Enhanced Tribological Performance of Low-Friction Nanocomposite WSexSy/NP-W Coatings Prepared by Reactive PLD

A novel laser-based method for producing nanocomposite coatings consisting of a tungsten sulfoselenide (WSexSy) matrix and W nanoparticles (NP-W) was developed. Pulsed laser ablation of WSe2 was carried out in H2S gas under appropriate laser fluence and reactive gas pressure. It was found that moderate sulfur doping (S/Se ~0.2–0.3) leads to significant improvement in the tribological properties of WSexSy/NP-W coatings at room temperature. Changes in the coatings during tribotesting depended on the load on the counter body. The lowest coefficient of friction (~0.02) with a high wear resistance was observed in a N2 environment at an increased load (5 N), resulting from certain structural and chemical changes in the coatings. A tribofilm with a layered atomic packing was observed in the surface layer of the coating. The incorporation of nanoparticles into the coating increased its hardness, which may have influenced the formation of the tribofilm. The initial matrix composition, which had a higher content of chalcogen atoms ((Se + S)/W~2.6–3.5), was altered in the tribofilm to a composition close to the stoichiometric one ((Se + S)/W~1.9). W nanoparticles were ground and retained under the tribofilm, which impacted the effective contact area with the counter body. Changes in the tribotesting conditions—lowering the temperature in a N2 environment—resulted in considerable deterioration of the tribological properties of these coatings. Only coating with a higher S content that was obtained at increased H2S pressure exhibited remarkable wear resistance and a low coefficient of friction, measuring 0.06, even under complicated conditions.


Introduction
Transition Metal Dichalcogenide (TMD) coatings are highly regarded for their exceptional solid-lubricating properties. Research into TMD coatings has been ongoing for several decades, with a growing emphasis on exploring their properties [1][2][3][4][5]. The antifrictional (low-friction) properties of TMD coatings are attributed to their hexagonal crystal lattice structure, which features layered atomic packing. The weak shear resistance between the basal planes is widely considered the primary factor contributing to the solid-lubricating behavior of TMD-based coatings [5][6][7][8][9]. However, this view of the friction mechanism does not fully explain the tribological variability seen in TMD coatings of varying phases, their chemical composition, or architecture. As a result, there is continuing interest in investigating the ways to enhance the wear resistance of TMD coatings in a variety of conditions (such as medium composition and temperature) and to achieve ultralow friction [4,5,[10][11][12][13][14][15].
The most widely utilized method for obtaining TMD-based coatings is ion sputtering, also known as magnetron deposition. This technique allows the composition of deposited coatings to be flexibly adjusted and the tribomechanical characteristics to be improved. This is achieved through forming multilayer or compound coatings with metals (Ti, Ni, Figure 1a-c shows SEM images of WSe x /NP-W and WSe x S y /NP-W coatings deposited on a substrate during 25 min. The coatings have a sufficiently dense structure, which is weakly dependent on the deposition conditions. The surface showed submicron rounded particles consisting of a cluster of smaller spherical nanoparticles. On the surface of the WSe x /NP-W coating, there was only a small number of spherical particles, with diameters of up to 0.5 µm. water vapor condensation and ice layer formation, the friction pair was treated in a N2enriched environment. An optimal N2 flow rate made it possible to keep the plate temperature at about −100 °C and remove the air from the area surrounding the tested sample. The characteristics of the wear tracks on the coatings, the wear scar on the balls, and the wear debris were obtained using an optical profilometer and optical microscopy. After tribotesting, additional studies of the wear tracks by STEM, EDS, and MRS were carried out.

Morphology, Composition, and Microstructure of WSex/NP-W and WSexSy/NP-W Coatings
Figure 1a-c shows SEM images of WSex/NP-W and WSexSy/NP-W coatings deposited on a substrate during 25 min. The coatings have a sufficiently dense structure, which is weakly dependent on the deposition conditions. The surface showed submicron rounded particles consisting of a cluster of smaller spherical nanoparticles. On the surface of the WSex/NP-W coating, there was only a small number of spherical particles, with diameters of up to 0.5 μm. The study of the WSe2 target subjected to pulsed laser ablation revealed the formation of rounded particles enriched in tungsten. The SEM, EDS, and XRD results of the laserirradiated WSe2 target are shown in Figures S1-S3 in the Supplementary Materials. The XRD analysis demonstrated the emergence of two new phases in the target post irradiation, which consisted of cubic α-W and a metastable phase characterized as W3O and β-W. The stabilization of the β-W phase is attributed to the introduction of O (and potentially Se/S) atoms. The analysis of peak intensities suggests that the formation of α-W was predominant on the surface of the target.
The mechanism behind the modification of phase composition in a laser-ablated WSe2 target requires further investigation. This mechanism encompasses the formation of superheated liquid W-Se phase, potentially accompanied by the selective evaporation of selenium and the coagulation of the thin film of W-enriched liquid into spherical particles that then solidify. When the laser impacts a local target area, the particles are likely to be captured by the laser plume and transported to the coating surface. Large particles have low velocities and are unlikely to adhere to the coating. Smaller particles (nanoparticles) exhibit high velocities (~2 × 10 4 cm/s) and are more likely to adhere to the coating surface [44].
XRD studies of the laser-deposited WSex/NP-W coating confirmed the formation of a structure consisting of an X-ray amorphous matrix and a β-W/W3O phase ( Figure 2). A detailed analysis of XRD data for coatings of this type was carried out earlier in [39]. Along with the β-W phase, the formation of the α-W phase was also found. The S atom penetration had no significant effect on the structure of the WSexSy/NP-W coatings; however, the contribution from inclusions of the α-W phase in the XRD patterns for these coatings was noticeably smaller than for the WSex/NP-W coatings. Perhaps, this is because S atoms are The study of the WSe 2 target subjected to pulsed laser ablation revealed the formation of rounded particles enriched in tungsten. The SEM, EDS, and XRD results of the laserirradiated WSe 2 target are shown in Figures S1-S3 in the Supplementary Materials. The XRD analysis demonstrated the emergence of two new phases in the target post irradiation, which consisted of cubic α-W and a metastable phase characterized as W 3 O and β-W. The stabilization of the β-W phase is attributed to the introduction of O (and potentially Se/S) atoms. The analysis of peak intensities suggests that the formation of α-W was predominant on the surface of the target.
The mechanism behind the modification of phase composition in a laser-ablated WSe 2 target requires further investigation. This mechanism encompasses the formation of superheated liquid W-Se phase, potentially accompanied by the selective evaporation of selenium and the coagulation of the thin film of W-enriched liquid into spherical particles that then solidify. When the laser impacts a local target area, the particles are likely to be captured by the laser plume and transported to the coating surface. Large particles have low velocities and are unlikely to adhere to the coating. Smaller particles (nanoparticles) exhibit high velocities (~2 × 10 4 cm/s) and are more likely to adhere to the coating surface [44].
