Room and High Temperature Tribological Behaviour of W-DLC Coatings Produced by DCMS and Hybrid DCMS-HiPIMS Configuration

coatings Article Room and High Temperature Tribological Behaviour of W-DLC Coatings Produced by DCMS and Hybrid DCMS-HiPIMS Configuration Manuel Evaristo 1,*, Filipe Fernandes 1,2,* and Albano Cavaleiro 1,3 1SEG-CEMMPRE: Department of Mechanical Engineering, University of Coimbra, Rua Luís Reis Santos, 3030-788 Coimbra, Portugal; [email protected] 2Department of Control Engineering, Czech Technical University in Prague, Technicka 2, 6 166 27 Prague, Czech Republic 3LED&Mat: Instituto Pedro Nunes, Rua Pedro Nunes, 3030-199 Coimbra, Portugal *Correspondence: [email protected] (M.E.); [email protected] (F.F.) Received: 24 February 2020; Accepted: 25 March 2020; Published: 27 March 2020   Abstract: Carbon-based coatings are used in many applications, particularly in sliding contacts to reduce friction and wear. To improve the tribological properties, these coatings are usually alloyed with metals; W is one of the most used since it helps improve the tribological performance at high temperatures. In this work, we compared the tribological performance of Diamond-Like Carbon alloyed with tungsten (DLC-W) films deposited by direct current magnetron sputtering (DCMS) with films deposited in a hybrid configuration DCMS +high power impulse magnetron sputtering (HiPIMS). The DLC-W coatings were produced with approximately the same W content. One hydrogenated film was deposited with the hybrid configuration for comparison purposes. Microstructure, structure, mechanical properties, and tribological behaviour were used to compare the coatings. All the films displayed a low-order structure of tungsten carbide embedded in an amorphous carbon matrix. The use of the hybrid HiPIMS/DCMS results in coatings with more compact morphologies due to the high ionization fraction of the species produced on the W target (W and Ar ionized species), which primarily will oppose the shadowing effect as the ions will reach the substrate at angles close to 90 ◦ . HiPIMS non-hydrogenated film is the more tribological, performing either at room or high temperature (150 ◦ C) due to the much more compact morphology, which avoids the detachment of hard W-C particles, which are responsible for more efficiently scratching the film surface. Experiments revealed that wear behaviour in all the films is governed by the contact of the tribolayer formed on the counterpart composed of W–C, C and W–O against the surface of the film. Keywords: magnetron sputtering; high power impulse magnetron sputtering—HiPIMS; tungsten doped DLC coatings; tribology 1. Introduction Diamond-like carbon (DLC) coatings are well established for many tribological applications such as internal combustion engines and biomedical devices due to their low wear rate (<10 −16 m 3 /Nm), low coefficient of friction (COF) (<0.2), corrosion resistance, and biocompatibility [ 1 ]. Specifically, hydrogen-free DLC coatings, with high levels of sp 3 content, displayed very high hardness associated with a low wear rate and friction coefficient when sliding against steel. These coatings perform globally better when compared with the nitrides and carbides-based wear protective coatings used in those industries. However, some important shortcomings must be solved to make this type of coating more attractive for a wider range of applications. First of all, the adhesion of the coatings to the substrate must be improved as well as their mechanical properties [ 2 , 3 ]. Another detrimental Coatings 2020,10, 319; doi:10.3390/coatings10040319 www.mdpi.com/journal/coatings Coatings 2020,10, 319 2 of 15 feature of DLC coatings is the influence of environmental conditions on their tribological performance such as temperature, and atmosphere to which they are exposed, resulting in different levels of wear rates and friction for the same coating [ 4 , 5 ]. In particular, the performance of DLC coatings at high temperature is limited due to their fast oxidation at temperatures higher than 200 ◦ C [ 6 ]. Thus, studies have been conducted to explore new ways to improve the adhesion and to extend their service life at high temperature conditions. To solve those problems, studies on different types of architectures of the interlayers (adhesion and/or gradient layers) and doping of DLC coatings with metallic elements, have been being carried out in the last few decades. The alloying elements used to tune the properties of DLC films are divided into carbides and non-carbide formers, as follows: Ti, Cr, F, W [ 7 – 9 ] and Cu and Ag [ 10 , 11 ] respectively. Among those elements, W is the most studied one due to its capacity to improve the tribological behaviour of DLC’s at increased temperatures. Voevodin et al. [ 12 ] observed a low COF of 0.30 for a coating with nanocrystalline WC particles embedded in a a-DLC matrix when tested against steel in an ambient air atmosphere (50% RH) at 25 ◦ C. Abou Gharam et al. [ 13 ], who investigated the tribological behaviour of WC/a-C coatings deposited by magnetron sputtering against aluminium antagonist sliding partner, reported low COF of 0.18 and 0.12 for 400 and 500 ◦ C, respectively. The formation of WO 3 was responsible for