Structure and Frictional Properties of Ultrahard AlMgB14 Thin Coatings

This paper presents the results of studies on AlMgB14-based ceramic coatings deposited on WC-Co hard alloy substrates using RF plasma sputtering. The aim of this work is to study the structure, phase composition, and mechanical properties of AlMgB14-based coatings depending on the sputtering mode. According to the results of the microstructural study, the bias voltage applied to the substrate during the sputtering process significantly contributed to the formation of the coating morphology. Based on the results of compositional and structural studies by energy dispersive X-ray spectroscopy, X-ray diffraction, and Raman spectroscopy, it was found that the coatings are composed of nanocrystalline B12 icosahedrons distributed in an amorphous matrix consisting of Al, Mg, B, and O elements. The nanohardness of the coatings varied from 24 GPa to 37 GPa. The maximum value of the hardness together with the lowest coefficient of friction (COF) equal to 0.12 and wear resistance of 7.5 × 10−5 mm3/N·m were obtained for the coating sputtered at a bias voltage of 100 V. Compared with the COF of the original hard alloy substrate, which is equal to 0.31, it can be concluded that the AlMgB14-based coatings could reduce the COF of WC-based hard alloys by more than two times. The hardness and tribological properties of the coatings obtained in this study are in good agreement with the properties of AlMgB14-based materials obtained by other methods reported in the literature.


Introduction
Improving wear resistance while reducing the coefficient of friction of the surface of materials used in the manufacture of friction pairs and machining tools is a current scientific and technical challenge. The most widespread and promising method is to modify the surface of parts by sputtering coatings, which can significantly increase hardness and wear resistance and reduce the coefficient of friction compared to the base material. Currently, a high level of development has been achieved in this direction. Nitride-based coatings such as Si 3 N 4 , TiN, and c-BN [1,2]; carbides: SiC and WC [1,3]; oxides: Al 2 O 3 and Cr 2 O 3 [4]; borides: TiB 2 [5]; and diamond-like carbon (DLC) coatings [6] are widely used. Research into fundamentally new materials that offer a multiple reduction in the coefficient of friction, while simultaneously increasing hardness and wear resistance, is of great scientific importance. Among these materials, aluminum magnesium boride AlMgB 14 -based ceramics could be highlighted [7][8][9].
According to [8], the coefficient of friction of AlMgB 14 -based coatings can reach 0.02 with lubrication, while the hardness reaches 35 GPa and higher. According to [9], AlMgB 14based coatings have a lower coefficient of friction than the "industry leading" DLC coating. In addition to testing the coefficient of friction under laboratory conditions, the authors [9] demonstrated the effectiveness of AlMgB 14 -based coatings when applied to the friction pairs of a hydraulic pump. The coefficient of friction of AlMgB 14 -based coatings is reduced The starting AlMgB 14 -based powder material was obtained by the SHS method developed in our previous work [17] from a premix of Al 12 Mg 17 alloy and amorphous boron powders in the stoichiometric atomic ratio of 2:14. The SHS was performed using a Ti-Si chemical furnace as the heat energy source. First, the stoichiometric Al 12 Mg 17 :B premix was mechanically activated for 3 hours in an argon atmosphere with steel grinding bodies at a ratio of 3:1 to the mass of the powder to ensure uniform mixing of the components and enhance their reactivity. The powder was then cold-pressed into 23 mm diameter specimens weighing 15 g each. To form a chemical furnace, the donor mixture of titanium and silicon in the stoichiometric ratio for Ti 5 Si 3 synthesis (74 wt. % Ti + 26 wt. % Si) was poured into a cylindrical cellulose paper container, into which the acceptor sample was placed. The optimum thickness of the chemical furnace was 3 mm, as determined in [17]. The resulting system was placed in a constant pressure reactor. The reactor was evacuated, and then its working space was filled with argon to a pressure of 2 bar. The synthesis reaction was initiated by short-term heating of the upper surface of the system with a molybdenum spiral. After synthesis, the specimen was separated from the chemical furnace and ground to powder by hand in a mortar. The obtained powder was further ground in a planetary ball mill at 14 Hz for 40 min with tungsten carbide grinding bodies at a ratio of 3:1 to the mass of the powder. The ground powder was then sieved through a 40 µm mesh. After grinding and sieving, the obtained AlMgB 14 -based powder had a Nanomaterials 2023, 13, 1589 3 of 12 uniform structure, represented by angular grains of AlMgB 14 . The average particle size of the powder was 3.5 µm. According to the results of X-ray phase analysis, the obtained powder material contained 91 wt. % of AlMgB 14 phase and 9 wt. % of the spinel MgAl 2 O 4 phase, formed due to the presence of oxygen in the initial boron powder [17].

