ICMCTF2011 Session B6-1: Application-Oriented Coating Design and Architectures

Reactive cathodic arc evaporation is a deposition method with high versatility for the design of oxide coatings. The utilization of binary composite targets in this method facilitates the synthesis of binary and ternary oxides. While the metallic ratio in the oxide coating can be adjusted by the composition of the target constituents, the chemical reactions occurring during the oxide synthesis can be controlled additionally by the oxygen flow.

It has been shown that for targets consisting of elemental powders, the target surface undergoes reactions during arc evaporation in oxygen. As a result of these reactions, solid solutions and intermetallic compounds are formed [1]. A simple empirical model based on the binary phase diagram can be utilized to predict the compounds produced at the target surface. The Al-Ni material system is utilized to demonstrate the design procedure of an oxide coating based on this model. Three different target compositions (Al₈₅Ni₁₅, Al₇₀Ni₃₀, Al₄₅Ni₅₅) were selected. The prediction of the phase formation at the target surfaces are compared with the analytical results obtained by X-ray diffraction measurements.

The influence of the material changes at the target surface on the evaporation process is discussed. The microstructure and composition of the synthesized Al-Ni-O layers are investigated by XRD, RBS and TEM. It is shown that the synthesized coatings consist either of mixtures of NiAl₂O₄ (spinel) in a nanocrystalline Al₂O₃ (corundum) matrix or of spinel and Ni-O phases. Also in this case, the phase diagram of the Al-Ni-O system can be utilized to explain the phase compositions in the synthesized oxide layer. As expected from the phase analysis, the synthesized coatings show high hardness and excellent interface matching with coatings of cubic structure.

[1] J. Ramm, A. Neels, B. Widrig, M. Döbeli, L. de Abreu Vieira, A. Dommann, H. Rudigier, accepted for publication in Surf. Coat. Technol., 10.1016/j.surfcoat.2010.08.152

Based on their specific microstructure and properties MAX phases have recently be shown to offer a substantial potential for erosion protection of gas turbine hardware. In this paper, single phase Cr₂AlC MAX phase coatings were deposited onto a nickel-base superalloy using a sintered powder target in conventional DC magnetron sputtering and high power impulse magnetron sputtering (HIPIMS) mode. Controlled phase transformation during heat treatment resulted in marked improvement of erosion resistance. The Rietveld method was used to quantify the phase transformations. Notably, the lattice parameter of DCMS and HIPIMS coatings responded differently to heat treatment. Coatings deposited at low temperatures were amorphous and could be transformed into crystalline MAX phase by heat treatment. Unlike highly textured columnar Cr₂AlC coatings sputtered at high temperatures, nearly no texture was observed in crystalline coatings transformed from the amorphous state. Local heat-treatment in vacuum triggered selective phase transformation. When aluminium was reduced in the coating during heat treatment, a continuous transition from Cr₂AlC MAX phase to chrome carbide occurred. This effect can be used to control the local properties in the coatings.

Probably, there are only few principles in materials science that are used as often as the “rule of mixture”. Based on the properties of the constituting phases this rule allows for estimating a variety of composite properties ranging from the heat capacity of ivory to effective Young’s modulus of concrete. For the calculation of the bulk modulus, for example, the only input data needed are the bulk moduli of the constituting phases and their volume fractions. This is commonly used for macroscopic composites as well as multilayers with layer thicknesses of a few nanometers. The applicability of the rule of mixture on smaller length scales, i.e. on the atomic level has not been considered. Here, the rule of mixture is extended to the sub unit-cell level. It is shown that the theoretical bulk modulus of a single phase with layered crystal structure is reliably estimated from calculated bulk moduli data of the constituents adopting the symmetry within the individual layers of the layered structure. The strength of the here-proposed model is demonstrated for M_n+1AX_n phases (M: early transition metal, A: A-group element, X: C or N, n = 1-3), M_nAl₄C_n+3 phases (M: Hf or Si, n = 1-3) as well the ionic quaternary compound KAg(CN)₂.

