ICMCTF2012 Session B3-1: Ion-Surface Interactions in Film Growth and Post-Growth Processes

Transition metals and their nitrides are widely used in the industry as protective coatings because of their excellent tribological properties. Among them, tantalum (Ta) demonstrated to provide high wear protection [1] as well good corrosion resistance and biocompatibility [2]. However, there are only a few studies up to date about tribological analysis of coatings based on this metal. This could be due to the difficulty in evaporating Ta in Physical Vapor Deposition systems because its relatively low thermal and electrical conductivity and high fusion temperature.

In this work three different evaporation techniques are used and compared. All of them are varieties of Magnetron Sputtering: pulsed DC, HiPIMS (High Power Impulse MS) and MPP (Modulated Power Pulse). With these last two techniques there is a potential improvement on coating quality because during the evaporation process we have peaks of high power density increasing the ionization, density of the coating, adhesion and wear resistance [3].

We have seen that h igh power techniques (HiPIMS and MPP) enhance the hardness of TaN coatings compared to conventional Pulsed DC MS. The interface TaN with stainless steel is denser with MPP technique and better adhesion of the coating is achieved. In corrosion tests all samples show passive behavior and low corrosion currents in the anodic branch. TaN by HiPIMS showed the highest corrosion resistance which increased when increasing the immersion time due to its denser microstructure and the stable electrochemical behavior of its passive film formed in PBS solution.

[1] J. Esteve, E. Martínez, A. Lousa, F. Montalà, L.L. Carreras, “Microtribological characterization of group V and VI metal-carbide wear-resistant coatings effective in the metal casting industry”, Surf. Coat. Technol. 133–134 (2000) 314–318.

[2] J. Black, "Biological performance of tantalum". Clin. Mater. 16 (3): 167–173. (1994). doi:10.1016/0267-6605(94)90113-9.

[3] A.P. Ehiasarian, book chapter: “Fundamentals and Applications of High Power Impulse Magnetron Sputtering”, Plasma Surface Engineering Research and its Practical Applications, p. 35 – 86 (2007), ISBN 978-81-308-0257-2.

Ions play an prominent role during reactive magnetron sputtering. Their influence can be quite explicit as for example when a substrate bias is applied during thin film growth. However, ions can also play a more hidden role. This paper aims to give an overview of the different processes in which ions play a key role.

The first, and most obvious during magnetron sputtering, is of course the sputter process as such. Although it seems straightforward to describe this, fundamental issues as the angular emission profile, compound sputter yield hampers a quantitative description of the deposition profile, and therefore the deposition rate at the substrate[1].

A similar question exists about the role of ions during the sustaining mechanism of the magnetron discharges. In recent years, substantial progress has been in the understanding of the behaviour of the electron emission yield when oxidizing the target[2]. As the latter behaviour also influences the emission of negative oxygen ions, a good understanding is needed because high energetic negative oxygen ions affect in an important way thin film growth. A few examples of this behaviour will be given[3].

As the ions bombard the target, they also become implanted. For inert gas atoms, their influence is minor. However, reactive ion implantation is an important pathway in the poisoning mechanism during reactive magnetron sputtering[4]. The paper will discuss the latest trends in the modelling of this process.

Finally, ions can be used as a tool to influence the thin film growth. As they are charged species, their energy can easily be influenced by biasing the substrate. Moreover, they can also be guided towards the substrate. This approach becomes even more interesting when most of the metal species are ionized as in HIPIMS plasmas. However, when studying thin film growth, one must realize that not only the ions are important, and other species play also their role. This will be discussed in the context of the characterisation of the different particle fluxes from the plasma towards the substrate [5].

[1] F Boydens, W P Leroy, R Persoons and D Depla, Submitting for publication to Physica status Solidi a

[2] D Depla, S Mahieu, R.De Gryse, Thin Solid Films 517 (2009) 2825

[3] S Mahieu, WP Leroy, K Van Aeken, D Depla, JAP 106 (2009) 093302

[4] D. Depla, X. Y. Li, S. Mahieu,K. Van Aeken,W. P. Leroy, J. Haemers, R. De Gryse, A. Bogaerts, JAP 107 (2010) 0113307

[5] S Mahieu, WP Leroy, K Van Aeken, M Wolter, J Colaux, S Lucas, G Abadias, P Matthys, D Depla, Solar Energy 85 (2011) 538

The incorporation of growth-induced defects from energetic deposited particles during sputter-deposition, known as the “atomic peening” effect, is revisited in low-mobility material thin films by combining in situ and real-time wafer curvature and ex situ X-ray diffraction techniques.

