P.J. Rudnik

High rate reactive sputtering in an opposed cathode closed-field unbalanced magnetron sputtering system

Abstract An opposed cathode sputtering system was constructed with the ability to coat parts with a size up to 15 cm in diameter by 30 cm long, but initial trials with this system revealed very low substrate bias currents. When the AlNiCo magnets in the two opposed cathodes were arranged in a mirrored configuration, the plasma density at the substrate was low, and the substrate bias current density was less than 1 mA cm-2. If the magnets were arranged in a closed-field configuration where the field lines from one set of magnets were coupled with the other set, the substrate bias current density was as high as 5.7 mA cm-2 when NdFeB magnets were used. In the closed-field configuration, the substrate bias current density was related to the magnetic field strength between the two cathodes and to the sputtering pressure. Hard, well-adhered TiN coatings have been reactively sputtered in this opposed cathode system in the closed-field configuration, but the mirrored configuration produced films with poor adhesion because of etching problems and low plasma density at the substrate.

The effect of target power on the nitrogen partial pressure level and hardness of reactively sputtered titanium nitride coatings

Target power and deposition rate have a significant effect on the properties of titanium nitride (TiN) coatings prepared by high rate reactive sputtering. As the target power is increased from 2.3 to 10.0 kW and the deposition rate from 1000 to 4800 A min-1, the Vickers microhardness (50 gf) increases from 970 to 3160 kgf mm-2 and the crystallographic orientation goes from a very strong (111) texture to a more random orientation for the TiN coatings. The partial pressures of the argon and nitrogen gases measured through the target show an apparent drop as power is applied to the titanium target. This apparent drop in pressure is actually a density reduction of the gases in front of the target due to gas rarefaction and heating. In order to maintain stoichiometry in the coatings, the partial pressure of the nitrogen reactive gas must be increased as the target power is increased to offset the N2 density reduction and to compensate for the increase in density of the titanium atoms.

The effect of N2 partial pressure, deposition rate and substrate bias potential on the hardness and texture of reactively sputtered TiN coatings

Abstract The N 2 partial pressure, target power and substrate bias potential all have a significant effect on the hardness and crystallographic orientation of TiN coatings prepared by high-rate reactive sputtering on cemented carbide substrates. Variations in both the substrate bias potential and the target power had a strong affect on the hardness and texture of the TiN coatings. Vickers microhardness values (50gf) of the TiN coatings varied between 970 and 3290 kgf mm -2 as the bias was stepped between 0 and –200 V and the target power between 2.3 and 10.0kW. The texture coefficient of the TiN coatings with a constant –100 V bias was strongly (111) at the low power level of 2.3 kW, but the texture became random as the power was increased to 4.0 kW and above with the same –100 V bias. At a fixed power level of 10 kW, the orientation of the TiN coatings changed from (111) at zero bias to (200) between –20 and –100 V, to strongly (220) at –150 and –200 V. The hardness of TiN coatings with N/Ti ratios of 0.70 - 1.01 produced by varying the N 2 partial pressure during deposition had a narrow range between 3140 and 3400 kgf mm -2 , but the texture coefficient varied from random to strongly (220). the adhesion of the TiN coatings was measured with the scratch test, and a critical load of 8.0 ± 0.5 kgf was the same for all samples except for the ones at the lowest target power (2.3 kW), low biases (0 and –20 V) and the lowest N/Ti ratio (0.70).

