ICMCTF2002 Session F1/E4-1: Mechanical Properties and Adhesion

Tuesday, April 23, 2002 8:30 AM in Room San Diego

Tuesday Morning

Start	Invited?	Item
8:30 AM		F1/E4-1-1 Finite Element Modelling Of Stress Development During Deposition Of Ion Assisted Coatings D.J. Ward, R.D. Arnell (University of Salford, United Kingdom) Ion Assisted Physical Vapour Deposited (IAPVD) Films typically have a high state of residual stress. This residual stress comprises two components: a thermal stress, which forms as the system cools to room temperature; and an intrinsic stress which is caused by the processes of deposition. Much work has been published on the tribology and mechanical behaviour of surface coatings without consideration of the residual stress. It was therefore considered desirable to develop a Finite Element Simulation to be used either as a precursor to any realistic mechanical study of the behaviour of such surface coatings, or to be used as a tool to study the effects of varying the deposition parameters. Previous experimental work has shown that the residual stress is related to deposition parameters, such as incident ion and atom fluxes and energies, and recent Molecular Dynamics studies have indicated that trapped inert gas species may play a major role in the mechanism for creation of the intrinsic stress. The F.E. simulation assumes that the processes of ion bombardment and material deposition are consecutive, but as the analysis time step tends to zero this assumption approximates the simultaneity of the processes. Suitable mathematical descriptions are employed in the bombarded region of the growing coating to simulate the macroscopic effects of the microscopic atomic collision phenomena and diffusion processes. Two finite element simulations are presented. The first is based on an analytical model, which has gained popular acceptance and this was presented in a previous year at this conference. The second builds on this to simulate wider aspects of known behavior and is presented in this follow up paper. The predicted trends of mean stress and its distribution are similar to those observed in published experimental work.
8:50 AM		F1/E4-1-2 Mechanical and Thermal Behaviour of Coating-Substrate-Systems Investigated with Parallel FE J. Leopold (Fraunhofer Institute for Machine Tools and Forming Technology (IWU), Germany); J.A. Oosterling (TNO, Netherlands); H. van den Berg (WidiaValenite, Germany); N.M. Renevier (TeerCoatings, United Kingdom) Using parallel FE simulation software, deformations, strains and stresses in coated tools caused by external and internal loads can be computed on a microscopic scale. This way, stress- and strain-values inside and between the coating layers become available. The knowledge of the stress-strain-distribution in a coating-substrate-system together with information's about the adhesive strength between the coating layers and the tool's material can be used to calculate critical stress values. This allows to predict the stability of the whole coating-substrate-system against mechanical and thermal loading. Using this approach, developing engineers are supported in the design of advanced coating-substrate systems for advanced industrial applications, like High-Speed-, High-Performance, Dry- or Micro-Machining, and in the judgement of the suitability of just established coating-substrate systems for special applications. Since both, the whole macroscopic tool (mm scale) and the microscopic coating layers (µm scale) must be included in the same geometrical simulation model, graded high-resolution FE meshes must be used. Nevertheless, the number of nodes in the 3-D computational FE grid reaches some 100 000s or million, leading to large computational time and storage requirements. For this reason, multi-processor hardware and parallel working software is used for the computations. Simulations of a multilayer (MOST, TiN, TiAlN, Al₂O₃) system with more than 2.4 million degrees of freedom have been performed with the available hard- and software within less than 1 hour time of computation.
9:10 AM		F1/E4-1-3 Determination of the Thermal Expansion Coefficient and of the Tensile Strength of a Polycrystalline TiAl₃ Layer, Formed by Solid State Reaction M Ignat (CNRS INP Grenoble, France) The interfacial reactions occuring during the ageing of electronical or micromechanical devices during operation, may drive them to failure. As a matter of fact , a solid state reaction, activated at an interface between thin films, can induce high internal intrinsic stresses, which can trigger irreversible damage mechanisms, as cracking and/or debonding. Determination of critical parameters, as for example the thermal expansion coefficient and the tensile strenght of a polycristaline film, formed by a solid state reaction, are crucial points which will allow a better understanding of the reliability of the mentioned devices. The example we shall comment, comes from Al/Ti interconnects. From residual stress determinations, we first deduced the thermal expansion coefficient of the TiAl₃, formed by solid state reaction. Then, its tensile strenght was deduced from analysing its cracking, through models describing the cracking process in thin films.