XRD studies of the laser-deposited WSe x /NP-W coating confirmed the formation of a structure consisting of an X-ray amorphous matrix and a β-W/W 3 O phase ( Figure 2). A detailed analysis of XRD data for coatings of this type was carried out earlier in [39]. Along with the β-W phase, the formation of the α-W phase was also found. The S atom penetration had no significant effect on the structure of the WSe x S y /NP-W coatings; however, the contribution from inclusions of the α-W phase in the XRD patterns for these coatings was noticeably smaller than for the WSe x /NP-W coatings. Perhaps, this is because S atoms are introduced into the molten/hot W particles transferred by the laser plume from the irradiated target to the coating and stabilize the β-W phase during their "quenching". introduced into the molten/hot W particles transferred by the laser plume from the irradiated target to the coating and stabilize the β-W phase during their "quenching".
. EDS (SEM) elemental analysis showed that in a relatively large volume of WSex/NP-W coating the x = Se/W ratio was 2.1, which was almost identical to the composition of the pristine target. For the WSexSy/NP-W coating deposited with a H2S pressure of 3.6 Pa, the S/Se ratio was 0.5, and the (Se+S)/W ratio was 2.4. An increase in the H2S pressure to 9 Pa led to a further rise in sulfur content, resulting in a S/Se ratio of 2.3 and a (Se + S)/W ratio of 4.3. According to EDS (SEM) measurements, the oxygen concentration in WSex/NP-W and WSexSy/NP-W coatings was ~3-4 at.%. Considering that the particles in these coatings contain mainly tungsten, the amorphous matrix of these coatings would be significantly enriched in chalcogen atoms. Figure 3a,b shows the results of a STEM study of a whole cross-section for WSexSy/NP-W coating prepared at a 3.6 Pa pressure of reactive H2S gas for 15 min. The coating had a thickness of approximately 900 nm. The high-angle annular dark-field images and elemental maps reveal that the WSexSy/NP-W coating was composed of a Se-and S-enriched matrix with embedded W-enriched nanoparticles. The coating contained regular, spherical particles with a diameter of up to 100 nm, in addition to particles of a deformed ellipsoid shape. The latter particles had a size of less than 50 nm. The density of the nanoparticles was high, while the coating exhibited a relatively dense structure with pores of up to 50 nm. Observations showed that the height of the surface protrusions on the coating did not exceed 100 nm. The analysis of high-resolution TEM (HRTEM) images and selected area electron diffraction (SAED) patterns confirmed the formation of nanoparticles with the characteristic crystal structure of the β-W phase (Figure 3c). The matrix exhibited an amorphous structure in which nanocrystals with layered atom packing were detected. These nanocrystals had a maximum size of 3 nm and a distance of 0.64 nm between the atomic planes. EDS (SEM) elemental analysis showed that in a relatively large volume of WSe x /NP-W coating the x = Se/W ratio was 2.1, which was almost identical to the composition of the pristine target. For the WSe x S y /NP-W coating deposited with a H 2 S pressure of 3.6 Pa, the S/Se ratio was 0.5, and the (Se + S)/W ratio was 2.4. An increase in the H 2 S pressure to 9 Pa led to a further rise in sulfur content, resulting in a S/Se ratio of 2.3 and a (Se + S)/W ratio of 4.3. According to EDS (SEM) measurements, the oxygen concentration in WSe x /NP-W and WSe x S y /NP-W coatings was~3-4 at.%. Considering that the particles in these coatings contain mainly tungsten, the amorphous matrix of these coatings would be significantly enriched in chalcogen atoms. Figure 3a,b shows the results of a STEM study of a whole cross-section for WSe x S y /NP-W coating prepared at a 3.6 Pa pressure of reactive H 2 S gas for 15 min. The coating had a thickness of approximately 900 nm. The high-angle annular dark-field images and elemental maps reveal that the WSe x S y /NP-W coating was composed of a Se-and Senriched matrix with embedded W-enriched nanoparticles. The coating contained regular, spherical particles with a diameter of up to 100 nm, in addition to particles of a deformed ellipsoid shape. The latter particles had a size of less than 50 nm. The density of the nanoparticles was high, while the coating exhibited a relatively dense structure with pores of up to 50 nm. Observations showed that the height of the surface protrusions on the coating did not exceed 100 nm. The analysis of high-resolution TEM (HRTEM) images and selected area electron diffraction (SAED) patterns confirmed the formation of nanoparticles with the characteristic crystal structure of the β-W phase (Figure 3c). The matrix exhibited an amorphous structure in which nanocrystals with layered atom packing were detected. These nanocrystals had a maximum size of 3 nm and a distance of 0.64 nm between the atomic planes.

Nanostructure of WSe x S y /NP-W Coatings
The relatively high density of the WSe x S y /NP-W coating can be attributed to the high surface mobility of the atoms deposited from laser plume. The atoms move into shadow regions created by the deposition of spherical particles, leading to the formation of a relatively dense coating. However, this phenomenon may not occur when the deposited particles are molten and uniformly cover the coating surface. In such cases, the shape of the particle changes, as can be observed in the cross-sectional images of smaller particles. EDS analysis was conducted on different local areas of the matrix, revealing a S/Se ratio of 0.2-0.3 and a (Se + S)/W ratio of approximately 2.6-3.5. The oxygen concentration did not exceed 6 at.%. A cross-sectional analysis of lager areas that included the NP-W showed a (Se + S)/W ratio of approximately 1.7, suggesting that the W atom content in the matrix and the NP-W was approximately equal. However, the EDS analysis of the nanoparticles indicated that they may also contain Se, S, and O atoms-possibly fragments of the matrix enriched with chalcogen atoms. With an estimated specific density of the amorphous matrix being 2-3 times lower than that of tungsten, the volume fraction of nanoparticles in the coating is estimated at 25-30%. The EDS elemental analysis of nanoparticles showed that they could contain Se, S, and O atoms, in addition to tungsten. Nevertheless, these results are not sufficiently reliable, as fragments of the matrix enriched with chalcogen atoms may have entered the measurement region.  The relatively high density of the WSexSу/NP-W coating can be attributed to the high surface mobility of the atoms deposited from laser plume. The atoms move into shadow regions created by the deposition of spherical particles, leading to the formation of a relatively dense coating. However, this phenomenon may not occur when the deposited particles are molten and uniformly cover the coating surface. In such cases, the shape of the particle changes, as can be observed in the cross-sectional images of smaller particles. EDS analysis was conducted on different local areas of the matrix, revealing a S/Se ratio of 0.2-0.3 and a (Se + S)/W ratio of approximately 2.6-3.5. The oxygen concentration did not exceed 6 at.%. A cross-sectional analysis of lager areas that included the NP-W showed a (Se + S)/W ratio of approximately 1.7, suggesting that the W atom content in the matrix and the NP-W was approximately equal. However, the EDS analysis of the nanoparticles indicated that they may also contain Se, S, and O atoms-possibly fragments of the matrix enriched with chalcogen atoms. With an estimated specific density of the amorphous matrix being 2-3 times lower than that of tungsten, the volume fraction of nanoparticles in the coating is estimated at 25-30%. The EDS elemental analysis of nanoparticles showed that they could contain Se, S, and O atoms, in addition to tungsten. Nevertheless, these results are not sufficiently reliable, as fragments of the matrix enriched with chalcogen atoms may have entered the measurement region.  Figure 4 shows MRS spectra measured for WSe x /NP-W and WSe x S y /NP-W coatings. The spectrum of WSe x /NP-W coatings is in many ways like that obtained earlier for WSe x -based coatings deposited by PLD and magnetron techniques [48,49]. All spectra demonstrated a broad feature between 200 and 300 cm −1 , which is characteristic of an amorphous/disordered local structure. The introduction of sulfur reduced the intensity of this broad peak, indicating an increased degree of local packing defectiveness in WSe x S y /NP-W films. The obtained films are in a highly nonequilibrium state, which can be easily altered by mild heating using a high-intensity laser beam. Upon laser heating, a narrower peak at approximately 250 cm −1 emerges (as shown in Figure 4). This peak may comprise two peaks (E 1 2g and A 1g ) from the WSe 2 compound. Nevertheless, peaks related to various Se molecular configurations are also present in this frequency range [50,51]. Thus, during the heat-induced crystallization of the matrix in the WSe x /NP-W coating, surplus selenium may be distributed along the interstitials of the 2H-WSe 2 lattice. Other changes, such as the WSexSy/NP-W films. The obtained films are in a highly nonequilibrium state, which can be easily altered by mild heating using a high-intensity laser beam. Upon laser heating, a narrower peak at approximately 250 cm −1 emerges (as shown in Figure 4). This peak may comprise two peaks (E 1 2g and A1g) from the WSe2 compound. Nevertheless, peaks related to various Se molecular configurations are also present in this frequency range [50,51]. Thus, during the heat-induced crystallization of the matrix in the WSex/NP-W coating, surplus selenium may be distributed along the interstitials of the 2H-WSe2 lattice. Other changes, such as the formation of elemental Se inclusions or sublimation/evaporation of excess selenium, are also possible.