the low COF. Jibin et al. [ 14 ] also reported a significant improvement in the tribological performances of the WC/a-C coating deposited by magnetron sputtering tested against AISI52100 steel balls at high temperature (200 ◦ C), due to the formation of a continuously compacted tribo-film of WO 3 . Banerji et al. [ 15 ] reported COF values of 0.06 for 100 ◦ C, 0.46–0.54 between 200 ◦ C and 300 ◦ C, 0.07 for 400 ◦ C, and 0.08 for 500 ◦ C, when tested against Ti alloy balls. Bhowmick et al. [ 16 ] reported COF of 0.11 when tested against Al-6.5% Si alloy under the influence of pure oxygen atmosphere. In summary, the incorporation of W in the range of 10–15 at. % was shown to improve the tribological properties of DLC coatings at elevated temperatures, when tested against different counterparts. Metal-doped carbon base coatings are being mainly produced/developed by magnetron sputtering Physical Vapor Deposition (PVD) techniques. Different conventional kinds of plasma generators are being used in the production of a-C, DLC coatings, such as: direct current (DC), radio frequency (RF) and pulsed direct current. Recently, a new power supply designated as high power impulse magnetron sputtering (HiPIMS) emerged in the market and started to be used on the production of different types of coatings, including DLC’s. The advantage of HiPIMS in relation to the conventional power supplies relies on the fact that it allows high ionization of the species being deposited and, consequently, can tune the morphology, structure and mechanical properties of films. Recently, Aijaz et al. [ 17 , 18 ] reported that HiPIMS is a suitable technique to produce hydrogen-free DLC films with properties comparable to other DLC films deposited by different state-of-the-art deposition processes. This work first reports the comparison of the tribological performance at room and high temperature (150 ◦ C) of W-DLC-doped coatings produced with approximately the same W content by conventional DCMS and hybrid DCMS-HiPIMS configurations. For hybrid configuration, one hydrogen free and one hydrogenated W-DLC film were produced to compare with the DCMS film. 2. Experimental Details The coatings were deposited by PVD magnetron sputtering, on mirror-polished Si wafer, and polished AISI M2 high-speed steel cylinders with a diameter of 50.0 mm and 5 mm thickness with mirror finishing of the surface to be coated. Prior to deposition, all substrates were ultrasonically cleaned in acetone and ethanol baths for 15 min. and then air dried by air blow. A semi-industrial closed field unbalanced magnetron sputtering machine (TEER Coating Ltd., Worcestershire, UK) equipped with four cathodes, evenly distributed in relation to the centre of the chamber, was used for the depositions. The targets (380 mm × 175 mm × 10 mm) were vertically aligned in the chamber. The samples to be coated were mounted in a substrate holder that was placed at a distance of 150 mm from the targets. Three different coatings were deposited in this work, with W content in the range of 10 to 15 at. %, using two different power supplies arrangements, i.e., one coating deposited Coatings 2020,10, 319 3 of 15 by DCMS and two coatings produced using a hybrid DCMS-HiPIMS configuration. The W-DLC coating produced by DCMS power supply was sputtered from two high purity C targets and a C target containing 14 W pellets with 20 mm of diameter, evenly distributed along the race track. In order to improve the coatings adhesion, a Cr target was used to produce interlayer and gradient layers as described later. In hybrid DCMS-HiPIMS depositions, the composite C–W target was replaced by a high purity W target and connected to HiPIMS power supply (HiPIMS Cyprium ™ III plasma generator, Zpulser Inc.), working in deep oscillation magnetron mode. An example of the HiPIMS DOMS discharge voltage and current oscillating waveforms used in this work is shown in Figure 1. Coatings 2020, 10, x FOR PEER REVIEW 3 of 15 improve the coatings adhesion, a Cr target was used to produce interlayer and gradient layers as described later. In hybrid DCMS-HiPIMS depositions, the composite C–W target was replaced by a high purity W target and connected to HiPIMS power supply (HiPIMS Cyprium™ III plasma generator, Zpulser Inc.), working in deep oscillation magnetron mode. An example of the HiPIMS DOMS discharge voltage and current oscillating waveforms used in this work is shown in Figure 1. Figure 1. Representative HiPIMS DOMS discharge current and voltage oscillating waveforms for the W target used for the depositions. For this configuration, one hydrogen free and one hydrogenated W-DLC film were produced. A schematic representation of the chamber configuration and targets distribution for each of the depositions is shown in Figure 2. (a) (b) (c) Figure 2. Schematic illustration of the deposition chamber and distribution of targets used for the films deposited for this work, (a) DCMS, (b) HIPIMS and (c) HIPIMS CH4. Prior to each deposition, the targets and the substrates were sputter cleaned in an Ar discharge at 0.37 Pa. This stage was made in a two-step procedure. The first step (common for all depositions configurations) was the cleaning of the C targets and substrates by applying a power of 