RF Plasma Sputtering
For sputtering of AlMgB 14 -based coatings, electron-ion-plasma equipment "COM-PLEX" (IHCE SB RAS, Tomsk, Russia) was used [18]. The equipment combines ion-plasma etching of the sample surface and the deposition of coatings from powder or solid targets. A schematic of the sputtering equipment is shown in Figure 1.
Nanomaterials 2023, 13, x FOR PEER REVIEW 3 of 12 at a ratio of 3:1 to the mass of the powder. The ground powder was then sieved through a 40 μm mesh. After grinding and sieving, the obtained AlMgB14-based powder had a uniform structure, represented by angular grains of AlMgB14. The average particle size of the powder was 3.5 μm. According to the results of X-ray phase analysis, the obtained powder material contained 91 wt. % of AlMgB14 phase and 9 wt. % of the spinel MgAl2O4 phase, formed due to the presence of oxygen in the initial boron powder [17].

RF Plasma Sputtering
For sputtering of AlMgB14-based coatings, electron-ion-plasma equipment "COM-PLEX" (IHCE SB RAS, Tomsk, Russia) was used [18]. The equipment combines ionplasma etching of the sample surface and the deposition of coatings from powder or solid targets. A schematic of the sputtering equipment is shown in Figure 1. The sputtering process consisted of the following steps. The substrate and AlMgB14 powder target were placed in a vacuum chamber and evacuated to a pressure of 5 × 10 −3 Pa. After argon was introduced into the chamber at a pressure of 0.1-0.4 Pa, the plasma generator was turned on. First, a negative bias voltage was applied to the substrate to clean and activate the surface using argon plasma. After the surface was treated and the substrate was heated to 450 °C, the RF generator connected to the target was turned on to initiate the sputtering of the coating material onto the substrate surface. After sputtering, the substrate was cooled to a temperature of less than 100 °C under vacuum. The coatings were sputtered at bias voltages ranging from 35 V to 150 V. Table 1 lists the values of the main sputtering parameters.

Characterization
The microstructures of the coatings were studied using an optical microscope (METAM LV-34, Saint Petersburg, Russia) and a scanning electron microscope (TESCAN Mira, Brno, Czech Republic). The morphology of the coatings was studied using an atomic The sputtering process consisted of the following steps. The substrate and AlMgB 14 powder target were placed in a vacuum chamber and evacuated to a pressure of 5 × 10 −3 Pa. After argon was introduced into the chamber at a pressure of 0.1-0.4 Pa, the plasma generator was turned on. First, a negative bias voltage was applied to the substrate to clean and activate the surface using argon plasma. After the surface was treated and the substrate was heated to 450 • C, the RF generator connected to the target was turned on to initiate the sputtering of the coating material onto the substrate surface. After sputtering, the substrate was cooled to a temperature of less than 100 • C under vacuum. The coatings were sputtered at bias voltages ranging from 35 V to 150 V. Table 1 lists the values of the main sputtering parameters.