This strategy enables a combination of combinatorics and DFT: a database can be built up by calculating a variety of hypothetical phases constituting the building blocks of the layered structure. Suitable combinations thereof may be identified, thus reducing calculation time by some orders of magnitude. This work is a significant step towards knowledge-based materials design since “sub unit-cell composite” materials with tailor-made elastic properties can be predicted based on hypothetical phases assembling the composite.

Coatings of ternary B1-type transition metal nitrides, such as Ti-Al-N, often exhibit superior hardness and protective properties in harsh environments compared to the constituent binaries. However, the performance of such coatings is eventually limited by brittleness. The Ti-W-N materials system presents an electronic structure with alternating high and low electron density regions that lead to an enhancement of the hardness/ductility relation [1]. Analogous to Ti-W-N, the V-W-N system has been predicted to have similar qualities [2], making it a promising hard coating material candidate with increased toughness.

In this work we have experimentally studied growth and properties of single-crystal V_xW_1-xN(001) thin films, with concentrations in the range 0.04 ≤ x ≤ 0.96, grown on MgO(001) substrates by reactive magnetron sputter epitaxy (MSE). Metallic V and W targets in a pure N₂ atmosphere was used and the deposition temperature was varied from 500 to 1000°C.

For temperatures higher than 800°C, a mixture of B1 VWN and W₂N was observed by XRD. At 700°C and x ≥ 0.4 (as determined by RBS), XRD indicates a single phase B1 V_xW_1-xN (001)-oriented growth with a lattice parameter slightly smaller than that of the MgO substrate. For x < 0.4, a phase mixture of W and VWN is observed. Cross-sectional HR-TEM and selected area electron diffraction of a V_0.4W_0.6N film confirmed the single crystal cubic B1 structure. High resolution XRD reciprocal space mapping showed that x=0.4 yields a tetragonally-strained film, with a lateral (in-plane) lattice parameter a=4.212 Å, equal to that of MgO, and a transverse lattice parameter c=4.201 Å. For x=0.62, on the other hand, the film is relaxed with c=4.202 Å and a=4.192 Å. Hardness and elastic moduli data obtained through nanoindentation will be reported.

[1] D. G. Sangiovanni, V. Chirita and L. Hultman, Phys. Rev. B 81, 104107 (2010)

[2] D. G. Sangiovanni et al, 57^th AVS Symposium, 17-22 October 2010, Abstract nr. 673.

[1] D.G. Sangiovanni, V. Chirita and L. Hultman, Phys. Rev. B 81, 104107 (2010).

[2] D. G. Sangiovanni et al, 57^th AVS Symposium, 17-22 October 2010, Abstract # 673.

Titanium carbide/carbon hard coatings were deposited by a hybrid plasma process combining physical vapour deposition and plasma enhanced chemical vapour deposition at room temperature. Argon gas was used to sputter a titanium target in an RF magnetron discharge; whereas methane was used as a source of hydrogenated carbon. An additional RF plasma, independent of the magnetron plasma, was generated by a coil located between the target and the substrate. By adjusting the Ar/CH₄ ratio and the species impinging on the substrate during the deposition, films with a variety of chemical composition and microstructure were prepared. The evolution of the microstructure and the chemical composition were studied by X-ray photoelectron spectroscopy, X-ray diffraction, transmission electron microscopy and Raman spectroscopy. Titanium carbide nanoparticles of different sizes embedded in an amorphous carbon matrix can be formed depending on plasma conditions. The grain size of the different coatings were determined by using Scherrer’s formula or from a statistical study extracted from the high resolution transmission electron microscopy micrographs. The internal residual stress in the coatings deposited under different deposition conditions was calculated by the curvature method using Stoney’s formula. In addition, nanoindentation tests were performed on different samples in order to investigate the mechanical properties of the coatings. It was observed that the measured hardness varies between 15 and 30 GPa according to the carbon concentration present in the coatings. The evolutions of the internal residual stress, of hardness and of Young’s modulus are correlated to the evolution of the microstructure of the coatings which depends on the carbon content in the coatings.

