ICMCTF2016 Session TS4-1: Plasma Diagnostics and Modeling

Non-equilibrium atmospheric pressure plasmas are currently receiving an increased interest for a variety of applications ranging from environmental remediation, material processing and synthesis to envisioned medical applications such as wound healing.

The high pressure conditions coinciding in many cases with complex non-equilibrium chemistry due to the presence of specific precursors for deposition applications form a formidable challenge for plasma diagnostics. In addition, while low pressure plasmas are diffuse, atmospheric pressure plasmas are often filamentary in nature. The existence of these filaments is correlated with strong gradients in plasma properties both in space and time that can significantly affect the plasma chemistry. As these filaments are often randomly appearing in space and time, it poses great challenges for diagnostics often requiring the stabilization of the filament to study the in situ plasma kinetics.

In this presentation, I will review the current status of plasma diagnostics focusing on reactive oxygen and nitrogen species production as important in the field of plasma-bio interactions and surface activation of plastics. This will be extended to more complex (precursor) chemistries used in plasma deposition including an outlook to the challenges ahead in this particular area.

In recent years, various plasma waves and instabilities have been discovered in magnetron discharges operating in the high power impulse magnetron sputtering (HiPIMS) mode but later also for more conventional direct current magnetron sputtering (DCMS). The overwhelming majority of those studies were done on small, round, planar magnetrons, and there is a need to investigate plasma properties of linear magnetrons as they are also prevalent for many coatings applications. In this work, we report on plasma instabilities studied on various target materials in a relatively small (25 cm), commercial linear magnetron, including non-reactive and reactive and non-reactive DCMS HiPIMS situations. This work focuses on the plasma behavior on the linear racetrack and at the corner of the racetrack in terms of two types of low frequency oscillations: azimuthal and axial, thereby providing insights relevant to industrial magnetron sputtering.

In sight of replacing hard anodizing for the growth of protective coatings on light alloys (Al, Mg, Ti), Plasma Electrolytic Oxidation (PEO) is encountering an increasing interest. Indeed PEO does not only prevent from using the hazardous compounds required by anodizing, it also allows improving the performance of the grown coatings in term of maximum thickness, crystallinity, micro-hardness, corrosion resistance and growth rate. However, the high energy consumption is viewed as a serious constraint of the industrial development of the process. Indeed Mohedano et al. [1] have estimated the energy required for anodizing to be as low as few 0.1 kW.h.m^-2.µm^-1 while several kW.h.m^-2.µm^-1 are required for PEO. However one should keep in mind that, unlike anodizing, PEO involves local heating which allows obtaining fully crystallised coatings made of high-temperature phases (α-Al₂O₃, γ-Al₂O₃) on relatively low melting point substrates (Al). If one assumes that a similar heat treatment would be available for anodized parts, the energy consumption in order to bring only the coating to the transition temperatures of the aforementioned phase would be in the range of the kW.h.m^-2.µm^-1, that is to say, the same order of magnitude as PEO energy consumption.

This promising conclusion has encouraged the present contribution that aims at establishing the energy balance at the scale of a single micro-discharge and, therefore, identifying the major energy loss. To do so, electrical measurement synchronised with ultra-high speed video imaging has been implemented in the single discharge set-up developed by Dunleavy et al. [2]. The electrical measurements have shown an energy input around 1mJ per discharge. The video recordings have shown the ignition of a plasma discharge followed by a bubble that grows up to 500 µm within ~100 µs. The Rayleigh-Plesset equation suggests that the bubble growth is driven by an overpressure of 1-2 bar. It then appears that the vaporisation of the electrolyte in contact with plasma is the main energy dissipation. The other contribution such as coating production and heating, plasma ignition and sustaining, or conduction through electrolyte are in the range of magnitude of some tens of µJ i.e few percent of the energy input.

The complete energy audit will be detailed and ways of improvement of PEO energy efficiency will be proposed.

[1] Mohedano et al.Surf. Interface Anal. (2015) DOI 10.1002/sia.5815

[2] Dunleavy et al. Appl. Surf. Sci 268(2013) 397

Since 2010 the Institute of Electrical Engineering and Plasma Technology at the Ruhr-University Bochum has been investigating a large-area, capacitively coupled multi-frequency plasma source (MFCCP) for deposition of ceramic layers.