Characterization
The microstructures of the coatings were studied using an optical microscope (METAM LV-34, Saint Petersburg, Russia) and a scanning electron microscope (TESCAN Mira, Brno, Czech Republic). The morphology of the coatings was studied using an atomic force microscope (AFM) NT MDT (Moscow, Russia). The elemental composition of the coating cross-section was studied using energy dispersive X-ray spectroscopy (EDX) (Oxford Instrument, Abingdon, UK) during microstructural studies. The crystal structure was studied by X-ray diffraction analysis using a Shimadzu XRD 6000 diffractometer (Kyoto, Japan) with CuKα radiation and Raman spectroscopy using a Renishaw spectrometer (Wotton-under-Edge, UK) at a wavelength of 785 nm. XRD analysis of the coatings was performed in sliding beam mode with a sliding angle of 2 • . Nanoindentation of the coatings was performed using a CSM-Instruments (Peseuxm Switzerland) desktop nanoindentation system with a load of 15 mN and exposure time of 5 s. Nanohardness was determined using the Oliver and Farr method [19]. The coefficient of friction and wear rate were determined using the pin-on-disc method with a 100Cr6 steel ball under dry conditions at room temperature with a load of 1 N and test speed of 25 mm/s using a TRIBOtechnic (Clichy, France) tribometer. Figure 2 shows the surface microstructure images of AlMgB 14 -based coatings obtained at different bias voltages. force microscope (AFM) NT MDT (Moscow, Russia). The elemental composition of the coating cross-section was studied using energy dispersive X-ray spectroscopy (EDX) (Oxford Instrument, Abingdon, UK) during microstructural studies. The crystal structure was studied by X-ray diffraction analysis using a Shimadzu XRD 6000 diffractometer (Kyoto, Japan) with CuKα radiation and Raman spectroscopy using a Renishaw spectrometer (Wotton-under-Edge, UK) at a wavelength of 785 nm. XRD analysis of the coatings was performed in sliding beam mode with a sliding angle of 2°. Nanoindentation of the coatings was performed using a CSM-Instruments (Peseuxm Switzerland) desktop nanoindentation system with a load of 15 mN and exposure time of 5 s. Nanohardness was determined using the Oliver and Farr method [19]. The coefficient of friction and wear rate were determined using the pin-on-disc method with a 100Cr6 steel ball under dry conditions at room temperature with a load of 1 N and test speed of 25 mm/s using a TRIBOtechnic (Clichy, France) tribometer. Figure 2 shows the surface microstructure images of AlMgB14-based coatings obtained at different bias voltages. As shown in Figure 2, the microstructure of the AlMgB14-based coating sputtered at a bias voltage of up to 50 V is represented by angular regions with sizes of 2-5 μm. As the bias voltage increases, the microstructure of the coating surface is represented by rounded As shown in Figure 2, the microstructure of the AlMgB 14 -based coating sputtered at a bias voltage of up to 50 V is represented by angular regions with sizes of 2-5 µm. As the bias voltage increases, the microstructure of the coating surface is represented by rounded clusters. The higher the bias voltage is, the larger are the individual clusters. Thus, for the coating sputtered at a bias voltage of 100 V, the average size of the clusters representing Nanomaterials 2023, 13, 1589 5 of 12 the surface microstructure is 2-3 µm, whereas for the coating sputtered at a bias voltage of 150 V, the size of individual clusters reaches 5 µm.

Microstructure
AFM images representing the morphology of the sputtered AlMgB 14 -based coatings are shown in Figure 3.
clusters. The higher the bias voltage is, the larger are the individual clusters. Thus, for the coating sputtered at a bias voltage of 100 V, the average size of the clusters representing the surface microstructure is 2-3 μm, whereas for the coating sputtered at a bias voltage of 150 V, the size of individual clusters reaches 5 μm.
AFM images representing the morphology of the sputtered AlMgB14-based coatings are shown in Figure 3. According to the AFM images ( Figure 3), the angular regions formed on the surface of the AlMgB14-based coatings sputtered at a bias voltage of 35 V were up to 300 nm in height. As the bias voltage increased up to 150 V, the morphology of the coating surface was represented by spherical cavities with depths of up to 300 nm. The average roughness calculated from the AFM results for the 100 × 100 μm regions is 22.05 nm for the AlMgB14based coating sputtered at a bias voltage of 35 V and 75.3 nm for the coating sputtered at a bias voltage of 150 V.  The SEM image ( Figure 4) shows a clear boundary between the coating and substrate. As shown in Figure 4, the sputtered coating was 3 μm thick. The EDX mapping shows According to the AFM images (Figure 3), the angular regions formed on the surface of the AlMgB 14 -based coatings sputtered at a bias voltage of 35 V were up to 300 nm in height. As the bias voltage increased up to 150 V, the morphology of the coating surface was represented by spherical cavities with depths of up to 300 nm. The average roughness calculated from the AFM results for the 100 × 100 µm regions is 22.05 nm for the AlMgB 14based coating sputtered at a bias voltage of 35 V and 75.3 nm for the coating sputtered at a bias voltage of 150 V. clusters. The higher the bias voltage is, the larger are the individual clusters. Thus, for the coating sputtered at a bias voltage of 100 V, the average size of the clusters representing the surface microstructure is 2-3 μm, whereas for the coating sputtered at a bias voltage of 150 V, the size of individual clusters reaches 5 μm. AFM images representing the morphology of the sputtered AlMgB14-based coatings are shown in Figure 3. According to the AFM images (Figure 3), the angular regions formed on the surface of the AlMgB14-based coatings sputtered at a bias voltage of 35 V were up to 300 nm in height. As the bias voltage increased up to 150 V, the morphology of the coating surface was represented by spherical cavities with depths of up to 300 nm. The average roughness calculated from the AFM results for the 100 × 100 μm regions is 22.05 nm for the AlMgB14based coating sputtered at a bias voltage of 35 V and 75.3 nm for the coating sputtered at a bias voltage of 150 V.  The SEM image ( Figure 4) shows a clear boundary between the coating and substrate. As shown in Figure 4, the sputtered coating was 3 μm thick. The EDX mapping shows The SEM image ( Figure 4) shows a clear boundary between the coating and substrate. As shown in Figure 4, the sputtered coating was 3 µm thick. The EDX mapping shows that the coating is represented by uniformly distributed elements of the Al-Mg-B system. In addition to Al-Mg-B, oxygen was uniformly distributed in the coating structure. This can Nanomaterials 2023, 13, 1589 6 of 12 be explained by the presence of oxygen in the initial AlMgB 14 target powder. The EDX maps obtained for W and Co are typical for WC-Co hard alloy substrate materials.