1000 W on each C target and a bias voltage of 600 V on the substrates for 20 min. The next step was the cleaning of the Cr and C–W composite targets for DCMS by applying a power of 2,000 W to the Cr target and a power of 1,000 W to the C–W composite target. In the case of hybrid configuration, the following pulse was applied on the W target connected to the HiPIMS power supply: average peak power (Pa) of 500 W, charging voltage (DCint) of 250 V, constant voltage time on (ton) of 6 μs, period (T) of 50 μs, and pulse duration (D) of 500 μs. C targets were powered with 2,000 W. To improve the adhesion of the films the following sequence of interlayers were deposited: i) Cr layer by applying 2,000 W to the Cr target for 10 min and a pulsed negative bias of -110 V (250 kHz) to the substrate holder (deposition pressure of 0.37 Pa); ii) gradient layer, decreasing the power applied to Cr target to 0 W and increasing the power applied to the C targets and composite C–W target during 10 min up to 1,750 and 1,000 W, 0 20 40 60 80 100 120 140 Ip (A) 0125 250 375 500 -1200 -1000 -800 -600 -400 -200 0 Time (s) Vp (V) Figure 1. Representative HiPIMS DOMS discharge current and voltage oscillating waveforms for the W target used for the depositions. For this configuration, one hydrogen free and one hydrogenated W-DLC film were produced. A schematic representation of the chamber configuration and targets distribution for each of the depositions is shown in Figure 2. Coatings 2020, 10, x FOR PEER REVIEW 3 of 15 improve the coatings adhesion, a Cr target was used to produce interlayer and gradient layers as described later. In hybrid DCMS-HiPIMS depositions, the composite C–W target was replaced by a high purity W target and connected to HiPIMS power supply (HiPIMS Cyprium™ III plasma generator, Zpulser Inc.), working in deep oscillation magnetron mode. An example of the HiPIMS DOMS discharge voltage and current oscillating waveforms used in this work is shown in Figure 1. Figure 1. Representative HiPIMS DOMS discharge current and voltage oscillating waveforms for the W target used for the depositions. For this configuration, one hydrogen free and one hydrogenated W-DLC film were produced. A schematic representation of the chamber configuration and targets distribution for each of the depositions is shown in Figure 2. (a) (b) (c) Figure 2. Schematic illustration of the deposition chamber and distribution of targets used for the films deposited for this work, (a) DCMS, (b) HIPIMS and (c) HIPIMS CH4. Prior to each deposition, the targets and the substrates were sputter cleaned in an Ar discharge at 0.37 Pa. This stage was made in a two-step procedure. The first step (common for all depositions configurations) was the cleaning of the C targets and substrates by applying a power of 1000 W on each C target and a bias voltage of 600 V on the substrates for 20 min. The next step was the cleaning of the Cr and C–W composite targets for DCMS by applying a power of 2,000 W to the Cr target and a power of 1,000 W to the C–W composite target. In the case of hybrid configuration, the following pulse was applied on the W target connected to the HiPIMS power supply: average peak power (Pa) of 500 W, charging voltage (DCint) of 250 V, constant voltage time on (ton) of 6 μs, period (T) of 50 μs, and pulse duration (D) of 500 μs. C targets were powered with 2,000 W. To improve the adhesion of the films the following sequence of interlayers were deposited: i) Cr layer by applying 2,000 W to the Cr target for 10 min and a pulsed negative bias of -110 V (250 kHz) to the substrate holder (deposition pressure of 0.37 Pa); ii) gradient layer, decreasing the power applied to Cr target to 0 W and increasing the power applied to the C targets and composite C–W target during 10 min up to 1,750 and 1,000 W, 0 20 40 60 80 100 120 140 Ip (A) 0125 250 375 500 -1200 -1000 -800 -600 -400 -200 0 Time (s) Vp (V) Figure 2. Schematic illustration of the deposition chamber and distribution of targets used for the films deposited for this work, (a) DCMS, (b) HIPIMS and (c) HIPIMS CH4. Prior to each deposition, the targets and the substrates were sputter cleaned in an Ar discharge at 0.37 Pa. This stage was made in a two-step procedure. The first step (common for all depositions configurations) was the cleaning of the C targets and substrates by applying a power of 1000 W on each C target and a bias voltage of 600 V on the substrates for 20 min. The next step was the cleaning of the Cr and C–W composite targets for DCMS by applying a power of 2000 W to the Cr target and a power of 1000 W to the C–W composite target. In the case of hybrid configuration, the following pulse was applied on the W target connected to the HiPIMS power supply: average peak power (Pa) of 500 W, charging voltage (DCint) of 250 V, constant voltage time on (t on ) of 6 µ s, period (T) of 50 µ s, and pulse Coatings 2020,10, 319 4 of 15 duration (D) of 500 µ s. C targets were powered with 2000 W. To improve the adhesion of the films the following sequence of interlayers were deposited: (i) Cr layer by applying 2000 W to the Cr target for 10 min and a pulsed negative bias of -110 V (250 kHz) to the substrate holder (deposition pressure of 0.37 Pa); (ii) gradient