Mechanical and tribological properties of B-C-N coatings sliding against different wood balls

2.1 Deposition of BCN coatings

The BCN coatings were simultaneously deposited on Si (100) wafers and 316L stainless steel substrates using magnetron sputtering system via adjusting N₂ flow. In order to obtain a stable glow, the B4C target was installed on RF target. The polished stainless steel disks (Φ30×4 mm²) and Si (100) wafers were cleaned ultrasonically in deionizer water and ethanol for 20 min before being installed on substrate holder. When the chamber was evacuated to 9×10⁻⁴ Pa, substrates were cleaned with Ar plasma at a bias voltage of −100 V for 30 min. The B₄C target power and substrate temperature were set as 100 W and 100 °C, separately. The BCN coatings were deposited for 1 hour at different N₂ flows (5 sccm, 10 sccm, 15 sccm and 20 sccm). According to N₂ flow, the BCN coatings were denoted as BCN-5, BCN-10, BCN-15 and BCN-20.

2.2 Characterization of BCN coatings

The bonding structures of BCN coatings were detected using X-ray photoelectron spectroscopy (XPS, AXIS Ultra-DLD), and the element content in coatings was measured at the same time. The Vickers hardness of coating was measured using a microhardness tester (HVS-1000, china) equipped with a 136° diamond probe. The applied load and holding time were set as 0.098 N and 10 s. The Vickers hardness was calculated according to the four-point measurement method. The coating adhesion to the 316L stainless steel substrate was defined in the scratch tester (WS-2005, Scratch Tester, China) equipped with a hemispherical diamond tip (R = 0.2 mm). The applied load, sliding speed and wear track length were set to 20 N, 20 N/min and 3 mm, separately. The scratch test was repeated three times to ensure its reliability.

2.3 Tribological Properties of BCN coatings

The friction behavior of BCN coating sliding against different wood balls (e.g. acerbic, beech and lauan wood with a diameter of 8 mm) was investigated using a homemade ball-on-disc tribo-meter shown in Figure 1. During experiment, the friction forces of tribo-pairs were detected by a LMA-A-10 N load cell (Kyowa Co. LTD., Japan). The load cell voltage was measured by a DPM-700B strain amplifier (Kyowa Co. LTD., Japan) and recorded by NR-110/150 data collection system (Keyence Co. LTD., Japan) with a compatible computer. The resolution of the tribo-meter used in this study was 0.001. The applied load, sliding speed and sliding distance were set as follows: 3 N, 0.1 ms⁻¹ and 300 m, separately. After friction test, the morphology of wear tracks was observed using an optical microscope.

Figure 1

Schematic diagram of ball-on-disc tribo-meter.

3.1 Microstructure of BCN coatings

As seen in Table 1, the N content was non-proportional to the N₂ flow, but the boron content gradually decreased from 14.3 at.% to 7.2 at.%. In addition, the carbon content slightly increased with the rise of high N₂ flows. Actually, the element contents in coating were determined synergistically by the following factors such as deposit temperature, target power, substrate bias and N₂/Ar ratio. For example, V. Linss et al. [10] pointed out that the B/C ratio in BCN coatings did not change with the N contents at low temperatures (≤ 100 °C) unless being deposited at elevated deposition temperature. Once the nitrogen plasma was saturated in chamber, neither increasing N₂ flow nor increasing N₂ pressure could cause regular change in nitrogen content [10, 17].

