Characterization and Wear Response of Magnetron Sputtered W–B and W–Ti–B Coatings on WC–Co Tools

In this work, α-WB2 and (W,Ti)B2 borides were applied as wear-resistant coatings to commercial WC–Co cutting inserts. Properties of coatings deposited by magnetron sputtering on WC–Co tools were studied. The crystal structure and chemical composition were analyzed. Vickers hardness and surface roughness were determined and wear test in semi-dry conditions was performed. The W–B and W–Ti–B coatings deposited on WC–Co substrate were smooth and very hard. However, titanium alloy W-B films with Vickers hardness of 3630 ± 260 HV0.02 were characterized by lower adhesion to the substrate, influencing the wear mechanism. Turning tests carried out on 304 stainless steel showed that the W–B film caused less wear compared to uncoated insert. Moreover, when W–B coating was applied, flank wear was reduced by 30% compared to uncoated WC–Co insert. Additionally, coating prevented chipping of the edge during cutting under test conditions. The research shows that W–B film deposited by magnetron sputtering has great potential as a coating for cutting tools for difficult-to-cut materials.


Introduction
New tool materials or coatings are needed for special applications, particularly in machining of difficult-to-machine materials [1]. Coatings should be characterized by hardness, strength, and chemical inertness. Such properties are exhibited by boride films [2][3][4][5][6][7]. There are numerous compounds with boron that have good wear resistance, such as ZrB 2 , TiB 2 , BN, and Ti(B,N) [8]. In addition, tungsten borides and ternary transition (TM) metal-diborides possess very high application potential. The ternary TM-diboride coating films showing exceptionally high phase stability and mechanical properties, even at high temperature or even after exposition to high temperatures [2]. Addition of TM such as Ta or Ti cause that coatings possess fracture toughness k IC (~3.5 MPa √ m) higher than well-known coatings such us (Ti,Al)N, CrN/TiN, Ti-Si-N, TiN, and TiB 2 and at the same time they are super hard H > 40 GPa [2]. So far, there has been no research on the behavior of such layers under operating conditions. In Ref. [9] a series of borides (CrB 2 , Mo 2 B 5 and WB) were applied as protective coatings on cutting tools. Turning tests carried out on titanium in dry conditions showed that WB crystalline coating was resistant to abrasive wear. Similar properties were observed by Chrzanowska-Giżynska et al. [10,11], who investigated the friction coefficient and wear resistance of αWB coatings on flat stainless steel substrates in scratch tests. αWB films deposited by radio frequency magnetron sputtering (RF MS) were characterized by good adhesion. The 1 µm films tested in the scratch test failed at 200 mN due to cracking; however, further increase in the load did not cause and titanium (45 µm, 99.98% purity; Sigma Aldrich). For deposition of α-WB 2 film, powder was mixed and sintered with a 4.5:1 (B/W) ratio [23]. In the case of the target with titanium alloy, the ratio of boron to transition metals was 4.5 and Ti/(Ti + W) = 0.16 [4]. The first target consisted of two WB 2 and WB 4 phases. The second target was a nanocomposite of WB 4 , WB 2 , and TiB 2 phases. Detailed information about the manufacturing process and the chemical structure and mechanical properties of targets were described in Refs. [23] and [4], respectively. Based on our earlier studies on sputtering power and gas pressure [11], substrate temperature [10], and composition [4], the optimal parameters of deposition were chosen. The one-inch diameter