layer, decreasing the power applied to Cr target to 0 W and increasing the power applied to the C targets and composite C–W target during 10 min up to 1750 and 1000 W, respectively, and applying same bias voltage as applied to the Cr interlayer; in the case of the coatings deposited with HiPIMS connected to the W target, the bias was set also for − 110 V and the power on the Cr target was decreased down to 0 in a period of 10 min. HiPIMS power supply was immediately turned on, as soon as the gradient layer started, with the following pulse characteristic: average power (Pa) of 500 W, DCint of 400 V, constant time (t on ) of 6 µ s, oscillation period (T) of 50 µ s, and a pulse duration (D) of 500 µ s. This pulse characteristics were optimized in such a way that the final W content on the coating could be similar to the coating produced by DCMS. In addition, for the hydrogenated W-DLC coating, 10 sccm of CH 4 gas was introduced after 5 min of the gradient layer. After the gradient layer, the depositions proceeded with the final power applied to the targets as set for the end of the gradient layer, during 100 min. A summary of the deposition parameters used on the coatings production is shown in Table 1. Table 1. Deposition conditions of films. Sample Ar Flow (sccm) CH4 Flow (sccm) Pressure (Pa) Power C. 1 (W) Power C. 2 (W) Power C–W Target (W) DCMS 46 0 0.45 1750 1750 1000 Average Power (W) DC int (V) Vp (V) Ip (A) HiPIMS 46 0 0.45 1750 1750 500 400 1112 130 HiPIMS CH4 46 10 0.50 1750 1750 500 400 1156 109 Hereinafter, for easy coatings identification, the coating deposited by DCMS will be labelled as “DCMS” and coatings deposited by combined DCMS with High-Power Impulse Magnetron Sputtering (HiPIMS) will be labelled as “HIPIMS” and “HIPIMS CH4”, the last deposited under a reactive CH 4 and Ar atmosphere. It should be here highlighted that the W target was decided to be connected to the HiPIMS power supply because of their high ionization sputtered potential as compared to C. A scanning electron microscope (SEM) (ZEISS Merlin—Field Emission Gun) (Zeiss, Oberkochen, Germany) equipped with wavelength dispersive spectroscopy (WDS) was used to observe the surface and cross section morphologies and to evaluate the chemical composition. SiC and WC reference materials were used as standards for calibration of the W and C signals for chemical composition evaluation. The structure of the coatings was evaluated by X–ray diffraction in grazing incidence of 3 ◦ in the range 10–90 ◦ with a PANalytical X’Pert Pro MPD system (Cu K α radiation k α =1.54 Å, 45 kV and 40 mA) and by green light Raman spectroscopy (Horiba Jobin Yvon microscope, 532 nm/2.33 eV laser excitation). The laser power had to be lowered to 1% of the maximum of 40 mW to avoid the bonding structure thermal crystallization by the laser excitation. The coatings adhesion was evaluated by scratch test with a Rockwell indenter with a diamond tip of 200 µ m radius with standard conditions from 0 to 50 N. The scratches were analysed by optical microscopy and the critical adhesion loads were taken from the analysis of those images. The hardness and reduced Young’s modulus of the films were measured by depth-sensing indentation (MicroMaterials NanoTest platform). A maximum load of 10 mN was selected in order Coatings 2020,10, 319 5 of 15 to avoid the influence of the substrate with the maximum indentation depth never reaching values higher than 1/10 of the coating thickness. Tribological characterization of the coatings was conducted in a standard lab-scale worldwide recognized technique, pin-on-disc sliding equipment, to evaluate the interaction between the films against AISI 52100 material under different environmental conditions. Experiments conducted on pin-on-disc were tested at room temperature (RT) and 150 ◦ C against 10-mm steel balls. The tests were performed under 5 N load, sliding speed of 0.1 m/s and a total sliding distance of 5000 and 3250 m for RT and 150 ◦C respectively. 3. Results 3.1. Coatings Charcterization WDS analysis does not allow measurement of the H concentration on hydrogenated produced film, but is a reliable technique to evaluate the W and C concentration; the W/C ratios in atomic concentration for the three coatings deposited are 0.18 for the DCMS coating and 0.15 for HIPIMS and HIPIMS CH4, which is very similar (although slight higher for DCMS film), confirming the production of films with approximately the same W concentration. The thicknesses of HiPIMS’s coatings are lower (1050 and 1150 nm for HIPIMS and HIPIMS CH4) than the coating deposited by DCMS (1730 nm). Since the power applied to C targets placed on cathodes 2 and 4 was the same in both DCMS and HiPIMS’s depositions and the slight higher sputtering rate of W as compared to carbon (0.6 against 0.2 (in Ar atmosphere under 600 eV), the lower thickness of the HiPIMS coatings is primarily attributed to the much lower power applied to the W target as compared to the composite C-W target that also contributes to the C in the coating. In addition, the expected metallic ion back-attraction to the target and the high voltages in HiPIMS power supply as compared to DCMS (1112–1156 against 480 V) can also contribute to the loss of deposition rate [ 19 ]. A slight