The Cr/ta-C multilayers are regarded as diffusion barriers due to the formation of chromium carbides with a narrow homogeneity range at their interfaces. In order to get insights into reactive diffusion kinetics in the Cr-C system running on the nanoscale, in-situ XRD measurements were carried out at the BM 20 beam line at the European Synchrotron Radiation Facility (ESRF) in Grenoble. The interface reactions were visualized via changes of the layer thickness, interface morphology and crystallinity during a heat treatment in a temperature range between room temperature and 600°C.

The individual layer thickness and the interface roughness were obtained from X-ray reflectivity measurements, additional information about the correlation of the interface corrugations from the resonant small-angle diffuse scattering. The changes in the interface morphology were confirmed by TEM. These nanostructure parameters were complemented by the phase composition concluded from the glancing angle X-ray diffraction (GAXRD) patterns.

The effect of the initial nanostructure of the multilayers on the reactive diffusion kinetics was investigated with the aid of multilayers having different individual layer thicknesses and interface morphologies. These layers were deposited using different energies of the carbon ions during the deposition of the tetrahedrally bonded carbon (ta-C) layers. The thickness of these layers was varied between 6–10nm. The Cr layer thickness was kept almost constant at 10nm. The interface roughness between both layers varied between 5-15Å. Additionally, it was found that the amount of the sp³ bonds in the carbon layers and the crystallinity of the Cr layers changed depending on the carbon ion energy.

The embedding of information on surfaces is state-of-art for various fields of application such as product identification (e.g. manufacturer, price, use-by date), material identification (e.g. type of material, brand name, batch number), identification of persons (e.g. ID-cards, driving licenses, admittance passes) and identification of documents (banknotes, securities, tickets). In many instances, the legal authentication of a product, a material, a person or a document is required.

Public, hidden and forensic features either encoded or directly legible is used for authentication. Fabry-Perot layer stacks as information carriers provide in conjunction with imaging ellipsometry as optical read-out system all-in-one anti-counterfeiting features. Different designs are described with respect to public features such as color either in reflection or transmission, tilt effects perceptible by the human eye, hidden features in the UV or IR spectral range or even forensic features such as the ellipsometric quantities ψ and Δ as a function of wavelength and angle of incidence. In addition to interference effects within the Fabry-Perot cavity, additional scattering effects can be superimposed which affect color and brightness and depend on the source of illumination and the angle of perception.

Besides stratified Fabry-Perot stacks, also patterned layer systems have been investigated. Bar codes and data matrices are used as information carrier. Layer pattern are produced by laser writing systems either in terms of a modification of the metallic film in the Fabry-Perot cavity or by pre-structuring of the substrate-based bottom reflector. In addition to bar-code or data-matrix readers, imaging ellipsometry is used both for forensic encoding of substrate or Fabry-Perot-based information. Object-in-object features and their specific ellipsometric fingerprint are shown and discussed in dependence on the stack design.

It could be shown that physically uncloneable functions (PUF) could be realized as a result of a multi-material and a multi-parameter deposition approach as well as due to specific design features of the Fabry-Perot layer stack. Hence, they are not subject to any reverse engineering strategies. Examples of micro-structured and laser-modified Fabry-Perot layer systems are considered that may be used at all perception levels (e.g. human eye, bar code reader, imaging ellipsometry) for authentication against product counterfeiting and related areas.

In the fields of material development and material optimization more and more often coatings are used to get the wanted and needed material properties. Unfortunately most of the analysis theories and measurement procedures can't be used for such coated systems.

This talk will present some extended analysis methods for the most of the measurement procedures classically used. With these new methods it is possible to determine the correct generic material parameters like Young's modulus and yield strength for coated, viscous, gradient and microstructured systems.

Beside the analysis of the measurement data to determine generic coating parameters the simulation of the real contact situation also plays a more important role. Based on such an extended test analysis true initial failure mechanism can be derived and used for later optimization and development [1]. With such simulations with completely analytical models (www.siomec.de/filmdoctor) one can save a great amount of development time and money, because inapplicable material structures, materials (or material combinations) can be excluded before the creation of new usually expensive prototypes. Moreover, the actual search vector for the optimization procedure can be defined/narrowed much better exluding irrelevant failure mechanisms with respect to the practical application.

In the second part of the talk some ways for the simulation of real applications and practical tests (e.g. scratch tests) will be presented.