PSI - Issue 13

Felipe C. da Silva et al. / Procedia Structural Integrity 13 (2018) 658–663 / Structural Integrity Procedia 00 (2018) 000 – 000

659

2

[3 – 5]. Nowadays TiN coatings are largely used in the automotive, aerospace, microelectronics and even in sanitary industries. It presents high corrosion resistance, excellent mechanical, electrical, optical (TiN presents a golden aspect which is aesthetically appreciated), and of course in tribology (due to high hardness) [6 – 10]. Physical vapor deposition techniques (PVD) are commonly used for TiN coating, and, in particular, the reactive magnetron sputtering technique has the advantage of using low deposition temperatures [10]. Sputtering techniques are characterized, however, by low deposition rates, and the triode magnetron sputtering technique (also known as grid-assisted magnetron sputtering, GAMS) allows an improvement by assisting plasma confinement [11]. An additional advantage of the GAMS technique is the possibility to customize some parameters during deposition, which may alter the film properties. Some of these variable, like the application of a polarization potential between the Ti target and the part (bias), may induce compressive residual stresses in the film, which are beneficial [12 – 14]. Another variable which can be controlled during deposition is the supply of nitrogen in the chamber atmosphere. Using this parameter it is possible to obtain a film with a gradient of properties along the thickness [15 – 20]. The aim of the present work is to study the effect of different bias potentials and of the nitrogen supply variation on the properties of TiN films, deposited over aluminum and over brass, obtained using the GAMS technique. Focus is on the mechanical behavior of the films (and, in particular, in its fracture behavior). 2. Experimental procedure 2.1. Thin film deposition The geometry of the samples are dog bone type. The used substrates were produced from commercial AA1100 and yellow brass (UNS C26800, 33 wt. % Zn) sheets, with 34 mm x 10 mm x 3 mm and 3.0 mm radius. The sheets were ground on one side with SiC paper down to #2000 granulometry, followed by polishing with diamond cloth down to 1  m granulometry, prior to film deposition . . Coated specimens were obtained under three levels of bias: - 40, -75 and -100 V, for two conditions: under constant and under variable nitrogen flow, using the GAMS technique [11] . The constant flow condition was set to ~8 sccm N2 throughout the experiment, and the variable condition started at zero flow, with a linear increase of ~ 0.4 sccm N2 for each minute until 8 sccm N2 was obtained. Nitrogen and Argon at a pressure of 0.4 Pa were used as precursor gases. A pure Ti sputtering target was used as Ti source. Distance between target and sample was fixed in 60 mm, and a 1mm X 1 mm grid was placed 20 mm distant from the Ti target, inside the glowing region of the operating plasma. Deposition was performed at 573 K in all experiments. In all experiments deposition started by a pure Ti layer, following a suggestion by ref. [14], to improve film substrate adherence. Total deposition time was fixed in 30 minutes . 2.2. Film characterization The obtained films were characterized both in structure/microstructure and in mechanical properties. The structural characterization relied on for the films deposited on aluminum with bias -40 V, also, Energy Filtered Transmission Electron Microscopy (EFTEM), in the case of the graded film. X-ray diffraction was additionally performed also in the Grazing Incidence mode (GIXRD) for the constant N2 flow films deposited on aluminum, to determine the residual stress levels induced in the plane of the film. These were determined using the so- called sin2 ψ method [21]. The GIXRD experiments were performed using angles of 3, 7 and 12 o . Both TiN thin films (conventional and graded) were primarily analyzed in a Hitachi H-9500 TEM operating at 200 keV with a LaB6 filament. A Gatan Image Filter (GIF) Quantum SE was also used for EFTEM analysis (only in the grade film). The mechanical characterization was performed by instrumented indentation using a Hysitron TI 950 Triboindenter with a Berkovich tip and 5 mN load, using the procedure suggested by Oliver and Pharr [22], and by scratching test using a CETR microtribometer (model UMT-2), with 1 N pre load and progressive load mode. For this experiment a 120 o spheroconical diamond indenter was used. In these experiments, load was progressively increased until a trail of about 3 mm was produced. The indenter displacement speed was set to 10 mm min-1 in both substrates. The test is evaluated in the form of the load corresponding to the observation of the first cracks (denoted Lc1, associated with a cohesive failure of the film) and the load corresponding to the observation of delamination (denoted Lc2, corresponding to an adhesive failure).