magnetically soft high moment FeN and FeTaN thin films prepared by dc magnetron sputtering. The FeTaN films were deposited on oxidized silicon (100) substrates with a series of FeTa targets in which the Ta content varies from 0 to 25wt%. The hardness (H) and Young's modulus (E) were measured by the Nano Indenter at nine indenter penetration depths: 20, 30, 40, 50, 60, 80, 100, 120 and 200 nm. The inherent hardness values of these films (no substrate effect) can be determined at penetration depths ranging from 20 to 60 nm for the 500 nm thick film used in the study. It was found that for the films deposited from the pure Fe target when the nitrogen flow rate increases from 0 to about 0.5 sccm the hardness of the film increases. However, a decreasing trend in hardness of these films was observed on further increasing the nitrogen flow rate. On the other hand, for the films prepared from the targets with the Ta content in the range of 5–15wt%, the hardness increases whenever Ta and N contents increase. These effects are clearly illustrated by 3-D and contour hardness and Young's modulus maps in this paper.

The modulating force method in nanoindentation gives a direct measure of contact stiffness, and being insensitive to drift, allows the accurate observation of creep in small indents to be carried out over long time periods. We present results for a range of metals at room temperature. Strain rate indices similar to those for macroscopic creep are found. Reverse creep occurs for step unloading greater than about half the starting load. For electropolished tungsten, we find quite different behaviour before and after the sudden pop-in. Afterwards, creep is as in other metals, but beforehand, it is essentially zero. The slight changes of stiffness observed at the very smallest loads are due to diffusion of adsorbed surface films into the contact zone. Our results show that the dislocations nucleated and multiplied at pop-in provide the mechanism of creep.

Nanoindentation is a common technique for measuring the mechanical properties of thin films. Here, we address the potential measurement errors caused by pile-up when soft films deposited on hard substrates are tested by nanoindentation methods. Pile-up is exacerbated in soft film / hard substrate systems because of the constraint the substrate exerts on plastic deformation of the film. To experimentally examine pile-up effects, aluminum films with thicknesses of 240 and 1700 nm were deposited on hard glass substrates and tested by standard nanoindentation techniques. The aluminum/glass system is interesting because the film and substrate have similar elastic moduli; thus, any unusual behavior in the nanoindentation results may be attributed to differences in the plastic flow characteristics alone. A detailed scanning electron microscopy examination of nanoindentation hardness impressions in the film revealed that common methods for analyzing nanoindentation data underestimate the true contact areas by as much as 80%, which results in overestimations of the hardness and modulus by as much as 80% and 35%, respectively. The sources of these errors and their influence on the measurement of hardness and elastic modulus are discussed, and a simple model for the composite hardness of the film/substrate system is developed. The model could prove useful in measuring the hardness and elastic modulus of soft-film / hard substrate systems when it is not possible to make indentations shallow enough to avoid the substrate influences.

CaF2 single crystals are interesting substrate materials for deposition of thin films. Its structure is cubic and it cleaves on {111} planes. CaF2, whose hardness has been reported to be 4 on the Moh's scale, is plastic and soft. In this study, the mechanical properties such as hardness(H) and Young's modulus(E) of single crystal CaF2 mineral were measured by using a nanoindenter with a Berkovich indenter normal to (100) and (111) planes. A normal indentation size effect (ISE) in accordance with the traditional power law and the proportional specimen resistance model (PSR) of Li and Bradt [1] was observed. The values of E and H on (100) plane are larger than those on (111) plane and these values on both planes decrease with increase in time during the hold segment. The effect of displacement rate on mechanical properties of (100) and (111) surfaces is also studied.

In-situ transmission electron microscopy (TEM) was performed to study grain growth and dislocation motion during temperature cycles of Cu films with and without a cap layer. In addition, the substrate curvature method was employed to determine the corresponding stresstemperature curves from room temperature up to 600°C. The results of the in-situ TEM investigations provide insight into the microstructural evolution which occurs during the stress measurements. Grain growth occurred continuously throughout the first heating cycle in both cases. The evolution of dislocation structure observed in TEM supports an explanation of the stress evolution in both capped and uncapped films in terms of dislocation effects.

A method is described where tensile tests can be performed in situ on micromachined structures. The testing equipment consists of a testing unit mounted on a micromanipulator in a Scanning Electron Microscope (SEM). The fracture loads of micromachined beam structures made from thick and thin film polysilicon as well as from electrodeposited nickel and nickeliron alloy were measured, and the fracture strengths then calculated via measurements of the test structures’ initial cross-sectional areas. The statistical scatter of the polysilicon fracture strength values were evaluated by Weibull statistics. The mean fracture strength and the Weibull modulus, a measure of the scatter, were obtained

Constant loading rate/load indentation tests (1/P dP/dt) and constant rate of loading followed by constant load (CRL/Hold) indentation creep tests have been conducted on high purity electropolished indium. It is shown that for a material with a constant hardness as a function of depth, a constant (1/P dP/dt) load-time history results in a constant indentation strain rate (1/h dh/dt). The results of the two types of tests are discussed and compared to data in the literature for constant stress tensile tests. The results from the constant (1/P dP/dt) experiments appear to give the best correlation to steady-state uniaxial data.

We have employed the infrared photoelasticity (PE) method to study the stress distribution in Si substrates under patterned thin film structures such as thermal oxide layers partially covered by metal films and oxide layers with long trench openings. It is demonstrated that a lot of information on the two dimensional stress distribution in the substrate under patterned thin film structures can be obtained from PE experiments. The capability, limitation, and further development of the PE method for semiconductor applications are discussed.

Nonlinear mechanical properties of layered systems of Ta2O5 and TiO2 films deposited on fused silica by reactive evaporation (RE), reactive ion plating (IP) and spin coating (SC) are investigated by means of an ultrasonic technique. The coatings with thickness of 100 nm possess differences in density and crystal structure, due to the different deposition conditions. The nonlinear acoustic response of the film/substrate systems depends on film material. Differences are observed in respect to film density as obtained by the alternate deposition methods. The origin of the differences in nonlinear acoustic response of the samples is discussed. The results are correlated to adhesion properties of the films determined by a scratch-test method.

The Young's modulus and density are analyzed for 280 and 500 nm thick TiO2 layers deposited by reactive evaporation (RE) and ion plating (IP) by means of surface acoustic wave spectroscopy (SAWS) and grazing incidence X-ray reflection spectroscopy (GIXR). The layers are amorphous or polycrystalline, and have densities between 2.9 g/cm3 and 3.8 g/cm3, depending on the deposition conditions. The measured Young's moduli vary between 65 GPa for RE films and 147 GPa for IP layers. They are independent of film thickness, but correlate with the density. A change of the Young's modulus due to the phase transition from amorphous to anatase is described, which occurs at temperatures above 210°C.