target was mounted in the water-cooled magnetron sputtering cathode (Kurt J. Lesker, Jefferson Hills, PA, USA) and positioned at a distance of 50 mm in front of the target. The RF sputtering power was 50 W and bias potential on the substrate was floating. The deposition process occurred in a vacuum chamber initially pumped to 0.02 Pa and then filled with argon to a working pressure of 0.9 Pa. The gas flow of argon was 19 mL/min. Prior to each deposition, the target was sputtered for 5 min in order to ensure a clean surface and stable sputtering conditions. Film was deposited for 120 min on flat, polished substrate cut from WC rod (1 mm thick, Ø = 15 mm diameter; Stjorsen Polska, Tychy, Poland) and on commercial WC cutting tools. Before deposition, all substrates were cleaned with subsequent rinses in acetone, alcohol, and deionized water. Both substrates were mounted on a rotating holder and heated to 540 • C.
The crystal structure of the coating was investigated by X-ray diffraction (XRD) using Bruker D8 Discover (Cu radiation, λ = 1.5418 Å, Coventry, UK). Film measurement was taken in 2θ scan mode, with the source fixed at 8 • position, which avoided signal from the substrate while maintaining high intensity of the signal originating primarily from the coating.
The surface was investigated with scanning electron microscopy (SEM; JEOL JSM6010PLUS/LV, Tokyo, Japan). Energy dispersive X-ray spectroscopy (EDS; JEOL (EUROPE) SA, Warszawa, Poland) microanalysis was used to study the elemental distribution. During EDS measurement, accelerating voltage of 7 kV was used. Due to problems with boron measurement and subsequently determining its quantity in combination with other, especially heavy elements, before measurement, the EDS spectrometer was calibrated based on polished W 2 B 5 commercial standard (Huizhou Tian Yi Rare Material Co, Guang Dong, China, 99.9% purity).
Vickers microhardness tests were performed under 20 g load on a Hanneman tester (Zeiss, Jena, Germany). Five imprints were done on each tested sample. To evaluate film hardness, the substrate hardness was taken into account.
Surface roughness was measured on a Talysurf stylus scanning profilometer (Taylor Hobsson, Leicester, UK) according to ISO standards. The tip radius was 2 µm, traverse length according to ISO 4288 [24] was 1.25 mm, and cut-off was 0.25 mm. A standardized phase-correct Gaussian high-pass filter (ISO 11562 [25]) was applied. The profiles and surface measurements were done before and after the deposition process.
Wear tests were done on a Skoda-Savin tester (Skoda Works, Plzeň, Czech Republic) in semi-dry conditions. The test allows determination of wear at the semi-dry sliding friction. Determination of wear was made during rotational movement of the counter-sample made of WC-Co alloy sliding on a stationary sample while wetting the surface with cooling fluid. The following wear processes may occur: abrasion, chipping, surface plastic deformation, and adhesion. Figure 1 shows the scheme of the wear test.
The roller was pressed against the tested surface at constant load of 15 kg. Relative motion speed was 1.5 m/s. Preliminary tests showed that after 500 rotations, the depth of the worn area was less than 1 µm, so each test was stopped after 500 rotations, corresponding to the length of the friction span of 170 m. The coolant, 0.05% K 2 CrO 4 solution in distilled water, was applied with flow efficiency of 1 L/min. Five tests were performed on each tested surface. After the tests, surface topography was studied by laser confocal microscopy (Keyence VK-X100, Osaka, Japan) and the volume of worn material was