Reactive unbalanced magnetron sputtering of the nitrides of Ti, Zr, Hf, Cr, Mo, Ti-Al, Ti-Zr and Ti-Al-V

Reactive unbalanced magnetron sputtering was used to deposit eight different nitride coatings on hardened 440C stainless steel rolling contact fatigue (RCF) test specimens. The target materials used in this study were Ti, Zr, Hf, Cr, Mo, Ti0.5Al0.5, Ti0.5Zr0.5 and Ti-Al-V (the aircraft alloy Ti-6 wt.%Al-4wt.%V). All the coatings were characterized using X-ray diffraction, microhardness and scratch adhesion tests. Ti, Zr and Hf form the simple binary nitrides of TiN, ZrN and HfN, respectively, with an f.c.c. structure and a range in stoichiometry. Cr and Mo form a range of nitrides, starting with a solid solution of nitrogen in each of them, a Cr2N or Mo2N phase, and a CrN or MoN phase. The MoN phase was not conclusively formed in this work. Mo2N forms in two different phases, i.e. β and γ phases, depending on the nitrogen partial pressure. The Ti-alloyed nitrides all form an f.c.c. structure. The coatings selected for the RCF tests ranged in hardness from a low of 1900 kg mm−2 for Ti0.5Al0.5N and CrN to a high of 2800 kg mm−2 for Ti0.5Zr0.5N. The deposition conditions for forming these nitrides are given in this paper. Also, the similarities and differences, both in the deposition conditions and in the resulting properties, are reviewed.

Advances in partial-pressure control applied to reactive sputtering

Summary High-rate reactive sputtering (HRRS) requires rapid and careful control of the reactive gas partial pressure to achieve high deposition rates and to maintain stoichiometry in the coating. HRRS of TiN was first achieved with feedback control of the N 2 utilizing a differentially pumped mass spectrometer system to sense the N 2 partial pressure. Recently, a new instrument called the optical gas controller (OGC) that operates at sputtering pressures became available for sensing the partial pressure of the gases in the sputtering atmosphere. The OGC was used during the reactive sputtering of TiO x with partial-pressure control, and the TiO x hysteresis loop exhibited a wide negative slope region which is much larger than the one found during the reactive sputtering of TiN x . Partial-pressure control, compared to mass flow control during oxide reactive sputtering, leads to better control of the process and to a 3.5 times higher deposition rate for highly resistive oxide coatings. With partial-pressure control, the anatase form of TiO 2 has been produced with lattice parameters of a 0 = 3.780 A and c 0 = 9.610 A.

Reactive d.c. magnetron sputtering of the oxides of Ti, Zr, and Hf

Abstract Conventional direct current (dc) magnetron sputtering of nonconducting oxide films via reactive sputtering is extremely difficult. Without good separation of the oxygen from the target, the surface of the target rapidly becomes covered with an oxide film or “poisoned” when there is a high enough partial pressure of the reactive gas to form the desired film composition on the substrate. Breakdown of the oxide film on the target surface leads to arcing, which can damage the power supply and which can also eject droplets into the growing film. Depending on the application, these droplets can degrade performance of the oxide film. Recent advances in power supply technology with the introduction of pulsed d.c. power has overcome many of the problems in using d.c. power for the reactive sputtering of oxide films. In this study, a d.c. power supply was used along with an arc suppression unit to provide pulsed d.c. power for the reactive sputter deposition of the oxides of Ti, Zr, and Hf. As long as partial pressure control of the reactive gas was used along with the pulsed d.c. power, crystalline TiO2, ZrO2, and HfO2 and suboxides of these materials were deposited in a well-controlled manner. Arcing on the sputtering target was virtually eliminated, and there were no forbidden composition zones. Details of the deposition process are provided along with the hysteresis curves and deposition data for the oxide systems.

Effects of an unbalanced magnetron in a unique dual-cathode, high rate reactive sputtering system

Abstract In the last few years a number of investigators have studied various“unbalanced” magnetron geometries, as a means of obtaining higher ion current densities at the substrate during deposition of hard coatings; however, previous discussions have been limited to single-cathode systems. Presented here for the first time are simple plasma and magnetic field measurements that illustrate the unique opportunities afforded by using unbalanced magnetrons in a dual-cathode system. The system studied employs a pair of opposed cathodes, 38 cm × 13 cm, placed 27.5 cm apart, to coat specimens mounted on a rotational substrate holder. Comparisons are drawn between the original “balanced” magnetron and several unbalanced configurations in terms of field strengths, deposition rates, etching characteristics, and substrate ion current densities for the growth of TiN films. Plasma probe and magnetic field studies were performed under a variety of. conditions in order to understand better the effects of “unbalancing” on the nature of the plasma within the three-dimensional geometry of the deposition chamber. All the unbalanced configurations examined provided enhanced ion bombardment at the surface of the growing film; however, the closed-field or opposed magnet geometry resulted in a threefold or greater increase in current density when compared with that obtained utilizing the corresponding mirrored geometry under the same conditions.

