Elsevier

Wear

Volume 264, Issues 9–10, 10 April 2008, Pages 885-892
Wear

The influence of metallic interlayers on the adhesion of PVD TiN coatings on high-speed steel

https://doi.org/10.1016/j.wear.2006.11.053Get rights and content

Abstract

In nearly all applications the adhesion of the coating to the substrate is crucial for the components performance and length of life. To enhance the adhesion it is common to use a metallic interlayer, most often titanium. In this study seven different metallic interlayers, namely W, Mo, Nb, Cr, Ti, Ag and Al, have been evaluated with respect to their influence on the adhesion of PVD TiN coatings to polished high-speed steel, ASP 2060. The purpose of this work is to investigate how some physical properties of a metal affect its capability to function as an adhesion interlayer. Samples were prepared using dc magnetron sputtering for the interlayer and reactive dc magnetron sputtering for the TiN coating. The deposition process included both pre-treatments and in situ treatments of the substrate surface in order to eliminate possible contaminations. The adhesion of the coatings was investigated with two different methods: scratch testing and Rockwell adhesion testing. The results indicate that differences in hardness between the metallic interlayers influence the practical adhesion more than differences in E-modulus. Furthermore, in order to optimize adhesion, the hardness of the interlayer should be close to the hardness of the substrate. It was also suggested that stresses, both in the TiN coating and in the metallic interlayer, affect the adhesion properties negatively. In addition, the necessity of interlayer in TiN on HSS can be questioned as the reference samples, without interlayer, showed adhesion properties comparable to the highest ranked interlayer containing samples in our assessment.

Introduction

Due to its extreme hardness, high thermal and chemical stability and low electrical resistivity TiN is frequently used as coating material in the industry since a few decades. The field of application is wide and varies from hard protecting coatings on mechanical tools to diffusion barriers in the micro-electronics industry. Even though other coatings are present on the market today, TiN is still used in certain areas and we consider it a suitable modeling material for this study. In the majority of the applications, a crucial factor for the length of life and the performance of the coated component is the coatings adhesion to the substrate. In areas where the coated surface is exposed to mechanical wear, flaking of the coating can lead to an aggravated wear situation. The resulting effect is not only that the underlying substrate surface becomes exposed; the flakes can also work as abrasive particles on the remaining surfaces. The desired course of event is instead a gradual wear of the coating during the tools total lifetime. A prerequisite for this is a good adhesion to the underlying material.

Mittal distinguish between three different forms of adhesions [1] namely; (i) fundamental adhesion, (ii) thermodynamic adhesion and (iii) practical adhesion. Fundamental adhesion is defined as the sum of all molecular and atomic interactions across the interface between the coating material and the substrate material. Thermodynamic adhesion signifies the change in free energy when an interface is formed (or separated). Practical adhesion is described as the force required to remove the coating from the underlying substrate irrespective to the locus of failure. In the present work the practical adhesion is the most interesting one because that is the adhesion measured experimentally.

Practical adhesion can be referred to as a function of fundamental adhesion and ‘other factors’. There are numerous such ‘other factors’, e.g. stresses in the coating, thickness and mechanical properties of the coating, mechanical properties of the substrate, deformation work consumed by plastic deformation and viscous dissipitation, mode of failure and the mode and rate of applying the force or the energy to detach the coating, i.e. the technique used for adhesion measurement. In other words, it is not only physical, mechanical and chemical features of the coating-substrate system that contribute to the result of an adhesion measurement but also the measurement technique itself. This is important to keep in mind when discussing measured adhesion properties. Accordingly, it is always best to use the adhesion measurement technique whose mode and rate of applied force best corresponds to the situation in the intended application.

In order to improve adhesion there are mainly three things to aim for:

  • Low energy in the interface; which is related to the micro-structural match, rather than miss-match, between substrate and coating.

  • The creation of strong and stable chemical bonds between substrate and coating.