determined. The roller was pressed against the tested surface at constant load of 15 kg. Relative motion spee s 1.5 m/s. Preliminary tests showed that after 500 rotations, the depth of the worn area was les n 1 μm, so each test was stopped after 500 rotations, corresponding to the length of the frictio an of 170 m. The coolant, 0.05% K2CrO4 solution in distilled water, was applied with flow efficienc 1 L/min. Five tests were performed on each tested surface. After the tests, surface topography wa died by laser confocal microscopy (Keyence VK-X100, Osaka, Japan) and the volume of wor terial was determined.
The cutting test was performed on a Haas CNC TL1 lathe (Haas Automation, Oxnard, CA, USA e turning process was performed on AISI 304 stainless steel. The constant cutting length was 7 and the following process parameters were applied: cutting speed vc = 80 m/min, cutting dept = 1.5 mm, and feed rate p = 0.09 mm/rev. Two commercially produced cemented carbide (WC-Co tting tool inserts, supplied by Sandvik Polska Sp. z o.o. (Warsaw, Poland), were tested with an thout deposited coatings. Nominal composition of inserts was 94% WC and 6% Co (H13A). Fiv ts were performed on each insert. The insert was rotated about 70° to perform the next test. Th tting tool with marked test location is illustrated in Figure 2. After the cutting tests, the wear o erts was observed by laser confocal microscopy and flank wear VB was determined.  The cutting test was performed on a Haas CNC TL1 lathe (Haas Automation, Oxnard, CA, USA). The turning process was performed on AISI 304 stainless steel. The constant cutting length was 70 mm and the following process parameters were applied: cutting speed v c = 80 m/min, cutting depth a p = 1.5 mm, and feed rate p = 0.09 mm/rev. Two commercially produced cemented carbide (WC-Co) cutting tool inserts, supplied by Sandvik Polska Sp. z o.o. (Warsaw, Poland), were tested with and without deposited coatings. Nominal composition of inserts was 94% WC and 6% Co (H13A). Five tests were performed on each insert. The insert was rotated about 70 • to perform the next test. The cutting tool with marked test location is illustrated in Figure 2. After the cutting tests, the wear of inserts was observed by laser confocal microscopy and flank wear V B was determined. The roller was pressed against the tested surface at constant load of 15 kg. Relative motion s 1.5 m/s. Preliminary tests showed that after 500 rotations, the depth of the worn area wa 1 μm, so each test was stopped after 500 rotations, corresponding to the length of the fr of 170 m. The coolant, 0.05% K2CrO4 solution in distilled water, was applied with flow effic /min. Five tests were performed on each tested surface. After the tests, surface topograph ied by laser confocal microscopy (Keyence VK-X100, Osaka, Japan) and the volume of rial was determined. The cutting test was performed on a Haas CNC TL1 lathe (Haas Automation, Oxnard, CA, U turning process was performed on AISI 304 stainless steel. The constant cutting length w and the following process parameters were applied: cutting speed vc = 80 m/min, cutting d .5 mm, and feed rate p = 0.09 mm/rev. Two commercially produced cemented carbide (WC ng tool inserts, supplied by Sandvik Polska Sp. z o.o. (Warsaw, Poland), were tested wit out deposited coatings. Nominal composition of inserts was 94% WC and 6% Co (H13A) were performed on each insert. The insert was rotated about 70° to perform the next tes ng tool with marked test location is illustrated in Figure 2. After the cutting tests, the we ts was observed by laser confocal microscopy and flank wear VB was determined.