Composition and Crystal Structure
The results of the XRD analysis of the AlMgB 14 -based coating sputtered at a bias voltage of 100 V are shown in Figure 5. The XRD analysis was carried out with a standard exposure time of 1 s ( Figure 5a) and with an exposure time of 10 s (Figure 5b) for the angles 4-20 • (area marked with the letter "b" in Figure 5a). that the coating is represented by uniformly distributed elements of the Al-Mg-B system. In addition to Al-Mg-B, oxygen was uniformly distributed in the coating structure. This can be explained by the presence of oxygen in the initial AlMgB14 target powder. The EDX maps obtained for W and Co are typical for WC-Co hard alloy substrate materials.
The results of the XRD analysis of the AlMgB14-based coating sputtered at a bias voltage of 100 V are shown in Figure 5. The XRD analysis was carried out with a standard exposure time of 1 s ( Figure 5a) and with an exposure time of 10 s (Figure 5b) for the angles 4-20° (area marked with the letter "b" in Figure 5a). According to Figure 5a, the XRD pattern obtained with 1 s exposure contains, in addition to reflexes related to the WC-Co substrate, a broad reflex at low angles in the region of 10°. Based on the XRD pattern shown in Figure 5a, this reflex could refer to both the crystalline and amorphous states. The XRD pattern obtained with a longer exposure of 10 s (Figure 5b) determines the nature of this broad reflex as an amorphous halo. Because the XRD pattern of the AlMgB14-based coating sputtered at a bias voltage of 100 V contains only the WC-Co substrate reflexes and an amorphous halo, it can be concluded that the coating was formed in the amorphous state during RF plasma sputtering. Figure 6 shows the Raman spectra of the AlMgB14-based coating sputtered at a bias voltage of 100 V. For comparison, the Raman spectra for the hot-pressed AlMgB14-based sample ( Figure 6, blue line) obtained in our previous work [20] are also shown. According to Figure 6, the obtained Raman spectra for the AlMgB14-based coating did not contain any narrow lines because the coating was formed in the amorphous state. According to Figure 5a, the XRD pattern obtained with 1 s exposure contains, in addition to reflexes related to the WC-Co substrate, a broad reflex at low angles in the region of 10 • . Based on the XRD pattern shown in Figure 5a, this reflex could refer to both the crystalline and amorphous states. The XRD pattern obtained with a longer exposure of 10 s (Figure 5b) determines the nature of this broad reflex as an amorphous halo. Because the XRD pattern of the AlMgB 14 -based coating sputtered at a bias voltage of 100 V contains only the WC-Co substrate reflexes and an amorphous halo, it can be concluded that the coating was formed in the amorphous state during RF plasma sputtering. Figure 6 shows the Raman spectra of the AlMgB 14 -based coating sputtered at a bias voltage of 100 V. For comparison, the Raman spectra for the hot-pressed AlMgB 14 -based sample ( Figure 6, blue line) obtained in our previous work [20] are also shown. Nanomaterials 2023, 13, x FOR PEER REVIEW 6 of 12 that the coating is represented by uniformly distributed elements of the Al-Mg-B system. In addition to Al-Mg-B, oxygen was uniformly distributed in the coating structure. This can be explained by the presence of oxygen in the initial AlMgB14 target powder. The EDX maps obtained for W and Co are typical for WC-Co hard alloy substrate materials. The results of the XRD analysis of the AlMgB14-based coating sputtered at a bias voltage of 100 V are shown in Figure 5. The XRD analysis was carried out with a standard exposure time of 1 s ( Figure 5a) and with an exposure time of 10 s (Figure 5b) for the angles 4-20° (area marked with the letter "b" in Figure 5a). According to Figure 5a, the XRD pattern obtained with 1 s exposure contains, in addition to reflexes related to the WC-Co substrate, a broad reflex at low angles in the region of 10°. Based on the XRD pattern shown in Figure 5a, this reflex could refer to both the crystalline and amorphous states. The XRD pattern obtained with a longer exposure of 10 s (Figure 5b) determines the nature of this broad reflex as an amorphous halo. Because the XRD pattern of the AlMgB14-based coating sputtered at a bias voltage of 100 V contains only the WC-Co substrate reflexes and an amorphous halo, it can be concluded that the coating was formed in the amorphous state during RF plasma sputtering. Figure 6 shows the Raman spectra of the AlMgB14-based coating sputtered at a bias voltage of 100 V. For comparison, the Raman spectra for the hot-pressed AlMgB14-based sample ( Figure 6, blue line) obtained in our previous work [20] are also shown. According to Figure 6, the obtained Raman spectra for the AlMgB14-based coating did not contain any narrow lines because the coating was formed in the amorphous state. According to Figure 6, the obtained Raman spectra for the AlMgB 14 -based coating did not contain any narrow lines because the coating was formed in the amorphous state. However, there is a broad band from 1000 cm −1 to 1200 cm −1 in the Raman spectra of both the coating and hot-pressed AlMgB 14 -based materials. According to the literature [21][22][23] this band can be attributed to the icosahedral B-B vibrational modes. This indicated that individual nanocrystalline B 12 icosahedrons were formed in the amorphous structure of the RF plasma-sputtered AlMgB 14 -based coating. However, there is a broad band from 1000 cm −1 to 1200 cm −1 in the Raman spectra of both the coating and hot-pressed AlMgB14-based materials. According to the literature [21][22][23] this band can be attributed to the icosahedral B-B vibrational modes. This indicated that individual nanocrystalline B12 icosahedrons were formed in the amorphous structure of the RF plasma-sputtered AlMgB14-based coating. Figure 7 shows the dependence of the nanohardness of the coatings on RF sputtering bias voltage. According to Figure 7, an increase in the RF sputtering bias voltage up to 100 V leads to an increase in the coating nanohardness up to 37 GPa. With a further increase in the RF sputtering bias voltage to 150 V, the nanohardness decreased to 32 GPa. The dependence of the tribological characteristics of the sputtered coatings is shown in Figure 8.  According to Figure 7, an increase in the RF sputtering bias voltage up to 100 V leads to an increase in the coating nanohardness up to 37 GPa. With a further increase in the RF sputtering bias voltage to 150 V, the nanohardness decreased to 32 GPa. The dependence of the tribological characteristics of the sputtered coatings is shown in Figure 8. However, there is a broad band from 1000 cm −1 to 1200 cm −1 in the Raman spectra of both the coating and hot-pressed AlMgB14-based materials. According to the literature [21][22][23] this band can be attributed to the icosahedral B-B vibrational modes. This indicated that individual nanocrystalline B12 icosahedrons were formed in the amorphous structure of the RF plasma-sputtered AlMgB14-based coating. Figure 7 shows the dependence of the nanohardness of the coatings on RF sputtering bias voltage. According to Figure 7, an increase in the RF sputtering bias voltage up to 100 V leads to an increase in the coating nanohardness up to 37 GPa. With a further increase in the RF sputtering bias voltage to 150 V, the nanohardness decreased to 32 GPa. The dependence of the tribological characteristics of the sputtered coatings is shown in Figure 8.  According to Figure 8a, the COF of the obtained AlMgB 14 -based coatings varies from 0.12 to 0.24 depending on the sputtering bias voltage. When the bias voltage is increased Nanomaterials 2023, 13, 1589 8 of 12 from 35 V to 50 V, the COF increases slightly by a value close to the measurement error. When the bias voltage is subsequently increased to 100 V, the COF decreases significantly to a minimum value of 0.12. For the bias voltage of 150 V, the COF increases again to the average value of 0.18. Thus, except for the maximum nanohardness of 37 GPa, the AlMgB 14 -based coating sputtered at a bias voltage of 100 V has the lowest coefficient of friction equal to 0.12 and a relatively low wear rate of 7.5 × 10 −5 mm 3 /N·m compared with the coatings sputtered at 35 V, 50 V, and 150 V bias voltages. Figure 9 shows the coefficient of friction as a function of sliding time for the WC-based substrate and the AlMgB 14 -based coating sputtered at 100 V bias voltage. Nanomaterials 2023, 13, x FOR PEER REVIEW 8 of 12