increase of the deposition rate could be noticed for the HIPIMS CH4 film as compared to HIPIMS film due to the fact that CH 4 atmosphere works as an additional source for C and H to the coating. A similar feature was also observed for DLC coatings deposited by HiPIMS with increasing of the deposition rate with the C2H2flow [20]. The cross section and surface morphologies of the as-deposited films are shown in Figure 3. The DCMS coating displays a columnar morphology. The column tops are separated by large void channels and consist of aggregates of smaller sized features, 20–30 nm wide, that give rise to an irregular surface. This kind of rough surface morphology is usually referred to as a cauliflower-type morphology and develops under limited diffusion of the deposited species. Both coatings deposited by HIPIMS show a much more compact morphology, with HIPIMS CH4 film showing a quasi-featureless morphology. Taking into account that the substrate bias was similar in all depositions ( − 110 V), the more compact morphology of both HiPIMS films is attributed to high ionization fraction of the species produced on the W target (W and Ar ionized species), which primarily will oppose the shadowing effect as the ions will reach the substrate at angles close to 90 ◦ and also will bombard the film with high energy. Thus, the deposited species have high energy and the presence of a potential on the substrate will accelerate the ions promoting a more compact morphology [ 21 ]. Consequently those films displayed a homogeneous and isotropic surface morphology consisting of fine features, 15–25 nm wide. The slight improvement on compactness of HiPIMS CH4 film corroborates the previous works dealing with the introduction of H on DLC films [20]. Coatings 2020,10, 319 6 of 15 Coatings 2020, 10, x FOR PEER REVIEW 6 of 15 (a) (b) (c) (a1) (b1) (c1) Figure 3. SEM images of the coatings cross section obtained by SEM (a) DCMS, (b) HIPIMS, (c) HIPIMS CH4, and surface morphology, (a1), DCMS (b1), HIPIMS (c1) HIPIMS CH4. The structure of the coatings was analysed by Raman spectroscopy and X-ray diffraction. When visible Raman excitation (532 nm) is used to analyse the structure of the coatings, sp2/sp3 ratio cannot be determined as the carbon D and G peaks are results of the vibration of sp2 sites; however, relevant information can be taken from the ID/IG ratio, G peak position and FWHM (full width at half maxim) of the G peak, as it is summarized in Table 2. The Raman spectra for DCMS and HIPIMS films are very similar as shown in Figure 4a. Although, a decrease of the G peak position and FWHM can be noticed, nothing can be inferred about the sp2/sp3 ratio variation because the ratio between ID/IG remains constant. However, for the HiPIMS film deposited under a reactive atmosphere of CH4, the G peak position downshift in conjunction with the ID/IG ratio and G FWHM decreases. This is characteristic of a reduction of the average size of graphite-type sp2-bonded clusters and the promotion of sp3 bond formation [22]. According to the literature, the formation of sp3 bonds in DLC films is mainly achieved by sub-plantation of energetic carbon ions into the subsurface region of the growing film, which promotes the formation of dense films [23,24]. However, the introduction of H has also been reported to promote sp3 bonds formation due to the random covalent network bonding between fourfold sp3 carbon, threefold sp2 carbon and singly coordinated hydrogen atoms. Although slight changes on the voltage and current characteristics of the targets were observed in HiPIMS’s deposition, no significant change on the bombarding conditions are expected as peak power on the targets does not significantly change (product of voltage versus current) when CH4 is introduced to produce the hydrogenated film. Thus, a higher level of sp3 bonds on hydrogenated film are likely due to the presence of H on the film. Figure 4b shows the XRD diffractograms of the different films in the region 10–90° (2). All three coatings displayed only a very broad peak positioned at 37°. This position corresponds well to the main standard peaks of several W–C phases such as WC1-x and WC (see ICDD cards 00-020-1316 and 00-072-0097 respectively). The conjunction of Raman and XRD results allows to suggest that the structure of films consists of a nanocomposite structure of tungsten carbide embedded in an amorphous carbon matrix. Similar results were found in the literature for WC coatings prepared by magnetron sputtering [12,25]. A slight broadening of the XRD diffraction peak of HiPIMS’s films as compared to DCMS film, can be observed, supporting the higher ionization fraction and energy of the deposited species induced by the former power supply. The very sharp peaks positioned at ~ 44 and 56° corresponds to signals coming from the Cr interlayer (deposited to improve adhesion) and Si substrate, respectively. Cr signal is not detected in DCMS film due to its much higher thickness. Figure 3. SEM images of the coatings cross section obtained by SEM ( a ) DCMS, ( b ) HIPIMS, ( c ) HIPIMS CH4, and surface morphology, (a1), DCMS (b1), HIPIMS (c1) HIPIMS CH4. The structure of the coatings was analysed by Raman spectroscopy and X-ray diffraction. When visible