Chemical States of WSex/NP-W and WSexSy/NP-W Coatings
Figure 5a-c shows XPS spectra measured for as-deposited WSex/NP-W and WSexSy/NP-W coatings. The analysis of the W4f spectra reveals the presence of W 0 , W 4+ , and W 6+ species [52][53][54]. The dominance of W 4+ and W 6+ states is due to tungsten's chemical bonds with chalcogen atoms (WSe2/WS2) and oxygen (WO3), respectively. An increase in sulfur concentration results in a shift of the W4f doublet peaks. The peaks for W4f7/2 and W4f5/2 in the WSex/NP-W coating had binding energies of 32.16 and 34.23 eV, respectively. For the WSexSy/NP-W_9 coating, these peaks were located at slightly higher binding energies of 32.47 and 34.52 eV. This energy shift and its increase with growing S concentration is a typical feature observed in metal diselenides upon partial replacement of Se atoms with S atoms [13,35,55]. The negligible contribution of the states of W 0 to the total spectrum indicated that the W nanoparticles were effectively enveloped by the matrix material, even on the surface. It is possible that the W nanoparticles located on the coating surface oxidized, leading to relatively intense peaks in the XPS spectra due to W 6+ . The binding energies of the W4f7/2 and W4f5/2 peaks caused by metal oxide inclusions were almost independent of the coating composition and were found to be approximately 35.7 and 37.8 eV, respectively.

Chemical States of WSe x /NP-W and WSe x S y /NP-W Coatings
Figure 5a-c shows XPS spectra measured for as-deposited WSe x /NP-W and WSe x S y / NP-W coatings. The analysis of the W4f spectra reveals the presence of W 0 , W 4+ , and W 6+ species [52][53][54]. The dominance of W 4+ and W 6+ states is due to tungsten's chemical bonds with chalcogen atoms (WSe 2 /WS 2 ) and oxygen (WO 3 ), respectively. An increase in sulfur concentration results in a shift of the W4f doublet peaks. The peaks for W4f 7/2 and W4f 5/2 in the WSe x /NP-W coating had binding energies of 32.16 and 34.23 eV, respectively. For the WSe x S y /NP-W_9 coating, these peaks were located at slightly higher binding energies of 32.47 and 34.52 eV. This energy shift and its increase with growing S concentration is a typical feature observed in metal diselenides upon partial replacement of Se atoms with S atoms [13,35,55]. The negligible contribution of the states of W 0 to the total spectrum indicated that the W nanoparticles were effectively enveloped by the matrix material, even on the surface. It is possible that the W nanoparticles located on the coating surface oxidized, leading to relatively intense peaks in the XPS spectra due to W 6+ . The binding energies of the W4f 7/2 and W4f 5/2 peaks caused by metal oxide inclusions were almost independent of the coating composition and were found to be approximately 35.7 and 37.8 eV, respectively.
The XPS spectra of the WSe x S y /NP-W coatings reveal the presence of S2p peaks, along with Se3p peaks in the background. The Se3p peaks were analyzed using information obtained from the Se3p spectra of the pure WSe x /NP-W coating. The two peaks at 161.8 and 163.0 eV shown in the XPS spectra of the WSe x S y /NP-W coatings can be assigned to S2p 3/2 and S2p 1/2 binding energies, respectively [13,35]. The binding energies are independent of the S concentration. In addition, two peaks assigned to Se2p 3/2 and Se2p 1/2 were observed at 160.9 and 166.6 eV, respectively.
When analyzing the Se3d region in the XPS spectrum of pure WSe x /NP-W coating, the overstoichiometric composition of the matrix was considered. Two pairs of peaks were necessary to fit each Se peak, corresponding to Se 2− species located at different nodes in the amorphous mesh of the WSe x matrix and to Se 0 species in Se-enriched clusters [35,49,54,56].
The binding energies of the Se3d 5/2 peak for these Se species are 54.4 and 55.3 eV, respectively. The S incorporation in the WSe x S y matrix slightly increases the binding energy for the Se 2− peak and introduces a broad band at higher binding energies (over 57 eV), which could be due to the introduction of O and/or S atoms into the Se clusters. Surface oxidation of the coatings caused by prolonged exposure to air after coating deposition may also influence the chemical state of tungsten and selenium. The best fit of the experimental Se3d spectrum was obtained when the Se3d 5/2 peak was at 57.8 eV, which corresponds to Se 2+ species. Se-enriched clusters may form due to excess Se atoms segregating. During the formation of the WSe x S y /NP-W coatings by reactive PLD, Se and S atoms were embedded into the coating matrix through different mechanisms. Se atoms were introduced through physical vapor-phase deposition, while S atoms were probably introduced through chemical interaction with surface coating atoms. High chalcogen concentrations in the WSe x S y matrix may result in Se phase segregation, as observed in previous studies of PLD processes in Se-based TMD films [48,56]. The XPS spectra of the WSexSy/NP-W coatings reveal the presence of S2p peaks, along with Se3p peaks in the background. The Se3p peaks were analyzed using information obtained from the Se3p spectra of the pure WSex/NP-W coating. The two peaks at 161.8 and 163.0 eV shown in the XPS spectra of the WSexSy/NP-W coatings can be assigned to S2p3/2 and S2p1/2 binding energies, respectively [13,35]. The binding energies are independent of the S concentration. In addition, two peaks assigned to Se2p3/2 and Se2p1/2 were observed at 160.9 and 166.6 eV, respectively.