icrostructure
The topography and chemical composition of deposited coating surface are presented in F th films were smooth with single impurities. EDS analysis ( Figure 3) showed that all elem uniformly distributed. As it is shown on exemplary distribution of tungsten (Figure 3a,b), the micro-segregation of elements.

Microstructure
The topography and chemical composition of deposited coating surface are presented in Figure 3. Both films were smooth with single impurities. EDS analysis ( Figure 3) showed that all elements were uniformly distributed. As it is shown on exemplary distribution of tungsten (Figure 3a,b), there is no the micro-segregation of elements.   Table 1 shows the chemical composition of deposited coatings. The addition of titanium caused a decrease in tungsten from 28.7 at.% to 23.1 at.%. However, it also changed the amount of oxygen from 4.1 at.% to 5 at.% and boron from 67.2 at.% to 68.3 at.%. A significant decrease in the amount of boron atoms in the deposited layers (B/(W + Ti) = 2.57) compared with the targets (4.5) can be explained by the reflection and scattering of the light element as a result of collisions with tungsten, which has an atomic mass 17 times greater [16]. Substitution of heavy tungsten by titanium causes more boron to reach the substrate, which is confirmed by the composition of the deposited layers (Table 1). In the case of oxygen, which possesses very high chemical activity [27], an increased amount is insignificant. It indicates that titanium can be in solid solution with crystal lattice of WB2. Taking into consideration that in the low-temperature region under equilibrium conditions only 2% of titanium can be in solid solution with WB2 lattice [28], the deposited coatings consist also of  Table 1 shows the chemical composition of deposited coatings. The addition of titanium caused a decrease in tungsten from 28.7 at.% to 23.1 at.%. However, it also changed the amount of oxygen from 4.1 at.% to 5 at.% and boron from 67.2 at.% to 68.3 at.%. A significant decrease in the amount of boron atoms in the deposited layers (B/(W + Ti) = 2.57) compared with the targets (4.5) can be explained by the reflection and scattering of the light element as a result of collisions with tungsten, which has an atomic mass 17 times greater [16]. Substitution of heavy tungsten by titanium causes more boron to reach the substrate, which is confirmed by the composition of the deposited layers (Table 1). In the case of oxygen, which possesses very high chemical activity [27], an increased amount is insignificant. It indicates that titanium can be in solid solution with crystal lattice of WB 2 . Taking into Coatings 2020, 10, 1231 6 of 13 consideration that in the low-temperature region under equilibrium conditions only 2% of titanium can be in solid solution with WB 2 lattice [28], the deposited coatings consist also of nanocomposite mixture of WB 2 and TiB 2 . Such a hypothesis can be confirmed by XRD ( Figure 4) and XPS studies, which were interpreted in more detail by Moscicki et al. [4].
As shown in Figure 4, the addition of titanium changed the structure of deposited coatings from crystalline ( Figure 4a) to amorphous (Figure 4b). The only reflection (0001) at angle 2θ = 28° indicates a strongly oriented columnar structure of WB2. The addition of Ti resulted in the appearance of a second nanocrystalline phase. Investigations on diborides show that those materials prefer to crystallize mainly in two related hexagonal structures: AlB2 structure type P6/mmm, preferably formed by titanium diborides, or W2B5−x-based structure P63/mmc, favored by WB2 [29]. However, vacancies, which are typical for magnetron sputtered materials, are beneficial for stabilizing the αstructure [29]. Additionally, the formation enthalpy is lower for α-than ω-structure for high Ti content, which promotes rebuilding of P63/mmc to P6/mmm structure and consequent amorphization of coating. The X-ray diffraction patterns of nanocrystalline (W,Ti)B2 coating can be seen in Figure 4b. Such a change increases the fracture toughness [14,20] but does not affect the very high hardness of deposited films.

Surface Topography
The thickness of the coatings was measured in a stepwise manner on polished samples by confocal microscope. It was seen that the thickness of W-B film was 1.81 ± 0.11 μm and W-Ti-B film was 1.43 ± 0.15 μm. Surface was smooth with single impurities.
Surface roughness was measured on flat polished samples and on cutting inserts before and after the deposition process. The surface topography of the substrate is still noticeable after the deposition process. Deposited films covered the substrate, keeping the shape of the roughness.
Roughness parameters are shown in Table 2. After deposition, the average value of the height of roughness (Ra, Rq), maximum profile valley depth (Rv), maximum height of profile (Rz) and maximum height of peaks (Rp) increases for both films. This is related to the presence of single impurities on the surface after deposition. Spatial parameter Sm does not change and corresponds to the distribution of carbides on the substrate. As shown in Figure 4, the addition of titanium changed the structure of deposited coatings from crystalline (Figure 4a) to amorphous (Figure 4b). The only reflection (0001) at angle 2θ = 28 • indicates a strongly oriented columnar structure of WB 2 . The addition of Ti resulted in the appearance of a second nanocrystalline phase. Investigations on diborides show that those materials prefer to crystallize mainly in two related hexagonal structures: AlB 2 structure type P6/mmm, preferably formed by titanium diborides, or W 2 B 5−x -based structure P63/mmc, favored by WB 2 [29]. However, vacancies, which are typical for magnetron sputtered materials, are beneficial for stabilizing the α-structure [29]. Additionally, the formation enthalpy is lower for αthan ω-structure for high Ti content, which promotes rebuilding of P63/mmc to P6/mmm structure and consequent amorphization of coating. The X-ray diffraction patterns of nanocrystalline (W,Ti)B 2 coating can be seen in Figure 4b. Such a change increases the fracture toughness [14,20] but does not affect the very high hardness of deposited films.