Raman excitation (532 nm) is used to analyse the structure of the coatings, sp 2 /sp 3 ratio cannot be determined as the carbon D and G peaks are results of the vibration of sp 2 sites; however, relevant information can be taken from the I D /I G ratio, G peak position and FWHM (full width at half maxim) of the G peak, as it is summarized in Table 2. The Raman spectra for DCMS and HIPIMS films are very similar as shown in Figure 4a. Although, a decrease of the G peak position and FWHM can be noticed, nothing can be inferred about the sp 2 /sp 3 ratio variation because the ratio between I D /I G remains constant. However, for the HiPIMS film deposited under a reactive atmosphere of CH 4 , the G peak position downshift in conjunction with the I D /I G ratio and G FWHM decreases. This is characteristic of a reduction of the average size of graphite-type sp 2 -bonded clusters and the promotion of sp 3 bond formation [ 22 ]. According to the literature, the formation of sp 3 bonds in DLC films is mainly achieved by sub-plantation of energetic carbon ions into the subsurface region of the growing film, which promotes the formation of dense films [ 23 , 24 ]. However, the introduction of H has also been reported to promote sp 3 bonds formation due to the random covalent network bonding between fourfold sp 3 carbon, threefold sp 2 carbon and singly coordinated hydrogen atoms. Although slight changes on the voltage and current characteristics of the targets were observed in HiPIMS’s deposition, no significant change on the bombarding conditions are expected as peak power on the targets does not significantly change (product of voltage versus current) when CH 4 is introduced to produce the hydrogenated film. Thus, a higher level of sp 3 bonds on hydrogenated film are likely due to the presence of H on the film. Table 2. Raman analysis of the coatings. Sample G Pos. cm−1ID/IGG FWHM cm−1 DCMS 1561.6 5.1 146.2 HIPIMS 1557.4 5.3 141.4 HIPIMS CH41557.4 3.4 130.6 Coatings 2020,10, 319 7 of 15 Coatings 2020, 10, x FOR PEER REVIEW 7 of 15 (a) (b) Figure 4. (a) Raman spectra and (b) X-ray diffraction patterns of the coatings. Table 2. Raman analysis of the coatings. Sample G Pos. cm-1 ID/IG G FWHM cm−1 DCMS 1561.6 5.1 146.2 HIPIMS 1557.4 5.3 141.4 HIPIMS CH4 1557.4 3.4 130.6 The adhesion, hardness and reduced Young’s modulus of the films were evaluated by scratch testing and nanoindentation. Figure 5a shows one representative scratch test for each of the coatings. Three types of failure modes were observed on the scratch tracks of the coatings: the first cracking (Lc1), first coating shipping (Lc2) and more than 50% coating failure (Lc3). Globally, HiPIMS’s coatings displayed higher critical loads than DCMS film with the HiPIMS CH4 being the better performing. Hardness of HiPIMS film is similar to the hardness of DCMS film, in spite of the much more compact morphology of the former. This suggests that the compact morphology of the HiPIMS is mainly influenced by the higher level of ionization fraction of species, which will fight the shadowing effect. This agrees with the similar Raman spectra between those coatings. HiPIMS CH4 displayed the higher hardness among all the coatings, corroborating the featureless morphology and increase of sp3 bonds shown above. Considering that the only difference in the deposition of this film in relation to HiPIMS is the use of the reactive CH4 atmosphere, we may assume that the increase in hardness is due to the increase of sp3 content of the coatings. This behaviour was also observed by other authors when introducing C2H2 to the discharge [20]. Concerning the evolution of the reduced Young’s modulus, there is a gradual decrease from the DCMS to the HIPIMS CH4 from approximately 200 to 193 GPa. Since the reduced Young’s modulus is related with the type of bonds in the material, the decrease from DCMS to HiPIMS (196 GPa) can be related with the slight decrease of the W content from 14.6 to 12.8 at. %, since there is no indication from the Raman spectra that there are any changes in the structure of the amorphous carbon matrix. A further reduction of the modulus was observed to the HIPIMS CH4, which can be related with the presence of C–H bonds, which avoids establishing sp3 bonds between C atoms, thus lowering the bonding energy. 400 800 1200 1600 2000 DCMS Intensity (a. u) Raman shift (cm-1) HIPIMS HIPIMS CH4 10 20 30 40 50 60 70 80 90 Intensity (a.u) 2 (º) DCMS HIPMS HIPMS CH4 Figure 4. (a) Raman spectra and (b) X-ray diffraction patterns of the coatings. Figure 4b shows the XRD diffractograms of the different films in the region 10–90 ◦ (2 θ ). All three coatings displayed only a very broad peak positioned at 37 ◦ . This position corresponds well to the main standard peaks of several W–C phases such as WC1-x and WC (see ICDD cards 00-020-1316 and 00-072-0097 respectively). The conjunction of Raman and XRD results allows to suggest that the structure of films consists of a nanocomposite structure of tungsten carbide embedded in an amorphous carbon matrix. Similar results were found in the literature for WC coatings prepared by magnetron sputtering [ 12 , 25 ]. A slight broadening of the XRD diffraction peak of HiPIMS’s films as compared to DCMS film, can be