When analyzing the Se3d region in the XPS spectrum of pure WSex/NP-W coating, the overstoichiometric composition of the matrix was considered. Two pairs of peaks were necessary to fit each Se peak, corresponding to Se 2− species located at different nodes in the amorphous mesh of the WSex matrix and to Se 0 species in Se-enriched clusters [35,49,54,56].
The binding energies of the Se3d5/2 peak for these Se species are 54.4 and 55.3 eV,

Frictional Behavior of WSe x /NP-W and WSe x S y /NP-W Coatings
Figure 6a-c shows the low-friction properties of WSe x /NP-W and WSe x S y /NP-W coatings at room temperature in environments consisting of moist air and a N 2 -enriched environment under loads of 1 and 5 N. The coating deposition time was 25 min. For the WSe x /NP-W coating, measurements in the air under a load of 1 N exhibited a behavior characterized by an initially higher coefficient of friction (CoF~0.09), which transitioned (run-in) to a lower steady-state CoF~0.07 with increased sliding cycles. This run-in behavior changed when this coating was doped with sulfur. For the WSe x S y /NP-W_3.6 coating, the initial CoF value was 0.05. During the run-in phase, the CoF increased to 0.06 and then slowly increased to 0.07 by the end of the test. For the WSe x S y /NP-W_9 coating, the CoF increased relatively quickly from an initial value of 0.06 to 0.12 when tested for 1000 cycles. The testing of this coating was then stopped.

Frictional Behavior of WSex/NP-W and WSexSy/NP-W Coatings
Figure 6a-c shows the low-friction properties of WSex/NP-W and WSexSy/NP-W coatings at room temperature in environments consisting of moist air and a N2-enriched environment under loads of 1 and 5 N. The coating deposition time was 25 min. For the WSex/NP-W coating, measurements in the air under a load of 1 N exhibited a behavior characterized by an initially higher coefficient of friction (CoF~0.09), which transitioned (run-in) to a lower steady-state CoF~0.07 with increased sliding cycles. This run-in behavior changed when this coating was doped with sulfur. For the WSexSy/NP-W_3.6 coating, the initial CoF value was 0.05. During the run-in phase, the CoF increased to 0.06 and then slowly increased to 0.07 by the end of the test. For the WSexSy/NP-W_9 coating, the CoF increased relatively quickly from an initial value of 0.06 to 0.12 when tested for 1000 cycles. The testing of this coating was then stopped. The tribotests in a N2-enriched environment with loads of 1 N showed only a slight improvement in the friction properties of the WSex/NP-W coating, particularly during the running-in phase, where no increased coefficient of friction was observed. Throughout the tribotesting, the steady-state CoF was ~0.05-0.06. In contrast, the addition of sulfur considerably improved the low-friction properties of the WSexSy/NP-W_3.6 coating in this environment, resulting in a steady-state CoF of ~0.045. Changing the test environment did not improve the tribological properties of the WSexSy/NP-W_9 coating.
Increasing the load on the counter body had a different effect on the tribological behavior of the WSex/NP-W and WSexSy/NP-W coatings when tested in a N2-enriched environment. For WSex/NP-W, the CoF value decreased to 0.03-0.04, but a sharp increase in CoF was observed at a certain stage of the tribotest. The durability of the coating did not exceed ~1500-2000 sliding cycles. This behavior was repeated in three tribotests. The best tribological properties were found for WSexSy/NP-W_3.6, with a CoF value of 0.02-0.025 throughout the entire tribotest period. Increasing the load caused some improvement in the tribological properties of the WSexSy/NP-W_9 coating. During a sufficiently long sliding period (up to 2500 cycles), the CoF value did not exceed 0.06.
The results of the tribotesting of the WSexSy/NP-W_9 coating indicate that coatings with a high concentration of chalcogen atoms had unsatisfactory tribological properties ( Figure 6). The tribotests in a N 2 -enriched environment with loads of 1 N showed only a slight improvement in the friction properties of the WSe x /NP-W coating, particularly during the running-in phase, where no increased coefficient of friction was observed. Throughout the tribotesting, the steady-state CoF was~0.05-0.06. In contrast, the addition of sulfur considerably improved the low-friction properties of the WSe x S y /NP-W_3.6 coating in this environment, resulting in a steady-state CoF of~0.045. Changing the test environment did not improve the tribological properties of the WSe x S y /NP-W_9 coating.
Increasing the load on the counter body had a different effect on the tribological behavior of the WSe x /NP-W and WSe x S y /NP-W coatings when tested in a N 2 -enriched environment. For WSe x /NP-W, the CoF value decreased to 0.03-0.04, but a sharp increase in CoF was observed at a certain stage of the tribotest. The durability of the coating did not exceed~1500-2000 sliding cycles. This behavior was repeated in three tribotests. The best tribological properties were found for WSe x S y /NP-W_3.6, with a CoF value of 0.02-0.025 throughout the entire tribotest period. Increasing the load caused some improvement in the tribological properties of the WSe x S y /NP-W_9 coating. During a sufficiently long sliding period (up to 2500 cycles), the CoF value did not exceed 0.06.
The results of the tribotesting of the WSe x S y /NP-W_9 coating indicate that coatings with a high concentration of chalcogen atoms had unsatisfactory tribological properties ( Figure 6). However, under specific conditions of low temperature and in a N 2 -enrichment environment, this coating showed the best tribological behavior (Figure 7). In these conditions, weak water vapor condensation and the formation of a thin ice film were possible. The WSe x S y /NP-W_9 coating had an initial CoF of 0.04 and reached a steady-state CoF~0.06 after 500 cycles. The CoF of the other two coatings increased rapidly to values above 0.08 within 500 to 1000 cycles.
The introduction of sulfur into the WSe x /NP-W coating caused a decrease in hardness and elasticity. The hardness of the WSe x /NP-W coating was approximately 9.2 GPa, i.e., higher than the 2-3 GPa hardness of magnetron-deposited WSe x coating (e.g., [32,49]). Young's modulus of the WSe x /NP-W coating was found to be between 110-120 GPa, which is significantly higher than that of magnetron-deposited WSe x coatings, where Young's modulus typically does not exceed 60 GPa [32,49]. The hardness of the WSe x S y /NP-W_3. 6 and WSe x S y /NP-W_9 coatings was reduced to 7.5 and 4.3 GPa, respectively, due to the matrix modification. Young's modulus of the WSe x S y /NP-W_3. 6 and WSe x S y /NP-W_9 coatings was 107 and 67 GPa, respectively. Thus, the addition of S to the matrix of nanocom-posite WSe x S y /NP-W coatings caused a decrease in their mechanical strength. However, the W nanoparticles present in the bulk prevented excessive softening of the coatings. However, under specific conditions of low temperature and in a N2-enrichment environment, this coating showed the best tribological behavior (Figure 7). In these conditions, weak water vapor condensation and the formation of a thin ice film were possible. The WSexSy/NP-W_9 coating had an initial CoF of 0.04 and reached a steady-state CoF~0.06 after 500 cycles. The CoF of the other two coatings increased rapidly to values above 0.08 within 500 to 1000 cycles. The introduction of sulfur into the WSex/NP-W coating caused a decrease in hardness and elasticity. The hardness of the WSex/NP-W coating was approximately 9.2 GPa, i.e., higher than the 2-3 GPa hardness of magnetron-deposited WSex coating (e.g., [32,49]). Young's modulus of the WSex/NP-W coating was found to be between 110-120 GPa, which is significantly higher than that of magnetron-deposited WSex coatings, where Young's modulus typically does not exceed 60 GPa [32,49]. The hardness of the WSexSy/NP-W_3. 6 and WSexSy/NP-W_9 coatings was reduced to 7.5 and 4.3 GPa, respectively, due to the matrix modification. Young's modulus of the WSexSy/NP-W_3.6 and WSexSy/NP-W_9 coatings was 107 and 67 GPa, respectively. Thus, the addition of S to the matrix of nanocomposite WSexSy/NP-W coatings caused a decrease in their mechanical strength. However, the W nanoparticles present in the bulk prevented excessive softening of the coatings.