Surface Topography
The thickness of the coatings was measured in a stepwise manner on polished samples by confocal microscope. It was seen that the thickness of W-B film was 1.81 ± 0.11 µm and W-Ti-B film was 1.43 ± 0.15 µm. Surface was smooth with single impurities.
Surface roughness was measured on flat polished samples and on cutting inserts before and after the deposition process. The surface topography of the substrate is still noticeable after the deposition process. Deposited films covered the substrate, keeping the shape of the roughness.
Roughness parameters are shown in Table 2. After deposition, the average value of the height of roughness (R a , R q ), maximum profile valley depth (R v ), maximum height of profile (R z ) and maximum height of peaks (R p ) increases for both films. This is related to the presence of single impurities on the Coatings 2020, 10, 1231 7 of 13 surface after deposition. Spatial parameter S m does not change and corresponds to the distribution of carbides on the substrate. The topography of the surface can be better revealed by 3-D surface roughness measurement. Figure 5 shows surface topography measured on a scanning profilometer. Waviness related to polishing process is observed for each sample. In the case of W-Ti-B coating, more impurities are visible as picks ( Figure 5c) and the value of area roughness S a , surface roughness height of peaks S p and valley S v , as well as maximum height S z (Table 3), are bigger than for WC-Co and W-B film, similar to the profile parameters.  The topography of the surface can be better revealed by 3-D surface roughness measurement. Figure 5 shows surface topography measured on a scanning profilometer. Waviness related to polishing process is observed for each sample. In the case of W-Ti-B coating, more impurities are visible as picks (Figure 5c) and the value of area roughness Sa, surface roughness height of peaks Sp and valley Sv, as well as maximum height Sz (Table 3), are bigger than for WC-Co and W-B film, similar to the profile parameters.   The profile roughness measurement was also done on cutting tools. Figure 6 shows the topography of the surface of cutting inserts observed on confocal microscope without and with deposited W-B film. In the case of WC-Co insert, small carbides in the matrix with characteristic sharp shape for WC grains were observed (Figure 6a). After deposition of W-B coating, the surface morphology changed, however single carbides were still observed (Figure 6b).  The profile roughness measurement was also done on cutting tools. Figure 6 shows the topography of the surface of cutting inserts observed on confocal microscope without and with deposited W-B film. In the case of WC-Co insert, small carbides in the matrix with characteristic sharp shape for WC grains were observed (Figure 6a). After deposition of W-B coating, the surface morphology changed, however single carbides were still observed (Figure 6b). The average value of profile roughness height (Ra, Rq) and maximum height Rz of the peak are smaller after the deposition process. Deposited material is located on the roughness valley, causing a smaller value of parameter Rv (Table 4) compared to the uncoated insert.

Microhardness
Microhardness tests showed that for the sample with W-B coating deposited on WC-Co substrate, the hardness value was 2680 HV0.02, with standard deviation of 360. As a result that the depth of imprints was about 0.35 μm, it was necessary to take into account the hardness of the substrate to evaluate the hardness of the film. Based on the formula in [30] and substrate hardness 2000 HV0.02, the hardness of the W-B film was estimated at 2960 ± 440 HV0.02. This value is close to the one obtained for a similar coating presented in [9]. The sample with W-Ti-B coating was harder (2810 ± 130 HV0.02) and estimated hardness of the film without the influence of substrate was 3630 ± 260 HV0.02. The increased hardness of titanium alloyed coatings can be related to different mechanisms, e.g., solid solution hardening [31] or precipitation hardening [32]. In the case of W-Ti-B nanocomposite, which consists of coherently coupled grains of two phases with a hexagonal crystal structure (α-WB2 and TiB2), the binding strength is increased. According to XRD studies (Figure 4), such a composite is formed during RF magnetron sputtering. The creation of coherent boundaries between the crystallites of different phases increases the hardness of the composite. The formation of W Figure 6. Surface topography of (a) cutting insert WC-Co, (b) insert with deposited W-B film.
The average value of profile roughness height (R a , R q ) and maximum height R z of the peak are smaller after the deposition process. Deposited material is located on the roughness valley, causing a smaller value of parameter R v (Table 4) compared to the uncoated insert.