Mechanical and Tribological Properties
According to Figure 8a, the COF of the obtained AlMgB14-based coatings varies from 0.12 to 0.24 depending on the sputtering bias voltage. When the bias voltage is increased from 35 V to 50 V, the COF increases slightly by a value close to the measurement error. When the bias voltage is subsequently increased to 100 V, the COF decreases significantly to a minimum value of 0.12. For the bias voltage of 150 V, the COF increases again to the average value of 0.18. Thus, except for the maximum nanohardness of 37 GPa, the AlMgB14-based coating sputtered at a bias voltage of 100 V has the lowest coefficient of friction equal to 0.12 and a relatively low wear rate of 7.5 × 10 −5 mm 3 /N•m compared with the coatings sputtered at 35 V, 50 V, and 150 V bias voltages. Figure 9 shows the coefficient of friction as a function of sliding time for the WC-based substrate and the AlMgB14-based coating sputtered at 100 V bias voltage. According to Figure 9, the COF of the WC-Co substrate starts from 0.15 and increases with the time of the tribological test up to an average value of 0.31. The increase in COF is related to the abrasive destruction of the WC-Co surface. The COF of the AlMgB14-based coating starts from the higher value of 0.18 and increases sharply to 0.2 with the start of the tribological test, which can be defined as the lapping stage caused by the irregularities of the surface microstructure. After the first 50 s, the COF of the AlMgB14-based coating decreased to an average value of 0.12.