Wear Behavior of WSex/NP-W and WSexSy/NP-W Coatings
Studies into the tribological properties of WSex/NP-W and WSexSy/NP-W coatings have revealed that a comparison of the wear behavior of WSex/NP-W and WSexSy/NP-W_3.6 coatings is of particular interest under conventional test conditions (as regards ambient temperature and environment). The wear investigation focused on the central area of the track, as there was no significant correlation between wear and ball sliding speed in different areas of the track. Under specific test conditions, such as low temperatures and an environment containing water vapor, the sliding properties of these coatings were considerably impaired compared with the WSexSy/NP-W_9 coating. Additional investigations are needed to determine the underlying reasons for this behavior. The Supplementary Materials ( Figure S4) provide information solely on the ball scars and wear tracks that resulted from severe stress testing of these coatings, with a comparison with the wear tracks and scars of the balls for other prepared coatings. Figure 8 shows the wear of the WSex/NP-W and WSexSy/NP-W_3.6 coatings and their corresponding counter bodies after tribotests in air. The coatings were deposited for 25 min. Notably, a correlation between the frictional and wear behavior of these coatings can be observed. The WSexSy/NP-W_3.6 coatings and counter bodies experienced slightly higher wear than their WSexSy/NP-W counterparts, resulting in a wider and deeper wear track and increased counter-body wear scar size. The depth profile of the wear track on the WSexSy/NP-W_3.6 coating differed slightly from that of the WSex/NP-W coating, with

Wear Behavior of WSe x /NP-W and WSe x S y /NP-W Coatings
Studies into the tribological properties of WSe x /NP-W and WSe x S y /NP-W coatings have revealed that a comparison of the wear behavior of WSe x /NP-W and WSe x S y /NP-W_3.6 coatings is of particular interest under conventional test conditions (as regards ambient temperature and environment). The wear investigation focused on the central area of the track, as there was no significant correlation between wear and ball sliding speed in different areas of the track. Under specific test conditions, such as low temperatures and an environment containing water vapor, the sliding properties of these coatings were considerably impaired compared with the WSe x S y /NP-W_9 coating. Additional investigations are needed to determine the underlying reasons for this behavior. The Supplementary Materials ( Figure S4) provide information solely on the ball scars and wear tracks that resulted from severe stress testing of these coatings, with a comparison with the wear tracks and scars of the balls for other prepared coatings. Figure 8 shows the wear of the WSe x /NP-W and WSe x S y /NP-W_3.6 coatings and their corresponding counter bodies after tribotests in air. The coatings were deposited for 25 min. Notably, a correlation between the frictional and wear behavior of these coatings can be observed. The WSe x S y /NP-W_3.6 coatings and counter bodies experienced slightly higher wear than their WSe x S y /NP-W counterparts, resulting in a wider and deeper wear track and increased counter-body wear scar size. The depth profile of the wear track on the WSe x S y /NP-W_3.6 coating differed slightly from that of the WSe x /NP-W coating, with the former exhibiting large areas where the coating was completely removed from the substrate. The sharp edges of these areas suggest that the WSe x S y /NP-W_3.6 coating may have peeled off in relatively large fragments, whereas track deepening due to wear and tear on the coating was less prominent.
The weaker adhesion of the WSe x S y /NP-W_3.6 coating to the substrate compared with the WSe x /NP-W coating could be attributed to the decreased energy of deposited atoms and ions resulting from deposition in a reaction gas. This decrease in energy reduces the efficiency of chemical bond formation at the coating-substrate interface, and in turn, the efficiency of surface sputtering [41]. As a result of the suppression of self-sputtering and S atom deposition, the thickness of the coating formed by RPLD increased, with the WSe x /NP-W coating measuring 800 nm and the WSe x S y /NP-W_3.6 coating 1200 nm. the former exhibiting large areas where the coating was completely removed from the substrate. The sharp edges of these areas suggest that the WSexSy/NP-W_3.6 coating may have peeled off in relatively large fragments, whereas track deepening due to wear and tear on the coating was less prominent. The weaker adhesion of the WSexSy/NP-W_3.6 coating to the substrate compared with the WSex/NP-W coating could be attributed to the decreased energy of deposited atoms and ions resulting from deposition in a reaction gas. This decrease in energy reduces the efficiency of chemical bond formation at the coating-substrate interface, and in turn, the efficiency of surface sputtering [41]. As a result of the suppression of self-sputtering and S atom deposition, the thickness of the coating formed by RPLD increased, with the WSex/NP-W coating measuring 800 nm and the WSexSy/NP-W_3.6 coating 1200 nm. Figure 9 shows the wear of the WSex/NP-W and WSexSy/NP-W_3.6 coatings and counter bodies after the tribotests in a N2-enriched environment under a load of 1 N. The WSexSy/NP-W_3.6 coating showed clearly improved wear resistance compared with the WSex/NP-W coating. At the same time, the wear of the counter body was significantly reduced. The addition of S caused a reduction in track width from 150 µ m to 80 µ m. The maximum track depth reduced from 240 to 30 nm by a factor of 8. A comparison of the wear of these coatings in wet air and nitrogen showed that in N2 environment the wear decreased for both coatings. This behavior is typical of all TMD-based coatings. However, for WSexSy/NP-W_3.6 coating, this reduction was the most significant.  Figure 9 shows the wear of the WSe x /NP-W and WSe x S y /NP-W_3.6 coatings and counter bodies after the tribotests in a N 2 -enriched environment under a load of 1 N. The WSe x S y /NP-W_3.6 coating showed clearly improved wear resistance compared with the WSe x /NP-W coating. At the same time, the wear of the counter body was significantly reduced. The addition of S caused a reduction in track width from 150 µm to 80 µm. The maximum track depth reduced from 240 to 30 nm by a factor of 8. A comparison of the wear of these coatings in wet air and nitrogen showed that in N 2 environment the wear decreased for both coatings. This behavior is typical of all TMD-based coatings. However, for WSe x S y /NP-W_3.6 coating, this reduction was the most significant. During tribotesting of the WSex/NP-W coating in N2-enriched environment with an increased counter body load of 5 N (Hertzian contact stress ~0.86 GPa), a marked increase in wear was observed. Figure 10 illustrates the wear pattern of the coating under these conditions, where the track depth on the coating reached nearly 500 nm after 1500 sliding During tribotesting of the WSe x /NP-W coating in N 2 -enriched environment with an increased counter body load of 5 N (Hertzian contact stress~0.86 GPa), a marked increase in wear was observed. Figure 10 illustrates the wear pattern of the coating under these conditions, where the track depth on the coating reached nearly 500 nm after 1500 sliding cycles. The tribotest ceased upon a rapid escalation of the friction coefficient. The analysis of the wear track suggested that the cause of this phenomenon may be attributed to the localized deepening of the track down to the substrate. While the increase in load accelerated the wear of the WSe x S y /NP-W_3.6 coating, the increment was moderate. The average depth of the track rose to 40 nm, with some regions of the track exhibiting a wear depth increase to 100 nm. The width of the well also increased from 80 nm to 140 nm, with a concomitant growth in the wear scar of the ball. Notably, the exposure of the counter body to a high load resulted in the delamination of the coatings from the substrate in some localized areas of the track, as evidenced by the formation of visible swellings in the microphotographs of the tracks (Figures 9 and 10). Wear analysis of the WSe x S y /NP-W_3.6 coating through the integration of the volume of material removed from the track demonstrated that the wear rate of this coating in a N 2 -enriched environment was independent of counter-body load, being registered at 1.6 × 10 −7 mm 3 N −1 m −1 .