Microhardness
Microhardness tests showed that for the sample with W-B coating deposited on WC-Co substrate, the hardness value was 2680 HV 0.02 , with standard deviation of 360. As a result that the depth of imprints was about 0.35 µm, it was necessary to take into account the hardness of the substrate to evaluate the hardness of the film. Based on the formula in [30] and substrate hardness 2000 HV 0.02 , the hardness of the W-B film was estimated at 2960 ± 440 HV 0.02 . This value is close to the one obtained for a similar coating presented in [9]. The sample with W-Ti-B coating was harder (2810 ± 130 HV 0.02 ) and estimated hardness of the film without the influence of substrate was 3630 ± 260 HV 0.02 . The increased hardness of titanium alloyed coatings can be related to different mechanisms, e.g., solid solution hardening [31] or precipitation hardening [32]. In the case of W-Ti-B nanocomposite, which consists of coherently coupled grains of two phases with a hexagonal crystal structure (α-WB 2 and TiB 2 ), the binding strength is increased. According to XRD studies (Figure 4), such a composite is formed during RF magnetron sputtering. The creation of coherent boundaries between the crystallites of Coatings 2020, 10, 1231 9 of 13 different phases increases the hardness of the composite. The formation of such boundaries between grains causes suppression of grain-boundary sliding in nanocrystalline materials [15]. This leads to increased film strength and hardness.

Wear Test
Wear test was performed for WC-Co flat samples with W-B and W-Ti-B coatings. Results were compared with wear resistance of WC-Co samples without coating. After testing the surface of samples, a groove was observed. For WC-Co samples, 1.9 µm groove depth was noted. Figure 7a shows an example of worn groove on W-B coating. In this case, the maximum depth of grooves after the wear test was 1.4 µm, which was less than the thickness of the film (1.8 µm). Figure 7 shows that the surface of the grooves was smooth with some irregularities on boundary, with unworn material. In the case of W-Ti-B film, cracks were observed on the groove boundary (Figure 8b). Table 5 shows the volume of lost material for tested samples after wear tests measured with confocal microscope and calculated based on the analysis program.
Coatings 2020, 10, x FOR PEER REVIEW 9 of 14 such boundaries between grains causes suppression of grain-boundary sliding in nanocrystalline materials [15]. This leads to increased film strength and hardness.

Wear Test
Wear test was performed for WC-Co flat samples with W-B and W-Ti-B coatings. Results were compared with wear resistance of WC-Co samples without coating. After testing the surface of samples, a groove was observed. For WC-Co samples, 1.9 μm groove depth was noted. Figure 7a shows an example of worn groove on W-B coating. In this case, the maximum depth of grooves after the wear test was 1.4 μm, which was less than the thickness of the film (1.8 μm). Figure 7 shows that the surface of the grooves was smooth with some irregularities on boundary, with unworn material. In the case of W-Ti-B film, cracks were observed on the groove boundary (Figure 8b). Table 5 shows the volume of lost material for tested samples after wear tests measured with confocal microscope and calculated based on the analysis program.  The highest wear resistance was noted for the sample with W-B film. The volume of lost material was about 40% less than for the uncoated WC-Co sample. In the case of W-Ti-B film, wear resistance was the worst and even lower than the uncoated sample.
Microscopic observation of worn areas revealed that the W-B film surface was smooth without cracks. Figure 8a shows a boundary area without cracks and other imperfections. Results indicated that abrasive wear was a dominant process in this case. For the W-Ti-B film, cracks and delamination of coating were observed on groove boundaries (Figure 8b). This phenomenon can explain the higher material loss of W-Ti-B film compared to other samples and the varying values of the volume loss of material in this case. Very hard delaminated fragments of the films can accelerate wear before they are removed from the treatment zone by the coolant.

Turning Test
Cutting test was performed for WC-Co insert and for insert with W-B coating. After the cutting tests, the inserts were investigated by laser confocal microscopy. Wear was measured on the flank of the insert and the flank wear VB was determined. For both tested inserts, smooth abrasive wear was observed. Figure 9 shows cutting edges after the test. Flank wear was less for W-B coated inserts (Figure 9a) than for WC-Co inserts (Figure 9d). The detailed observation reveals additional wear  The highest wear resistance was noted for the sample with W-B film. The volume of lost material was about 40% less than for the uncoated WC-Co sample. In the case of W-Ti-B film, wear resistance was the worst and even lower than the uncoated sample.
Microscopic observation of worn areas revealed that the W-B film surface was smooth without cracks. Figure 8a shows a boundary area without cracks and other imperfections. Results indicated that abrasive wear was a dominant process in this case. For the W-Ti-B film, cracks and delamination of coating were observed on groove boundaries (Figure 8b). This phenomenon can explain the higher material loss of W-Ti-B film compared to other samples and the varying values of the volume loss of material in this case. Very hard delaminated fragments of the films can accelerate wear before they are removed from the treatment zone by the coolant.