High rate reactive sputtering of MoNx coatings

Abstract High rate reactive sputtering of MoN x films was performed using feedback control of the nitrogen partial pressure. Coatings were made at four different target powers: 2.5, 5.0, 7.5 and 10 kW. No hysteresis was observed in the nitrogen partial pressure vs. flow plot, as is typically seen for the Ti-N system. Four phases were determined by X-ray diffraction: molybdenum, Mo-N solid solution, β-Mo 2 N and ψ-Mo 2 N. The hardness of the coatings depended upon composition, substrate bias, and target power. The phases present in the hardest films differed depending upon deposition parameters. For example, the β-Mo 2 N phase was hardest (load 25 gf) at 5.0 kW with a value of 3200 kgf mm −2 , whereas the hardest coatings at 10 kW were the γ-Mo 2 N phase (3000 kgf mm −2 ). The deposition rate generally decreased with increasing nitrogen partial pressure, but there was a range of partial pressures where the rate was relatively constant. At a target power of 5.0 kW, for example, the deposition rates were 3300 A min −1 for a N 2 partial pressure of 0.05–1.0 mTorr.

Reactive sputtering in the ABSTM system

Abstract The cathodic arc and unbalanced magnetron (UBM) sputtering processes have been combined into one unit called the ABS TM coating system, the very first of which was installed at BIRL. The arc process is usually used for the substrate sputter etching step, and shutters remain in front of the cathodes to intercept macroparticles. A conventional argon sputter etch is also available, but it takes approximately 100 min to heat a full load of substrates to 300 °C, against 15 min for the arc etch. The longer argon etch roughens the substrate surface, whereas the shorter arc etch can actually lead to a reduction in surface roughness. The UBM process with its high degree of ion bombardment is used for reactive deposition of the coatings. TiN and Ti 0.5 Al 0.5 N coatings produced in the ABS TM system have excellent hardness and scratch test critical load values. The hardnesses for TiN and Ti 0.5 Al 0.5 N coatings are 2200 and 2250 kgf mm -2 respectively, and the critical load for TiN on M2 tool steel is 50–60 N for a 2.0 μm thick film. For a 3.0 μm thick Ti 0.5 Al 0.5 N coating on tool steel, the critical load is 60 N, and on cemented carbide it is greater than 100 N. Field trials with Ti 0.5 Al 0.5 N coated saw blades and end mills show a 2–3 times improvement in performance compared with TiN coated tools. Cross-sectional transmission electron microscopy shows that the structure of a Ti 0.5 Al 0.5 N coating on steel is columnar and that it is fully dense.

Residual stress and strain distribution anomalies in TiN films deposited by physical vapor deposition

Abstract The residual stress, lattice parameters and strain-broadening of diffraction peaks have been studied in a series of TiN films deposited on a C3 cemented carbide substrate by reactive magnetron sputtering. The substrate bias voltage and the sputter target input power were varied. It was found that complex residual stress situations could exist where, for example, (220) planes could exhibit high compressive and shear stress, (422) planes, simple low tensile stress and (333)–(511) planes, simple low compressive stress conditions within a given film. The observed residual stress behavioral patterns fell into three groups depending upon the deposition conditions. In addition, the lattice parameters and peak broadening showed positive or negative deviations from the average values of the remaining planes, specific within each behavioral pattern range. It is thought that these effects are associated with the dramatic increases in the defect and stacking fault population found with increasing bias, and with the ultramicrocracking on (220) planes, which have been reported in the literature from transmission electron microscopy studies of TiN films made by physical vapor deposition methods.