HiPIMS Discharges Research Articles

Ionized particle transport in reactive HiPIMS discharge: correlation between the energy distribution functions of neutral and ionized atoms

We investigated the transport of titanium ions produced in a reactive high-power impulse magnetron sputtering device used for titanium nitride coating deposition. Time-resolved mass spectrometry measurements of ionized sputtered atoms correlated to time-resolved tuneable diode-laser induced fluorescence (TR-TDLIF) measurements of neutral sputtered atoms were used to understand transport features. Based on ion energy distributions of Ti+, we identified four populations of ions and explore their physical origins. The signals of all ion populations decrease strongly when only 1% N2 is added to the Ar/N2 gas mixture. Time resolved mass spectrometry confirms the result reported in previous work: the fast target poisoning when nitrogen is added in HiPIMS discharges. Based on the measured energy distribution functions of Ti2+, N+, N2 +, and Ar+, we discuss the production of these ions in HiPIMS discharges. The temperature of thermalized sputtered neutral atoms determined by previous TR-TDLIF measurements evidences the physical origin of an ion population with energies lower than 4 eV. According to discharge pressure, cathode voltage, and ion type, we also discuss the physical origins of high-energy ions (>4 eV).

Characterization of transport of titanium neutral atoms sputtered in Ar and Ar/N2 HIPIMS discharges

In this work we report on the investigation of the transport behavior of Ti neutral atoms sputtered in a reactive high power impulse magnetron sputtering device used for TiN coating deposition. The time-resolved tunable diode laser induced fluorescence (TR-TDLIF), previously developed to study the transport of tungsten atoms, was improved to measure Ti neutral atom velocity distribution functions. We find that the TR-TDLIF signal has to be fitted using three Gaussian distributions, corresponding to the energetic, thermalized, and quasi-thermalized (atoms with non-zero mean velocity) atom populations. The ability to distinguish populations of atoms and to determine their corresponding deposited flux and energy may be of great interest to control film properties as desired for targeted applications. From the fitting, the vapor transport parameters (flux and energy) are calculated and studied as a function of distance from the target, pressure, and percentage of nitrogen in an Ar/N2 gas mixture. The study focuses on the effect of added nitrogen on the transport of sputtered atoms.

Correlation between mass-spectrometer measurements and thin film characteristics using dcMS and HiPIMS discharges

In this work, chromium thin films were deposited using dcMS and HiPIMS technologies. To compare these technologies, we analyzed the ion flux and the Cr coating microstructure in the same plasma conditions. Ion flux was measured with a mass spectrometer in time-averaged for both discharge and time-resolved for HiPIMS discharge. Time-averaged measurements provided important information. First, the low energetic part of the ion energy distribution function (IEDF) was similar in dcMS and HiPIMS and second the high energetic component was more prominent in the HiPIMS discharge. Time-resolved measurements showed that the high energetic part of the ion flux reached the mass spectrometer faster than the lowest part. It is only after the pulse end that most of the thermalized ions arrived and then cooled the flux. The correlation of these results with microstructure analysis shows that energetic particles induced a higher film density and a smoother surface in HiPIMS compared to dcMS discharge.

A phenomenological model for the description of rotating spokes in HiPIMS discharges

In the ionization region above circular planar magnetrons, well-defined regions of high emissivity are observed, when the discharge is driven in the HiPIMS regime. These regions are characterized by high plasma density and are often referred to as ‘spokes’. Once their mode is stabilized, these structures rotate in the E × B direction with a constant rotation frequency in the hundreds of kHz range. A phenomenological model of the phenomenon is developed, in the form of a system of nonlinear coupled partial differential equations. The system is solved analytically in a frame co-moving with the structure, and its solution gives the neutral density and the plasma density once the electron density shape is imposed. From the balance of ionization, electron loss and constant neutral refilling, a steady-state configuration in the rotating frame is achieved for the electron and neutral densities. Therefore, the spoke experimentally observed can be sustained simply by the combination of this highly reduced number of phenomena. Finally, a study of the sensitivity of neutral and plasma densities to the physical parameters is also given.

A comparison of reactive plasma pre-treatments on PET substrates by Cu and Ti pulsed-DC and HIPIMS discharges

PET web samples have been treated by magnetically enhanced glow discharges powered using either medium frequency pulse direct current (p-DC) or low frequency high power pulse (HIPIMS) sources. The plasma pre-treatment processes were carried out in an Ar–O 2 atmosphere using either Cu or Ti sputter targets. XPS, AFM and sessile drop water contact angle measurements have been employed to examine changes in surface chemistry and morphology for different pre-treatment process parameters. Deposition of metal oxide onto the PET surface is observed as a result of the sputter magnetron-based glow discharge web treatment. Using the Cu target, both the p-DC and HIPIMS processes result in the formation of a thin CuO layer (with a thickness between 1 and 11 nm) being deposited onto the PET surface. Employing the Ti target, both p-DC and HIPIMS processes give rise to a much lower concentration of Ti (< 5 at.%), in the form of TiO 2 on the PET treated surface. The TiO 2 is probably distributed as an island-like distribution covering the PET surface. Presence of Cu and Ti oxide constituents on the treated PET is beneficial in aiding the adhesion but alone (i.e. without oxygen plasma activation) is not enough to provide very high levels of hydrophilicity as is clear from sessile drop water contact angle measurements on aged samples. Exposure to the plasma treatments leads to a small amount of roughening of the substrate surface, but the average surface roughness in all cases is below 2.5 nm. The PET structure at the interface with a coating is mostly or wholly preserved. The oxygen plasma treatment, metal oxide deposition and surface roughening resulting from the HIPIMS and p-DC treatments will promote adhesion to any subsequent thin film that is deposited immediately following the plasma treatment.