(c) (d) During tribotesting of the WSex/NP-W coating in N2-enriched environment with an increased counter body load of 5 N (Hertzian contact stress ~0.86 GPa), a marked increase in wear was observed. Figure 10 illustrates the wear pattern of the coating under these conditions, where the track depth on the coating reached nearly 500 nm after 1500 sliding cycles. The tribotest ceased upon a rapid escalation of the friction coefficient. The analysis of the wear track suggested that the cause of this phenomenon may be attributed to the localized deepening of the track down to the substrate. While the increase in load accelerated the wear of the WSexSy/NP-W_3.6 coating, the increment was moderate. The average depth of the track rose to 40 nm, with some regions of the track exhibiting a wear depth increase to 100 nm. The width of the well also increased from 80 nm to 140 nm, with a concomitant growth in the wear scar of the ball. Notably, the exposure of the counter body to a high load resulted in the delamination of the coatings from the substrate in some localized areas of the track, as evidenced by the formation of visible swellings in the microphotographs of the tracks (Figures 9 and 10). Wear analysis of the WSexSy/NP-W_3.6 coating through the integration of the volume of material removed from the track demonstrated that the wear rate of this coating in a N2-enriched environment was independent of counter-body load, being registered at 1.6 × 10 −7 mm 3 N −1 m −1 .  When tested in a N 2 -enriched environment at a very low temperature, the WSe x S y /NP-W_9 coating with high sulfur content showed the least wear. After 3000 sliding cycles, a smooth track with minimal wear debris formed on the surface of the coating. Relatively low wear and a weak adherence of wear debris were also observed on the counter body. In the other coatings, the analysis of the wear tracks and the counter bodies showed that the rapid increase in the coefficient of friction, as seen in Figure 7, was due to the significant wear of the coatings and the counter bodies ( Figure S4).

Discussion
The WSe x S y /NP-W_3.6 coating has tribological properties comparable to those of the WSe x /NP-W coating in moist air; it outperforms the WSe x /NP-W coating when tested in a N 2 -enriched environment at room temperature. Therefore, a detailed investigation of tribo-induced modification of the WSe x S y /NP-W_3.6 coating was carried out. Figure 11a,b shows Raman spectra for the wear tracks and wear debris formed after tribotesting the WSe x /NP-W and WSe x S y /NP-W_3.6 coatings. Under both test conditions, the coatings exhibited similar wear patterns, as indicated by the appearance of a relatively narrow peak near 250 cm −1 , which is characteristic of the atomic packing of 2H-WSe 2 . However, the main difference in the Raman spectra measured in the wear tracks for the WSe x /NP-W and WSe x S y /NP-W_3.6 coatings was the shape of the peaks. In the spectrum of the WSe x /NP-W coating, the maximum peak was in the range of 235-245 cm −1 . This shift from 250 cm −1 towards lower frequencies may be attributed to the influence of excess selenium, which can cause peaks at approximately 235 cm −1 in some nanoforms [57]. In contrast, the spectrum of the WSe x S y /NP-W_3.6 coating showed a peak maximum at 252 cm −1 and an asymmetrical peak shape. The shift towards higher frequencies was likely caused by the introduction of S atoms into the WSe 2 lattice [58]. The asymmetry of the peak shape may be due to differences in the texture of the modified layers in these coatings, as the texture can affect the ratio of intensities of A 1g and E 1 2g peaks, which merge into one broadened peak at a high defectiveness of atomic packing. Additionally, the formation of a structure containing WSe 2 and W(Se/S) 2 nanoclusters cannot be ruled out.

Discussion
The WSexSy/NP-W_3.6 coating has tribological properties comparable to those of the WSex/NP-W coating in moist air; it outperforms the WSex/NP-W coating when tested in a N2-enriched environment at room temperature. Therefore, a detailed investigation of triboinduced modification of the WSexSy/NP-W_3.6 coating was carried out. Figure 11a,b shows Raman spectra for the wear tracks and wear debris formed after tribotesting the WSex/NP-W and WSexSy/NP-W_3.6 coatings. Under both test conditions, the coatings exhibited similar wear patterns, as indicated by the appearance of a relatively narrow peak near 250 cm −1 , which is characteristic of the atomic packing of 2H-WSe2. However, the main difference in the Raman spectra measured in the wear tracks for the WSex/NP-W and WSexSy/NP-W_3.6 coatings was the shape of the peaks. In the spectrum of the WSex/NP-W coating, the maximum peak was in the range of 235-245 cm −1 . This shift from 250 cm −1 towards lower frequencies may be attributed to the influence of excess selenium, which can cause peaks at approximately 235 cm −1 in some nanoforms [57]. In contrast, the spectrum of the WSexSy/NP-W_3.6 coating showed a peak maximum at 252 cm −1 and an asymmetrical peak shape. The shift towards higher frequencies was likely caused by the introduction of S atoms into the WSe2 lattice [58]. The asymmetry of the peak shape may be due to differences in the texture of the modified layers in these coatings, as the texture can affect the ratio of intensities of A1g and E 1 2g peaks, which merge into one broadened peak at a high defectiveness of atomic packing. Additionally, the formation of a structure containing WSe2 and W(Se/S)2 nanoclusters cannot be ruled out. Raman studies of wear debris showed that, in addition to the S/Se-containing phase, there was a metal oxide phase as well (Figure 11a,b). The formation of metal oxide inclusions caused the appearance of a wide band from 700 to 1000 cm −1 . The broad band extending from 600 to 800 cm −1 is situated in the region of W-O-W stretching bonds of WO3 [49,59]. The presence of shoulders extending over 700 and 900 cm −1 indicated a high degree of WO3-phase defectiveness and points to the possible formation of Fe-containing oxides Raman studies of wear debris showed that, in addition to the S/Se-containing phase, there was a metal oxide phase as well (Figure 11a,b). The formation of metal oxide inclusions caused the appearance of a wide band from 700 to 1000 cm −1 . The broad band extending from 600 to 800 cm −1 is situated in the region of W-O-W stretching bonds of WO 3 [49,59]. The presence of shoulders extending over 700 and 900 cm −1 indicated a high degree of WO 3 -phase defectiveness and points to the possible formation of Fe-containing oxides (FeO, Fe 3 O 4 , and FeWO 4 ) [60]. It is important to note that the N 2 -enriched environment reduced the influence of oxygen and water vapor on tribo-chemical processes in the friction zone but did not eliminate it. This phenomenon is more pronounced for the S-containing WSe x S y /NP-W coating as compared with pure WSe x /NP-W coating.