A series of metastable Mo_1-xSi_x solid solution films were deposited by magnetron sputtering at low pressure (0.12 Pa) in Ar plasma discharge onto crystalline (110) Mo template layers. Unbalanced magnetron configuration and substrate bias voltages up to -120V were used to increase the contribution of energetic ions to the total deposited energy. The stress evolution was monitored in real-time using a multiple-beam optical stress sensor (MOSS) designed by kSA and implemented in the deposition chamber. The stress-field was determined from XRD using the sin²ψ technique adapted for the case of textured/epitaxial layers. Post-growth ion irradiation using 310 keV Ar⁺ ions at low dose(< 0.2 dpa) were carried out to identify the nature of point-defects and associated stress field.

Compressive stress evolution due to atomic peening is observable only above a first critical energy threshold. The stress-field appears in that case fully biaxial. Grain-size dependence of stress confirms that defect creation is confined to the grain boundaries. Further increase of the deposited energy, above a second threshold, results in the creation of additional volume defects inside the grains, i.e. expansion of the unit cell. Examination of sin²ψ plots evolution on irradiated films shows that defects incorporated in the grain boundaries are stable, while those created inside the grain are highly unstable.

In a deposited-energy/composition space diagram, these thresholds depict the existence of biaxial or hydrostatic stress domains related to these two distinct defect creation mechanisms. The strong variation of the critical energy thresholds with the Si content points out the sensitivity of defects creation to the intrinsic properties of the material.

Temperature and substrate bias play a key role in the structural evolution of aluminium oxide (Al₂O₃) during the deposition process. On the one hand, crystallinity depends on the mobility of the particles in the growing film, which is influenced by the substrate temperature. On the other hand, correct tailoring of the substrate bias allows to selectively control the energy distribution function of the ions impinging on the substrate (IEDF). Thus, film characteristics such as hardness, adhesion, crystallinity, or wear resistance can be controlled. In this work, manipulation of the substrate bias is performed by variation of the frequency, amplitude and shape of the applied bias signal. We will show how different bias functions affect the shape of the IEDF while keeping the mean energy constant at 55 eV, and how this in turn has a clear influence on the crystallinity of the film.

The films are deposited in a RF magnetron discharge, driven by 13.56 MHz and 71 MHz. The target is mounted on the powered electrode and a silicon substrate is placed on a biased electrode at the opposite side.

Our previous experiments have already shown a preferred orientation in the film when using a rectangular pulsed bias with 1 MHz frequency and an on-time of 5 µs. Here, we will present a thorough investigation of how variations in frequency and duty cycle in the bias signal affect the IEDF and film properties for constant mean energy and an on-time of 5 µs. Additionally, the influence of growth temperature (500 °C, 550 °C and 600 °C) will be shown.

Film characterization is performed using FTIR and XRD to determine the orientation/crystallinity of the films. The measurements are correlated with the measured and simulated IEDFs in each case. We will show how the tailoring of the IEDF through bias shape manipulation is excellent tool for controlling film structure.

The work is funded by DFG within SFB-TR 87.

In this study, extremely short pulse, whose pulse width was about 50 ns and called "nanopulse", assisted hot-filament chemical vapor deposition method was used for the diamond synthesis, and depositions at low substrate temperature below 500 °C were attempted decreasing filament temperature and using high-voltage nanopulse assist. In addition, characteristics of nanopulse plasma were investigated through optical emission spectroscopy (OES) analysis as well as time resolved optical emission spectroscopy (TROES) analysis. Spectrometer used was special and the time resolution of 10 ns was enough to trace nanopulse plasma and investigate the effects of nanopluse on diamond growth.

As a result, we had succeed in depositing diamond at the substrate temperature of 400 °C using 1500 °C filament and high-voltage nanopulse assist. Moreover, according to OES and TROES measurements of nanopulse plasma, the relations between diamond properties and deposition conditions were found. Atomic hydrogen played an important role on diamond growth and the diamond was able to be deposited with sufficient concentration of CH radicals at the afterglow of nanopulse plasma.