Discussion
Regardless of the RF plasma sputtering mode, according to the results of XRD analysis, the AlMgB14-based coatings were formed in the amorphous state. Such an amorphous structure of sputtered coatings is typical for RF plasma sputtering. This can be explained by the highly intensive surface treatment by the ions of the sputtered material. Consequently, a regular crystal structure was not formed. This is in good agreement with the results presented by other authors [15,21,24], which also showed the formation of an amorphous structure of AlMgB14-based coatings, regardless of the deposition mode. However, the Raman spectra show the formation of B12 icosahedral structures in the amorphous Al-Mg-B matrix, which in turn defines the obtained high hardness values of up to 37 GPa.
The morphology of the coatings was found to be highly dependent on the sputtering mode. For example, at low bias voltages (35 V), the coating surface was represented by the angular regions of the sputtered material. As the bias voltage increased, spherical cavities were formed on the surface of the coatings as the surface treatment by the ions of the sputtered material intensified. As a result, as the RF sputtering bias voltage increases from 35 V to 150 V, the average roughness of the coating increases from 22.05 nm to 75.3 nm. According to Figure 9, the COF of the WC-Co substrate starts from 0.15 and increases with the time of the tribological test up to an average value of 0.31. The increase in COF is related to the abrasive destruction of the WC-Co surface. The COF of the AlMgB 14 -based coating starts from the higher value of 0.18 and increases sharply to 0.2 with the start of the tribological test, which can be defined as the lapping stage caused by the irregularities of the surface microstructure. After the first 50 s, the COF of the AlMgB 14 -based coating decreased to an average value of 0.12.