Results of the STEM analysis of the WSe x S y /NP-W coating in the central region of the wear track formed after 500 sliding cycles in a humid air environment are presented in Figure 12a-e. The application of a 1 N load during sliding smoothed the roughness of the coating surface, while its porosity was retained. HRTEM and SAED studies revealed significant structural changes in the matrix throughout the coating, resulting in the formation of nanocrystals with a layered WSe 2 -like structure. The dimensions of the ordered areas with layered packing of atoms reached up to 20 nm, and their orientation was mainly influenced by the nearest W nanoparticle. The predominant orientation of the basic planes was along the surface of the nanoparticle, potentially due to the movement or vibration of the nanoparticle within the amorphous matrix. The interplanar spacing varied between 0.65 to 0.7 nm. No evidence of tribofilm formation with WSe 2 nanocrystals oriented along the surface was found, nor was any change in the Se/S and (Se + S)/W ratios in the near-surface layer of the matrix detected. Exposure to moist air resulted in an increase in the O atoms concentration in the matrix surface layer to 11.7 at.%, with O atoms penetrating to a depth of 30 nm.
coating surface, while its porosity was retained. HRTEM and SAED studies revealed significant structural changes in the matrix throughout the coating, resulting in the formation of nanocrystals with a layered WSe2-like structure. The dimensions of the ordered areas with layered packing of atoms reached up to 20 nm, and their orientation was mainly influenced by the nearest W nanoparticle. The predominant orientation of the basic planes was along the surface of the nanoparticle, potentially due to the movement or vibration of the nanoparticle within the amorphous matrix. The interplanar spacing varied between 0.65 to 0.7 nm. No evidence of tribofilm formation with WSe2 nanocrystals oriented along the surface was found, nor was any change in the Se/S and (Se+S)/W ratios in the nearsurface layer of the matrix detected. Exposure to moist air resulted in an increase in the O atoms concentration in the matrix surface layer to 11.7 at.%, with O atoms penetrating to a depth of 30 nm. Increasing the counter-body load to 5 N resulted in noticeable compaction of the WSe x S y /NP-W_3.6 coating and the effective formation of a W(Se/S) 2 -based tribofilm on the surface. Figure 13a-d presents HRTEM images and SAED patterns of the cross-section of a tribotested film in a N 2 -enriched environment after 1000 sliding cycles. The thickness of the tribofilm possessing ordered W(Se/S) 2 fringes reached 15 nm, and the basic planes had a preferential orientation parallel to the coating surface. The distance between the basic planes was approximately 6.7 nm. At the boundary between the film and the matrix, the atomic planes exhibited arbitrary orientations. The thickness of such a transition layer did not exceed 10 nm. In deeper layers, the amorphous matrix structure was mostly preserved. The tribo-induced modification of the surface-layer structure involved the partial removal of S and Se atoms, resulting in a decrease in the (Se + S)/W ratio to~1.9. However, the O atoms' concentration did not change significantly in the surface layer after tribotesting. preferential orientation parallel to the coating surface. The distance between the basic planes was approximately 6.7 nm. At the boundary between the film and the matrix, the atomic planes exhibited arbitrary orientations. The thickness of such a transition layer did not exceed 10 nm. In deeper layers, the amorphous matrix structure was mostly preserved. The tribo-induced modification of the surface-layer structure involved the partial removal of S and Se atoms, resulting in a decrease in the (Se + S)/W ratio to ~1.9. However, the O atoms' concentration did not change significantly in the surface layer after tribotesting. Raising the load to 5 N did not lead to an effective crystallization of the matrix in the coating volume, as was observed in the lower load tests. This lack of structural transformation could be due to the strong compaction of the coating induced by a higher load, which may have inhibited the reduction in local atomic packing density required for the transformation of an amorphous structure into a layered one. However, the transformation of an amorphous matrix with an excessive content of chalcogen atoms ((Se + S)/W ~2.6-3.8) into crystalline 2H-W(Se/S)2 can cause the intercalation effects that increase the interplanar distance. This is evident from the increase in the base-plane spacing to 0.7 nm, as shown in Figure 12. Despite this, the formation of nanoclusters of these components cannot be excluded, although they were not detected in the original coating structure during XPS studies. Assuming that the tribo-induced structural changes in the WSexSy/NP-W_3.3 coating in a N2-enriched environment at a load of 1N occurred in a similar way, the matrix crystallization throughout the coating can provide sufficient low-friction properties, even at a high concentration of chalcogen atoms-(Se + S)/W ~2.6-3.8. However, the results of tests under a higher load show that the densification of the structure and the formation of a tribofilm with a composition close to stoichiometric by the (Se + S)/W Raising the load to 5 N did not lead to an effective crystallization of the matrix in the coating volume, as was observed in the lower load tests. This lack of structural transformation could be due to the strong compaction of the coating induced by a higher load, which may have inhibited the reduction in local atomic packing density required for the transformation of an amorphous structure into a layered one. However, the transformation of an amorphous matrix with an excessive content of chalcogen atoms ((Se + S)/W~2.6-3.8) into crystalline 2H-W(Se/S) 2 can cause the intercalation effects that increase the interplanar distance. This is evident from the increase in the base-plane spacing to 0.7 nm, as shown in Figure 12. Despite this, the formation of nanoclusters of these components cannot be excluded, although they were not detected in the original coating structure during XPS studies. Assuming that the tribo-induced structural changes in the WSe x S y /NP-W_3.3 coating in a N 2 -enriched environment at a load of 1N occurred in a similar way, the matrix crystallization throughout the coating can provide sufficient low-friction properties, even at a high concentration of chalcogen atoms-(Se + S)/W~2.6-3.8. However, the results of tests under a higher load show that the densification of the structure and the formation of a tribofilm with a composition close to stoichiometric by the (Se + S)/W parameter are more important. The mechanisms responsible for removing excess S and Se atoms require further investigation. These atoms may be removed through the formation of volatile compounds with oxygen and hydrogen present in the applied media, such as wet air and a N 2 -enriched environment with an RH~7 %.