observed, supporting the higher ionization fraction and energy of the deposited species induced by the former power supply. The very sharp peaks positioned at ~44 and 56 ◦ corresponds to signals coming from the Cr interlayer (deposited to improve adhesion) and Si substrate, respectively. Cr signal is not detected in DCMS film due to its much higher thickness. The adhesion, hardness and reduced Young’s modulus of the films were evaluated by scratch testing and nanoindentation. Figure 5a shows one representative scratch test for each of the coatings. Three types of failure modes were observed on the scratch tracks of the coatings: the first cracking (Lc1), first coating shipping (Lc2) and more than 50% coating failure (Lc3). Globally, HiPIMS’s coatings displayed higher critical loads than DCMS film with the HiPIMS CH4 being the better performing. Hardness of HiPIMS film is similar to the hardness of DCMS film, in spite of the much more compact morphology of the former. This suggests that the compact morphology of the HiPIMS is mainly influenced by the higher level of ionization fraction of species, which will fight the shadowing effect. This agrees with the similar Raman spectra between those coatings. HiPIMS CH4 displayed the higher hardness among all the coatings, corroborating the featureless morphology and increase of sp 3 bonds shown above. Considering that the only difference in the deposition of this film in relation to HiPIMS is the use of the reactive CH 4 atmosphere, we may assume that the increase in hardness is due to the increase of sp 3 content of the coatings. This behaviour was also observed by other authors when introducing C2H2to the discharge [20]. Coatings 2020,10, 319 8 of 15 Coatings 2020, 10, x FOR PEER REVIEW 8 of 15 (a1) (b) (a2) (a3) Figure 5. Adhesion scratches and critical loads of: (a1) DCMS, (a2) HIPMPS, (a3) HIPIMS CH4; Hardness and reduced Young´s modulus of the films (b). 3.2. Tribological Behaviour 3.2.1. Wear Rate and Friction Coefficient at Room Temperature and 150 °C The specific wear rate and the average friction coefficient of the coatings tested at RT and 150 °C are shown in Figure 6. At RT, HiPIMS coating displayed the lowest specific wear rate among all the three films, with a COF value of 0.29; the DCMS film has a similar friction coefficient to the HiPIMS one but displays a higher specific wear rate. HiPIMS CH4 coating is the lowest performer despite its lower friction coefficient (0.11) and higher hardness as compared to the other coatings. According to some authors [26], the addition of hydrogen to the coatings results in a decrease of the friction, due to the easier graphitization of the C and an increase of the wear rate associated to the easier removal of the film due to the less bonding energy caused by the presence of H. The specific wear rate and the friction coefficient values are in the range of those reported in the literature for DLC-W coatings with similar chemical compositions [12,14]. The friction coefficient curve of HiPIMS film shows an initial stage of friction lasting 500 m, corresponding to the running in period (see Figure S1 in supplementary material). After that, steady state friction is reached; however, the curves present regular oscillations up to the end of the test. These fluctuations are often assigned with the dynamic process of the tribolayer formation [27]. The friction coefficient curve of DCMS film shows an initial stage with high friction, lasting 1,300 m. After this stage, stable values are observed, although some periods with lower friction can be observed. The HiPIMS CH4 coating presents the lowest friction with gradual decrease until the first 1,000 m and then increases up to 1,500 m, reaching a stable COF value of 0.1 approximately. The experiments at 150 °C evidenced in all films an increase of specific wear rate in relation to the tests performed at room temperature. Similar to room temperature tests, the film with the best wear resistance was the HIPIMS film, whilst the lowest performer remains the HIPIMS CH4 film, as shown in Figure 6. DCMS film has a specific wear rate almost three times higher than HIPIMS film, and its friction coefficient is the highest amongst all of the films, two times higher than the value measured at RT. For HiPIMS coating, the friction coefficient decreased from 0.3 to 0.1 and for HiPIMS, DCMS HIPIMS HIPIMSCH4 10 12 14 16 18 20 22 180 185 190 195 200 205 210 215 220 225 230 235 240 Hardness Hardness (GPa) Coating Reduced Modulus Reduced Modulus (GPa) Figure 5. Adhesion scratches and critical loads of: ( a1 ) DCMS, ( a2 ) HIPMPS, ( a3 ) HIPIMS CH 4 ; Hardness and reduced Young´s modulus of the films (b). Concerning the evolution of the reduced Young’s modulus, there is a gradual decrease from the DCMS to the HIPIMS CH4 from approximately 200 to 193 GPa. Since the reduced Young’s modulus is related with the type of bonds in the material, the decrease from DCMS to HiPIMS (196 GPa) can be related with the slight decrease of the W content from 14.6 to 12.8 at. %, since there is no indication from the Raman spectra that there are any changes in the structure of the amorphous carbon matrix. A further reduction of the modulus was observed to the HIPIMS CH4, which can be related with the presence of C–H bonds, which avoids establishing sp 3 bonds between C atoms, thus lowering the bonding energy. 