Turning Test
Cutting test was performed for WC-Co insert and for insert with W-B coating. After the cutting tests, the inserts were investigated by laser confocal microscopy. Wear was measured on the flank of the insert and the flank wear V B was determined. For both tested inserts, smooth abrasive wear was observed. Figure 9 shows cutting edges after the test. Flank wear was less for W-B coated inserts (Figure 9a) than for WC-Co inserts (Figure 9d). The detailed observation reveals additional wear mechanisms that are different for each tool. Figure 9b,c show adherent material. Figure 9c shows a profile of the cutting edge after machining. A change in insert geometry caused by build-up was observed in both cases. This mechanism is caused by pressure welding of the chip to the insert. It is most common during machining of sticky materials with low cutting speed. To prevent this kind of wear, the highest speed should be applied.
For the WC-Co insert, flank wear was combined with chipping of the cutting edge visible on Figure 9e,f. This mechanical wear is more dangerous than adhesive wear and can lead to rapid tool failure.

Turning Test
Cutting test was performed for WC-Co insert and for insert with W-B coating. After the cutting tests, the inserts were investigated by laser confocal microscopy. Wear was measured on the flank of the insert and the flank wear VB was determined. For both tested inserts, smooth abrasive wear was observed. Figure 9 shows cutting edges after the test. Flank wear was less for W-B coated inserts (Figure 9a) than for WC-Co inserts (Figure 9d). The detailed observation reveals additional wear mechanisms that are different for each tool. Figure 9b,c show adherent material. Figure 9c shows a profile of the cutting edge after machining. A change in insert geometry caused by build-up was observed in both cases. This mechanism is caused by pressure welding of the chip to the insert. It is most common during machining of sticky materials with low cutting speed. To prevent this kind of wear, the highest speed should be applied.
For the WC-Co insert, flank wear was combined with chipping of the cutting edge visible on Figure 9e,f. This mechanical wear is more dangerous than adhesive wear and can lead to rapid tool failure. Table 6 shows measurement results of the mean value of flank wear VB for a constant cutting length of 70 mm. Wear of the insert with W-B film was smaller than 30% compared to uncoated insert ( Figure 10). That is compatible with the results of the wear test, in which 40% less volume of material was found (Table 5).    Table 6 shows measurement results of the mean value of flank wear V B for a constant cutting length of 70 mm. Wear of the insert with W-B film was smaller than 30% compared to uncoated insert ( Figure 10). That is compatible with the results of the wear test, in which 40% less volume of material was found (Table 5).

Conclusions
In this work, properties of boride coatings such as WB x and (W,Ti)B 2 deposited by RF magnetron sputtering and the possibility of using them to protect WC-Co cutting tools for difficult-to-machine materials are described. The main conclusions are as follows: • W-B and W-Ti-B coatings deposited on WC-Co substrate are smooth and very hard. However, titanium alloy films with hardness of 3630 ± 260 HV 0.02 are characterized by lower adhesion to the substrate, which influences the wear mechanism.

•
The wear test at semi-dry sliding friction revealed 40% better wear resistance for W-B coating compared to WC-Co and W-Ti-B coating. In case of the W-Ti-B film, delamination and cracking were noted.

•
In the turning test of difficult-to-cut 304 stainless steel, the W-B coated tool showed better wear resistance than the uncoated tool. • Flank wear was smaller by 30% when W-B film was applied, compared to uncoated WC-Co insert. Additionally, coating prevented chipping of the edge during cutting in tested conditions. • W-B film deposited by magnetron sputtering has great potential as a coating for cutting tools for difficult-to-cut materials, but it is necessary to carry out more tests under various cutting conditions.