9:30 AM		F1/E4-1-4 In Situ Observations of the Stress Evolution and Delamination Kinetics of Thin Ta films on Si (100) B.L. French, J.C. Bilello (University of Michigan, Ann Arbor) Polycrystalline Ta coatings on Si (100) substrates were thermally tested, while their stress evolution and eventual delamination were observed in situ and in real-time using a recently developed synchrotron radiation technique. This method utilized white beam synchrotron (on SSRL station 2-2) Laue transmission diffraction topography coupled with simultaneous direct radiography to record the thermomechanical response of these films during testing. The observations were made using an experimental apparatus consisting of a 600°C heating device/sample holder and a portable CCD x-ray imaging/storage system with ~20 µm resolution. The system is capable of recording 30 images/sec; hence, at moderate strain rates both the morphology of delamination and delamination kinetics are readily observable. Furthermore, strain in the coatings can be calculated from changes in feature-separation on the film surface. The mode of failure and adhesion strength are correlated with the physiochemical nature of the coating-substrate interface, and also with the microstructure of the film. This work is supported, in part, by the ARO under GRANT# DAAG 55-98-1-0382, and the authors also thank USDoE for support of work performed at SSRL.
9:50 AM		F1/E4-1-5 Fracture Mechanisms in Nano-scale Layered Hard Thin Films A. Karimi, Y. Wang (EPFL - DP, Switzerland); T. Cselle (Platit AG, Switzerland) Depth sensing nanoindentation and nanoscratch testing were combined with atomic force microscopy (AFM) and electron microscopy observations to study nanoscale mechanical properties and fracture behavior of a number of Ti_xAl_1-xN_yC_1-y hard thin films. Various failure modes were activated either by normal loading-unloading or by microscratching of the samples to provide an estimation of the fracture toughness and interfacial fracture energies. By changing chemical composition and deposition conditions various nanostructured thin films including monolitically grown single layer, nanocomposite, and multilayers were coated onto the tungsten carbide-cobalt substrates. All tested films exhibit elevated mechanical properties with high hardness (38 ? 45 GPa) and modulus (500 ? 570 GPa). Under sufficiently high load indentation the formation of corner Palmqvist type radial cracks were usually observed because of small modulus mismatch between coating and substrate, good adhesion, and in particular high toughness of both substrate and films in spite of great difference in their respective hardness. Various failure modes were activated and the sequences of fracture events were determined using stepwise or continuously increasing load scratch tests. Some films were found to be more sensitive to tensile stress behind the indenter which generates regular microcracks on the scratch track. Other films in particular multilayers were appeared more susceptible to compressive stress ahead of the indenter leading to local delamination at the interface between layers, as well as to the occurrence of irregular microcracks under the contact area. The paper attempts to model these behaviors in terms of film microstructures.
10:10 AM		F1/E4-1-6 An International Round-robin Experiment to Evaluate the Consistency of Nanoindentation Hardness Measurements of Thin Films Y.-W. Chung, K.W. Lee (Northwestern University); M.P. Delplancke-Ogletree (Universite Libre de Bruxelles, Belgium); D. Yang (Hysitron Inc.); T.E. Buchheit (Sandia National Laboratories); A. Karimi (Swiss Federal Institute of Technology, Switzerland); J. Patscheider (Swiss Federal Laboratory for Materials Testing and Research, Switzerland); S.T. Lee (Ciry University of Hong Kong) We conducted an international round-robin experiment to determine the consistency of nanoindentation hardness measurements of thin films among six different laboratories, using three different samples. These samples were chosen to present a challenge of indenting at small loads (micro-Newton range). They were: 250 nm thick TiN, 700nm thick TiC, and 500nm thick TiB₂/TiC multilayer coatings (each layer being 3 nm thick), prepared at Northwestern University using magnetron sputtering. Each research team was free to use whatever nanoindentor and analysis methods at its disposal. Researchers other than those from Northwestern University were given only the thickness of each thin film sample, but not its composition. In addition, other than passing the samples, researchers were not to exchange data until the end of the round-robin experiment. This round-robin experiment demonstrates that at least for the hardness range of interest (15 to 35 GPa) and using well-documented procedures and analysis methods, all laboratories give essentially the same answer, with a statistical spread of about ±10%. These procedures and analysis methods will be presented. Results will also be presented on the effect of surface contaminants and the role of substrates. These factors are especially important in measurements on superhard coatings.