In order to reveal the chemical structures of BCN coatings, the corresponding B1s, C1s and N1s XPS spectra are illustrated in Figure 2. It is clear that the low intensity peaks are detected in B1s spectrum for BCN-5 and BCN-10 coatings (10.6~14.3 at.% B). The peak located at 186.5~186.9 eV were assigned to B-C bond (B₄C) [18], whilst that located at 189.3~189.5 eV were attributed to B-N bond. In addition, the B-O bond was detected at the binding energy of 191.5 eV. As compared with the former reports [8, 9, 13], the position of B1s peaks shifted to low binding energies, which was attributed to the high C concentration in coatings [19]. However, no obvious peaks were detected in B1s spectrum for BCN-15 and BCN-20 coatings due to their low boron contents. As seen in Figure 2(b), two deconvoluted peaks located at 398 eV and 400 eV were detected in N1s spectrum, which corresponded to N-B [19, 20] and N-C bonds [6], separately. According to the former reports [19, 20], the sp²N-B and sp²C-N bonds usually located at 397.3~398.0 eV and 400.5~401.5 eV, separately. Thus, the N-B and N-C bonds located at binding energy of 398 eV and 400 eV were sp² hybridization. In terms of C1s spectrum, three deconvoluted peaks were detected at different binding energies. According to the former reports, the binding energies of 284.2~284.6 eV, 285.4~285.9 eV and 286.4~286.9 eV were assigned to C-C (sp²) [6, 8, 9], C-N (sp²) [8, 9, 21] and C-N (sp³) [9, 21], respectively. The binding energy usually shifted to high energy zone after N bonded with C atom because of the high electronegativity of N atom. Accordingly, as seen in Figure 2(c), the binding energy in the range of 284.4~284.8 eV corresponded to sp²C-C, and the binding energies of 285.3~285.9 eV and 286.5~286.7 eV were assigned to sp²C-N and sp³C-N, separately. In addition, the appearance of C-O bond indicated that the surface of coatings was contaminated by oxygen from air. The coexistence of B-N, B-C, and N-C bonds in BCN coatings indicated the formation of a ternary B-C-N hybridization [9] and ternary composite materials [10].

3.2 Mechanical properties of BCN coatings

As seen in Table 1, the highest Vickers hardness of 1614.7 HV was obtained for BCN-5 coating, which attributed to its high B content. Similarly, Deng et al. [6] also pointed out that the high B content in BCN coating was helpful to increase its hardness. However, increased N₂ pressure [9] or N₂ flow [8, 19] also resulted in the decline of hardness for BCN coatings. That was because the increased N proportion could promote N atoms to preferentially bond with C, and resulting in decreased B-C and increased C-N bond content [8]. Based on the previous reports, the sp²B-N and N=C bonds were verified as “soft” phases [22, 23, 24], whereas the B-C and sp³N-C bonds were thought as “hard” phases [25, 26]. Here, BCN-10 coating only has sp²C-C and sp²B-N bonds without sp³N-C in its C1s spectrum, resulting in the lowest hardness of 1316.9 HV.

The coating adhesion to the 316L stainless steel substrate was performed in the scratch tester as critical load was one of the more significant parameters. The acoustic emission signals and the corresponding friction force curves are illustrated in Figure 3. Usually, a sudden increase in acoustic emission signal means the appearance of tiny cracks on scratches, but the deposition defect or surface roughness also could cause a slight acoustic emission signal. Therefore, combining the friction force curve (e.g. slope or intensity) and scratch morphology is a more reasonable way to determine the critical load. As seen in Figure 3(a), the intensity of the acoustic emission signal suddenly increased after presenting a smooth straight line, and their friction force also fluctuated. The low critical loads (L_c1) were thought of crack initiation resistance [27, 28]. Accordingly, the low critical loads (L_c1) were determined, which increased from 1.5 N to 8.5 N with an increase of N₂ flows. In addition, the optical micrographs of scratches are illustrated in Figure 4s to determine the high critical loads (L_c2) corresponded to the appearance of flake crack or delamination on scratches [29]. As seen the amplification optical photographs in Figure 4(b), the scaly cracks appeared on their scratches at high critical loads. The corresponding acoustic emission signal and frictional force curves also presented some fluctuations in Figure 3. As a result, the L_c2 values were determined to be 13.0 N, 13.8 N, 13.2 N and 15.0 N, separately. Refs. [27, 28] pointed out that the scratch crack propagation resistance (CPRs) determined by critical loads could be used as quick qual itative indication of the film toughness or used in a quality control process for tough film. In that case, the CPR_s of coatings could be determined to be 17.3 N², 41.9 N², 39.2 N² and 55.3 N² by the following equation, indicating an improved crack propagation resistance of coatings with an increase of N₂ flow. That result was attributed to the increased fraction of N=C bond in N1s spectra (48.5%→ 67.0 %), where the “soft” phases such as N=C or sp²C-N were helpful to improve the of coating adhesion and crack propagation resistance [9].