The MFCCP is powered by three frequencies: 13.56 MHz, 27.12 MHz and 60 MHz. The two lower frequencies are driven in a phase-locked mode with an adjustable relative phase shift. By varying the phase shift and adjusting the power of the two lower frequencies for a constant self-bias at the target, the ion energy onto the grounded substrate can be changed very accurately without any effect on the ion flux. This decoupling of both parameters is known as the electrical asymmetry effect. By applying the power of 60 MHz a high plasma density and consequently a high ion flux can be generated.

The main research objective is to understand the coherences between the plasma parameters and layer properties. In regard, the particle transport from target to substrate is investigated in detail. While the sputter process at the target is simulated using a Monte Carlo code (TRYDIN), the interaction of the sputtered particles with the neutral gas is analyzed applying a direct simulation Monte Carlo code (DSMC). For comparison, space resolved densities of electrons and sputtered species measured by optical emission spectroscopy with Abel Inversion are presented.

In addition the decoupling of ion energy and ion flux is used to determine their influence on the properties of the deposited aluminium nitride layer (e.g. thickness, stress, structure). Thereby the ion flux and ion energy distribution (IED) is measured with a retarding field energy analyzer on the grounded substrate. These values will be compared with IEDs of an electromagnetic and kinetic PIC simulation of the MFCCP.

Finally, characterization of the deposited layers will be correlated with the ion energy and ion flux.

The MFCCP is a sub-project of the collaborative research centre SFB-TR 87 funded by the German Research Foundation (DFG). We thank Dr. Carles Corbella and Prof. Dr. Julian Schulze for supporting the IED measurements and the inspiring discussions. We also thank the Institute of Materials (coating analytics) and the Institute for Theoretical Electrical Engineering at the Ruhr-University Bochum for providing success for presenting the corporate results.

Reactive plasma deposition (RPD) is a commercially available ion plating system with dc arc discharge for thin film deposition. The RPD method enables us with the growth of transparent conductive oxide (TCO) films exhibiting low electrical resistivity and high visible transmittance, such as indium tin oxide (ITO) and gallium doped zinc oxide (GZO) films, at a low substrate temperature (T_s). The feature of RPD technique is to achieve TCO films with high quality at a high growth rate at any T_s compared with that of the conventional sputtering technique. We have investigated the factors limiting the deposition rate of GZO films deposited on glass substrates by RPD. As a qualitative research necessary to solve them, very recently, we have been developing an original analytical approach to clarify the characteristics of the incident particle flux that strongly depends on the deposition parameters; oxygen (O₂) gas flow rates (OFRs) introduced into the deposition chamber and discharge current (I_D). In our previous work on ITO films, we demonstrated a quantitative analysis of ionization rates of incident particles, such as neutral particles and ions, using a mass-energy analyzer (Hiden, EQP300) and a Langumuir probe during the deposition.

We deposited GZO films on glass substrate at a T_s of 200 °C by RPD. The evaporation source was sintered ceramic ZnO containing the Ga₂O₃ contents of 4 wt.%. The flow rate of Ar gas was at 140 sccm. The OFRs and I_D were varied from 0 to 20 sccm and from 100 to 140A, respectively. We found that the growth rates tend to increase with increasing I_D at any given OFR. For lower OFR values of 0 to 15 sccm, we observed that the growth rates were limited by the above two deposition parameters. With further increasing OFRs up to 20sccm, I_D became the dominant factor of the growth rates. At any given I_D, we found a little difference in the growth rates between OFRs values of 15 and 20 sccm. The analysis shows that the ratios of oxygen species such as O, O⁺ and O₂⁺ flux to zinc species such as Zn and Zn⁺ flux are less than 1 at any OFRs and I_D. The important finding obtained by the analysis on the incident particle flux combined with the experimental results on the film growth rates is as follows: The behavior of the growth rates was determined by the incident fluxes of the minor oxygen-related particles. The above findings imply the strong relationship among growth rates, a microstructure with point defects associated with O species and electrical and optical properties of GZO films. We will discuss it in more detail.