  • A low stress gradient in the interface; which improves the ability to sustain externally imposed shear stresses in the interface. Detrimental gradients can, for example, arise from differences in thermal expansion of substrate and coating.

Several techniques are used for improving adhesion. These can be divided into three different categories: (i) pre-treatment—cleaning of the substrate prior to insertion in the deposition chamber, (ii) in situ treatment—pre-heating and sputter cleaning of the substrate, and (iii) interlayer deposition—a thin layer of a material, usually a metal, deposited between the substrate and the coating. The idea of pre-treatment and in situ treatment is to remove contaminations, such as grease, oxides, etc., from the substrate surface to increase surface reactivity that facilitates the formation of strong chemical bonds. The purposes of applying an interlayer are to minimise stresses in the interface and/or to dissolve contaminations.

To improve the adhesion of TiN and similar coatings it is common to use a thin layer of titanium or in some cases chromium as an interlayer. The reason for using these metals is to a large degree process related in that it is convenient to build a nitride and a metallic interlayer using the same metal. But to be fair, titanium does have good properties in terms of governing adhesion which has been explained by two factors [2]: (i) titanium, which is a reactive metal, dissolves surface contaminants such as oxides, (ii) an interlayer of titanium can lead to an improved match between the mechanical properties of the substrate and coating, which minimises the stress gradient in the interface.

This paper presents a study of seven different metallic interlayers and aims to investigate their abilities to work as adhesion enhancing interlayers. Modern PVD processes most often involve thorough cleaning which makes the requirement of contaminant solubility questionable. The work was performed with the intentions to facilitate the development and successful deposition of new coating materials, which may be prone to adhesive failure, and to look at the possibility to further improve the adhesion of TiN. The metals investigated were W, Mo, Nb, Cr, Ti, Ag and Al and they were chosen to cover a large variation in properties such as hardness, heat expansion, E-modulus and crystal structure.

Section snippets

Materials

The substrates used were high-polished plates (20 mm × 20 mm × 3 mm) of ASP2060, a high-speed-steel (HSS) produced by ERASTEEL Kloster AB. HSS was chosen as substrate since it is a common material to be coated with PVD coatings in industrial applications. The interlayer metals chosen originate from the groups 4, 5, 6, 11 and 13 in the periodic table. All but one (Al) belongs to the transition metals of the d-block. Three of these: Cr, Mo and W, are all in group 6, i.e. they have the same number of

Coating morphology

All deposited TiN coatings have a columnar structure similar to the one shown in Fig. 3. The thicknesses of the TiN coatings were 3.1 ± 0.5 μm and the thicknesses of the metallic interlayers were all within the range 100–150 nm.

Scratch testing

Two identical scratches were made in each sample. The critical loads for adhesion failures caused by scratching are shown in Table 3. These loads represent adhesion related failures. Minute cracks were occasionally seen at lower loads but as they are related to the cohesive

Discussion

The hardness of each produced TiN coating was measured with nanoindentation. The hardness was roughly the same, varying within the margin of errors of the TiN coatings deposited without interlayers, and without any trends considering presence and type of interlayer.

Concerning the results from the scratch test it is notable that TiNR and TiNC, without interlayers, show different behaviours which results in different types of critical loads. This indicates that the TiN produced with the two

Conclusions

The following conclusions are made regarding how the metallic interlayers W, Mo, Nb, Cr, Ti, Ag and Al affects the practical adhesion of TiN to the high-speed-steel ASP 2060, when evaluated with scratch testing and Rockwell adhesion testing:

  • Mo and Nb both give the best adhesion closely followed by the more conventionally used Ti and Cr.

  • The elastic modulus of the metallic interlayer does not affect the adhesion as much as its hardness.

  • In order to get good adhesion the thermal expansion

Acknowledgement

The authors are grateful to Erasteel Kloster AB for providing the substrate material.

References (6)

There are more references available in the full text version of this article.

Cited by (164)

View all citing articles on Scopus
View full text