3.2. Tribological Behaviour 3.2.1. Wear Rate and Friction Coefficient at Room Temperature and 150 ◦C The specific wear rate and the average friction coefficient of the coatings tested at RT and 150 ◦ C are shown in Figure 6. At RT, HiPIMS coating displayed the lowest specific wear rate among all the three films, with a COF value of 0.29; the DCMS film has a similar friction coefficient to the HiPIMS one but displays a higher specific wear rate. HiPIMS CH4 coating is the lowest performer despite its lower friction coefficient (0.11) and higher hardness as compared to the other coatings. According to some authors [ 26 ], the addition of hydrogen to the coatings results in a decrease of the friction, due to the easier graphitization of the C and an increase of the wear rate associated to the easier removal of the film due to the less bonding energy caused by the presence of H. The specific wear rate and the friction coefficient values are in the range of those reported in the literature for DLC-W coatings with similar chemical compositions [ 12 , 14 ]. The friction coefficient curve of HiPIMS film shows an initial stage of friction lasting 500 m, corresponding to the running in period (see Figure S1 in Supplementary Material). After that, steady state friction is reached; however, the curves present regular oscillations up to the end of the test. These fluctuations are often assigned with the dynamic process of the tribolayer formation [ 27 ]. The friction coefficient curve of DCMS film shows an initial stage with high friction, Coatings 2020,10, 319 9 of 15 lasting 1300 m. After this stage, stable values are observed, although some periods with lower friction can be observed. The HiPIMS CH4 coating presents the lowest friction with gradual decrease until the first 1000 m and then increases up to 1500 m, reaching a stable COF value of 0.1 approximately. Coatings 2020, 10, x FOR PEER REVIEW 9 of 15 CH4 film remains approximately the same at 0.1. The friction coefficient curves did not display any significant running in period (see supplementary material). Steady state friction of HiPIMS films is smooth, whilst some oscillations can be observed for HiPIMS CH4 film. The friction curve of DCMS film displayed a similar evolution for RT, although with higher friction; however, oscillation periods with low friction are absent. (a) (b) Figure 6. Specific wear rate at high (150°C) and room temperature (a); average friction coefficient for both temperatures (b). DCMS HIPIMS HIPIMS CH4 0.0 2.0x10-7 4.0x10-7 6.0x10-7 8.0x10-7 2.0x10-6 3.0x10-6 4.0x10-6 0.0 2.0x10-8 4.0x10-8 6.0x10-8 8.0x10-8 1.0x10-7 1.2x10-7 1.4x10-7 1.6x10-7 1.8x10-7 2.0x10-7 High Temperature 150ºC Wear rate high temp. (150ºC) (mm3/Nm) Coating Wear rate room temp. (mm3/Nm) Room temperature DCMS HIPIMS HIPIMS CH4 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Friction coeficient (-) Coating High temperature 150ºC Room temperature Coatings 2020, 10, x FOR PEER REVIEW 9 of 15 CH4 film remains approximately the same at 0.1. The friction coefficient curves did not display any significant running in period (see supplementary material). Steady state friction of HiPIMS films is smooth, whilst some oscillations can be observed for HiPIMS CH4 film. The friction curve of DCMS film displayed a similar evolution for RT, although with higher friction; however, oscillation periods with low friction are absent. (a) (b) Figure 6. Specific wear rate at high (150°C) and room temperature (a); average friction coefficient for both temperatures (b). DCMS HIPIMS HIPIMS CH4 0.0 2.0x10-7 4.0x10-7 6.0x10-7 8.0x10-7 2.0x10-6 3.0x10-6 4.0x10-6 0.0 2.0x10-8 4.0x10-8 6.0x10-8 8.0x10-8 1.0x10-7 1.2x10-7 1.4x10-7 1.6x10-7 1.8x10-7 2.0x10-7 High Temperature 150ºC Wear rate high temp. (150ºC) (mm3/Nm) Coating Wear rate room temp. (mm3/Nm) Room temperature DCMS HIPIMS HIPIMS CH4 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Friction coeficient (-) Coating High temperature 150ºC Room temperature Figure 6. Specific wear rate at high (150 ◦ C) and room temperature ( a ); average friction coefficient for both temperatures (b). The experiments at 150 ◦ C evidenced in all films an increase of specific wear rate in relation to the tests performed at room temperature. Similar to room temperature tests, the film with the best wear resistance was the HIPIMS film, whilst the lowest performer remains the HIPIMS CH4 film, as shown in Figure 6. DCMS film has a specific wear rate almost three times higher than HIPIMS film, and its friction coefficient is the highest amongst all of the films, two times higher than the value measured at RT. For HiPIMS coating, the friction coefficient decreased from 0.3 to 0.1 and for HiPIMS, CH4 film remains approximately the same at 0.1. The friction coefficient curves did not display any significant running in period (see Supplementary Material). Steady state friction of HiPIMS films is smooth, whilst some oscillations can be observed for HiPIMS CH4 film. The friction curve of DCMS film displayed a similar evolution for RT, although with higher friction; however, oscillation periods with low friction are absent.