10:30 AM	Invited	F1/E4-1-7 Surface Acoustic Wave Techniques to Characterize Thin Film Mechanical Properties D.C. Hurley, V.K. Tewary (NIST) A variety of methods have been developed to measure the mechanical behavior of thin films. Here, we describe a relative newcomer to the field called surface acoustic wave (SAW) spectroscopy. SAW spectroscopy is a nondestructive approach that may be more suitable than other methods in some applications. The technique involves the generation of SAWs by a line-focused, pulsed laser. After propagating several millimeters across the sample surface, the SAWs are detected by optical or piezoelectric sensors with a wide bandwidth (typically several hundred megahertz). The frequency dependence of the SAW phase velocity - the SAW dispersion relation - is determined by measuring the wave displacement at two or more propagation distances. Quantitative values for film mechanical properties like Young’s modulus or Poisson’s ratio are obtained by comparing the dispersion data to analytical models for wave propagation in layered systems. We have recently extended the models to enable analysis for anisotropic film properties with an inversion algorithm based on the elastodynamic Green’s function. To illustrate the versatility of SAW techniques, we review results for specimens with a range of mechanical properties (nanoporous silica to diamond-like carbon) and film thicknesses (tens of nanometers to several micrometers). Where possible, results will be compared to those obtained by other techniques such as instrumented indentation. Advantages and limitations of SAW techniques will be discussed, as will measurement issues such as uncertainty and precision.
11:10 AM		F1/E4-1-9 Dynamic Mechanical Analyses of Thin Polymer Films K.J. Wahl (U.S. Naval Research Laboratory); S.A. Syed Asif (Hysitron, Inc.) Recent advances in atomic force microscopy (AFM) and nanoindentation enable examination of surface mechanical properties of ultrathin films and compliant materials with far greater resolution and accuracy than ever before. In our laboratory, we have implemented dynamic mechanical analyses of nanoscale contacts using a hybrid nanoindenter, coupling depth-sensing nanoindentation with AFM positioning capabilities. This combination allows surface sensitive, quantitative mechanical properties measurements of nanostructures and thin films, at a single point as well as while scanning. We illustrate these expanded capabilities with several examples: 1) a dynamic nanoscale Johnson-Kendall-Roberts (nano-JKR) adhesion test, and 2) scanning nanomechanics.¹ The nano-JKR test allows study of processes that occur during the formation and breaking of adhesive contacts with diameters smaller than the optical limit, and can be used to measure dynamic visco-elastic properties including loss and storage moduli, adhesion energy, etc. Scanning nanomechanics provides a means of directly imaging mechanical response and properties with sub-micron spatial resolution. We will discuss how these new capabilities can be used to measure near surface mechanical properties and test the models and limits of continuum mechanics. ¹ S.A. Syed Asif, K.J. Wahl, R.J. Colton, and O.L. Warren, J. Appl. Phys. 90 (2001) 1192.
11:30 AM		F1/E4-1-10 Measuring the Adhesion of Sol-Gel Coatings to a Ductile Substrate by an Indentation-based Method Y. Xie, H.M. Hawthorne (National Research Council Canada) The adhesion of sol-gel coatings to underlying substrates is a critical property of the coatings. However, measuring the adhesion can be very difficult because some sol-gel coatings tend to fail cohesively before a desired adhesive failure can be induced at the coating-substrate interface. In this study, an indentation-based method was used to measure the adhesion of sol-gel coatings to ductile substrates. A hard conical indenter or a hard sphere was driven through a sol-gel coating and into the underlying substrate. Extensive plastic deformation of the substrate caused an annular crack to develop at the coating-substrate interface. The radius of the annular crack was then measured and the interface toughness calculated from crack radius values at different indentation loads, along with coating thickness, residual stress and mechanical properties of the coating and substrate.
11:50 AM		F1/E4-1-11 New Strength Evaluation of Thin Films on Substrates in Terms of Adhesion and Cohesion S. Kamiya, H. Nagasawa, M. Saka, H. Abé (Tohoku University, Japan) Adhesion of hard coatings on substrates has been most commonly evaluated by scratching tests, where the vertical load required to break off the coating by scratching its surface is supposed to indicate the strength. However, it is also well-known that the critical load obtained by scratching test does not always indicate the correct order of practical adhesion of coated equipment under individual operation. Fracture behavior of coatings in scratching tests is generally quite complex and can be classified into even dozens of "typical" appearances. This complexity and resulting ambiguity of test results should be coming from various possible combinations of properties of films, substrates and their interfaces each of which influences the fracture behavior of systems made of these three components. A new evaluation for the strength of thin films on substrates is proposed in this report, which simplifies the current complex situation mentioned above. The toughness of films and their interfaces to substrates, that is to say the cohesive and adhesive toughness of coatings, is independently measured by using the technique recently developed by the authors. Practical data are obtained for the case of a number of systems such as various anti-abrasive coatings on cutting tools. Conventional scratching tests are also performed on the same systems. The results are compared in two methods of evaluation. As a result, various appearances of fracture in the scratching traces can be simply explained by the separate knowledge of adhesion and cohesion of the films. Moreover, the influences of film thickness and/or shape of stylus which is always subjected to the argument for the case of conventional tests can be completely eliminated. The potential ability and applicability of new evaluation is further discussed, which could give a new scientific standard for the strength of thin films on substrates.