The analysis of the behavior of W nanoparticles in the WSe x S y /NP-W_3.3 coating revealed that these nanoparticles did not migrate into the coating and were removed during wear with the matrix material, under both high and low loads. This was confirmed by the consistent (Se + S)/W~1.7 values measured by EDS in the cross-section of the coating. Yet, under lower loads, the nanoparticles were found to protrude above the coating surface, resulting in increased friction ( Figure 12). Under higher loads, the nanoparticles were found underneath the tribofilm. Figure 13 shows that the rounded particles on the surface were ground because of contact with the counter body. The presence of a W(Se/S) 2 tribofilm at the interface of the W nanoparticles and the Fe-based ball may induce tribo-chemical reactions, such as sulfurization and/or selenization of the nanoparticles and the steel ball, leading to phase changes and a reduction in friction between initially metallic surfaces.
A comparison of the tribological properties of WSe x /NP-W and WSe x S y /NP-W coatings revealed that moderate S doping does not significantly degrade the properties of these coatings in wet air but considerably enhances the performance in a N 2 -enriched environment at room temperature. At moderate doping levels characterized by (Se + S)/W < 3.5, sulfur atoms were dispersed throughout the matrix volume, potentially taking different states within the amorphous matrix packing. As a result, the hardness of the coating decreased, albeit the process being mitigated by the presence of W nanoparticles, promoting plastic deformation and densification of the WSe x S y /NP-W_3.6 coating under load. The introduction of S atoms appeared to alter the local atomic packing and may have contributed to the transformation of the amorphous structure into a crystalline/layered structure under tribo-impact. The partial substitution of Se atoms by S atoms in the layered structure is believed to have reduced the shear stresses between the basal planes in the 2H-W(Se/S) 2 lattice. Notably, Hu et al. [35] have shown that TMD coatings containing a mixture of S and Se atoms exhibit superior tribological performance attributed to increased crystallinity and larger basal plane separations in these composites. To ensure the effective performance of WSe x S y /NP-W coatings under extremely challenging environmental conditions, such as very low temperatures and low humidity, the S content may need to be significantly increased, resulting in a much softer matrix. W nanoparticles formed the hard "skeleton" of the WSe x /NP-W and WSe x S y /NP-W nanocomposite coatings. At relatively low counter-body loads, the nanoparticles in the coating could be removed from the surface without fracturing. Under higher loads, the particles underwent wear. However, the formation of a W(Se/S) 2 tribofilm between the steel counter body and the W nanoparticle prevented direct contact between the metals, leading to a reduced friction coefficient. Notably, the contact between the counter body and the nanocomposite coating may have been localized primarily on the W nanoparticles, thus reducing the actual contact area and ultimately contributing to a lower friction coefficient.
A review of the literature on the tribological properties of WSe x coatings synthesized by means other than PLD techniques reveals that the most comparable tribotesting conditions were achieved by Dominguez-Meister et al. [8,49,61], who employed nonreactive magnetron sputtering for WSe x coating deposition. Reciprocating tests were conducted at 2 N of applied load (a maximum contact pressure of~0.83 GPa), a stroke length of 2 mm, and a 2 mm s −1 of linear speed during 2500 cycles. The tribological behavior of pure WSe x remains almost unaltered, showing a similar friction coefficient (0.04-0.07) and wear rates (1.5-3 × 10 −7 mm 3 N −1 m −1 ), independent of the nature of the environment (in ambient air and in dry nitrogen). Mixing WSe x with carbon or MoS 2 did not enhance the low-friction properties [46,62]. However, nanocomposite WSe x S y /NP-W_3.6 coatings fabricated in this study by reactive PLD showed improved low-friction properties with a minimum friction coefficient of~0.02, compared with the already known magnetron-deposited WSe x coatings. The wear resistance of the WSe x S y /NP-W_3.6 coatings was not inferior to that of magnetron-deposited WSe x coatings. The deposition of WSe x coatings by magnetron sputtering has been extensively explored and optimized. However, the tribological performance of nanocomposite WSe x S y /NP-W coatings may be further improved by optimizing reactive PLD parameters.

Conclusions
This study has shown that subjecting the WSe 2 target to pulsed laser ablation under vacuum conditions produces a flux of W and Se atoms, along with a flux of W nanoparticles. The maximum size of the nanoparticles can reach several tens of nanometers, although the structure of the coating is mostly made up of round-shaped particles smaller than 20 nm.
The volume fraction of nanoparticles in the coatings can reach 25-30%. In the presence of H 2 S gas during laser ablation of WSe 2 target, S atoms are introduced into the amorphous matrix of the coatings along with W and Se atoms, and their concentration increases with increasing H 2 S pressure. The incorporation of NP-W leads to an increased hardness of the nanocomposite coatings and may impact the contact area between the counter body and the coating.
Under moderate S atom concentration, WSe x S y /NP-W_3.6 coatings exhibit high-level tribological properties when tested in ambient air and N 2 -enriched environments at room temperature. The nature of tribo-induced changes in these coatings is contingent upon the load on the counter body. At higher counter-body loads, the coatings exhibited the lowest coefficient of friction (0.02) and the highest wear resistance (~1.6×10 −7 mm 3 N −1 m −1 ) when sliding in nitrogen. During sliding, an up to 20 nm-thick tribofilm forms on the coating surface containing W(Se/S) 2 nanocrystals, whose basal planes align parallel to the coating surface. Importantly, the tribofilm effectively reduces the concentration of Se and S atoms to a value near stoichiometric ((Se + S)/W~1.9) and prevents W nanoparticles from making clean contact with the steel counter body. The tribological performance of these coatings, however, noticeably declines under more challenging conditions, such as low temperature (−100 • C) and N 2 -enriched environments with a relative humidity of~7%.
Notably, the excessive S content in WSe x S y /NP-W_9 coatings has an adverse effect on their tribological properties under "conventional" ambient conditions. However, these coatings demonstrate the best tribological performance when tested at very low temperatures (about −100 • C), with a low wear rate and a coefficient of friction of 0.06 over 3000 sliding cycles. Other coatings experience a sharp increase in the coefficient of friction to 0.1 after just 1000 sliding cycles.
Supplementary Materials: The following supporting information can be downloaded at https:// www.mdpi.com/article/10.3390/nano13061122/s1, Figure S1: SEM images of (a) as-prepared and (b,c) laser-ablated WSe 2 target with two magnifications; the area for EDS analysis is indicated in a red circle in (c); Figure S2: Distribution of elements in a local area of the WSe 2 target subjected to laser ablation. The area is indicated in Figure S1c, and it contains spherical particles which were formed after laser ablation; Figure S3: XRD pattern of laser-ablated WSe 2 target exhibits the content of (a) WSe 2 (P6 3 /mmc), (b) α-W (Im3m), and (c) W 3 O/β-W (Pm3n) phases; Figure S4: Optical images of the wear tracks and wear scars for (a) WSe x /NP-W, (b) WSe x S y /NP-W_3.6, and (c) WSe x S y /NP-W_9 coatings after tribotesting in N 2 -enriched environment at a load of 5 N at −100 • C. The test duration for each coating is indicated in Figure 7.