Discussion
Regardless of the RF plasma sputtering mode, according to the results of XRD analysis, the AlMgB 14 -based coatings were formed in the amorphous state. Such an amorphous structure of sputtered coatings is typical for RF plasma sputtering. This can be explained by the highly intensive surface treatment by the ions of the sputtered material. Consequently, a regular crystal structure was not formed. This is in good agreement with the results presented by other authors [15,21,24], which also showed the formation of an amorphous structure of AlMgB 14 -based coatings, regardless of the deposition mode. However, the Raman spectra show the formation of B 12 icosahedral structures in the amorphous Al-Mg-B matrix, which in turn defines the obtained high hardness values of up to 37 GPa.
The morphology of the coatings was found to be highly dependent on the sputtering mode. For example, at low bias voltages (35 V), the coating surface was represented by the angular regions of the sputtered material. As the bias voltage increased, spherical cavities were formed on the surface of the coatings as the surface treatment by the ions of the sputtered material intensified. As a result, as the RF sputtering bias voltage increases from 35 V to 150 V, the average roughness of the coating increases from 22.05 nm to 75.3 nm.
According to the results of tribological studies, the COF depends non-linearly on the bias voltage. Two factors, surface morphology and hardness, can cause COF changes. Thus, it appears that as the bias voltage increases from 35 V to 50 V, the roughness of the coating increases, which also causes the COF to increase but by an insignificant value, comparable to the measurement error. As the bias voltage is further increased to 100 V, the COF decreases sharply to the lowest value of 0.12. The low COF of the AlMgB 14 -based coating could be explained by its high hardness. Because the hardness of the AlMgB 14 -based coating obtained at a bias voltage of 100 V is higher than that of the coatings obtained at other bias voltages and a substrate, it exhibits less abrasive wear during friction. As a result, the COF of the coating was lower and more stable during the test compared to the substrate (Figure 9).The self-lubricating effect of AlMgB 14 described in other papers [8,[24][25][26][27] could also contribute to the reduction in the coating COF. Therefore, the AlMgB 14 -based coating RF sputtered at a bias voltage of 100 V reduced the COF of the WC-Co substrate by more than a factor of 2.5. Meanwhile, as the bias voltage is subsequently increased, the COF increases again due to a decrease in hardness and an increase in the roughness of the coating.
The properties of the AlMgB 14 -based coatings obtained in this study, compared with those of similar materials reported in the literature, are shown in Table 2.  [27] According to Table 2, the RF-sputtered AlMgB 14 -based coatings have a higher maximum hardness than the AlMgB 14 -based materials studied by other authors. At the same time, the tribological properties of the coatings sputtered in the current study are in good agreement with the tribological properties investigated in other studies. According to other studies, the lowest COF values of AlMgB 14 -based coatings were obtained when a lubricating medium was applied. The wear resistance of the coatings was mainly influenced by the counter ball material and the composition of the target material for coating sputtering. For example, according to [25], the wear rate of the AlMgB 14 -30 wt. % Si increases by one order when tested with a Si 3 N 4 -based ball compared to the wear rate when tested with a steel ball. It is evident that this change in wear rate is caused by an increase in the hardness of the counterbody material. In [8], it was shown that a significant increase in the wear resistance of AlMgB 14 -based coatings and materials was achieved by introducing TiB 2 . It has been reported that the wear rate of AlMgB 14 -50 wt. % TiB 2 coating is 6.5 × 10 −7 mm 3 /N·m [8], which is 2 orders lower than the wear rate of AlMgB 14 -based coating without additives obtained in the present work.

Conclusions
In this work, 3 µm-thick AlMgB 14 -based coatings were deposited on WC-Co hard alloy substrates by RF plasma sputtering with varying bias voltages applied to the substrate. According to the XRD studies, sputtered coatings were formed in an amorphous state. However, the Raman spectra showed the formation of nanocrystalline B 12 icosahedrons in the amorphous Al-Mg-B coating structure. The morphology of the coating surface and the mechanical and tribological properties depend on the bias voltage applied to the substrate during sputtering. Therefore, as the bias voltage increases from 35 V to 150 V, the average roughness of the coatings increases from 22.05 nm to 75.3 nm. The nanohardness of the obtained AlMgB 14 -based coatings varied from 24 to 37 GPa, depending on the RF sputtering bias voltage. The AlMgB 14 -based coating sputtered at a 100 V bias voltage had the highest nanohardness of 37 GPa along with a COF of 0.12 and a wear rate of 7.5 × 10 −5 mm 3 /N·m. The hardness and tribological characteristics of sputtered AlMgB 14 -based coatings are in good agreement with the hardness and tribological characteristics of AlMgB 14 -based materials reported in other studies. According to the analysis of the literature, an additional increase in the wear resistance of AlMgB 14 -based coatings could be achieved by introducing up to 50 wt. % TiB 2 in the composition. This provides a perspective for further research aimed at investigating the effect of additives, such as TiB 2 , on the structure and properties of AlMgB 14 -based coatings. In addition, further research could also include the study of AlMgB 14 -based coatings under near-operational conditions-for example, when sputtered on the surface of cutting tools or on friction pairs.

Data Availability Statement:
The data presented in this study are available in the article.