Figure 2

XPS spectra of BCN coatings.

Figure 3

Acoustic emission signals and friction force for BCN films by a scratch tester.

Figure 4

Optical micrographs of scratches (a) and corresponding amplification optical photographs (b) by a scratch tester.

3.3 Tribological properties of BCN coatings

The friction behavior of BCN coatings is shown in Figure 5. As sliding against acerbic balls, the friction coefficient gradually increased in the whole sliding distance except BCN-5 coating. As sliding against beech balls, BCN-5 and BCN-10 coatings presented a steady coefficient of friction. In contrast, the friction coefficient increased dramatically with an increase of sliding distance for BCN-15 and BCN-20 coatings, implying the change of friction characteristics. As sliding against lauan balls, only BCN-5 coating kept a steady coefficient of friction within the whole sliding distance. As for the influence of mating balls, the mean stable friction coefficient (0.76~0.83) was obtained as sliding against acerbic balls in Figure 5(d), while the friction coefficient fluctuated greatly as sliding against other two balls (beech or lauan balls). In addition, the friction coefficient showed a great fluctuation. That was because the plant materials have a heterogeneous internal structure and exhibit high susceptibility to external conditions, which results from their unstable physical properties contrary to solids [15]. As seen in Figure 5d, the relatively high friction coefficient was obtained as sliding against softwood. It was a result of big contact deformation and high adhesion at the friction interface for softwood.

Figure 5

Friction behavior and mean friction coefficient of BCN coating as sliding against wood balls.

Optical images of wear scars for tribo-pairs are shown in Figures 6-9. As seen in Figure 6, smooth wear scars without any obvious scratches were observed regardless of mating balls. In addition, a small amount of debris adhered on the wear marks as sliding against acerbic and beech balls, whilst the debris accumulated at the edge of the wear mark as sliding against lauan balls. That result indicated the mild friction behavior occurred during their friction process with slight adhesion wear. In terms of BCN-10 coating, shallow abrasion marks were observed in Figure 7, even the scratches on the substrates have not been worn off. The wear track on BCN-10 coating as sliding against acerbic balls was narrower than those sliding against other two balls (beech and lauan balls). As seen in Figure 8a, some scratches obviously appeared on the wear track as sliding against acerbic balls, implying that the main wear mechanism was abrasive wear. In contrast, the abrasive wear was alleviated as sliding against other balls whilst appearing a small amount of wear debris adhesion (Figure 8b and 8c). Similarly, as seen in Figure 9a, scratches and wear debris were also observed on the wear tracks as sliding against acerbic balls, indicating that the abrasive wear and adhesion wear occurred during the friction process. For other two mating balls, more wear debris accumulated at the edge of the wear track (Figure 9b and 9c). Regarding the influence of wood species, softwoods (i.e. beech or lauan) were easier to produce tiny granular debris. In that case, the strong hygroscopicity of the granular debris produce large deformation to increasethe real contact surface area, and then the adhesion between granular debris increase the friction force [15, 16]. This was one of the reasons to increase friction coefficient for softwood.

Figure 6

Optical images of wear scars for BCN-5 coating sliding against different wood balls.

Figure 7

Optical images of wear scars for BCN-10 coating sliding against different wood balls.

Figure 8

Optical images of wear scars for BCN-15 coating sliding against different wood balls.

Refs. [27, 28] pointed out that the low critical load (L_c1) and CPRs could evaluate the crack initiation resistance and the persistence before a catastrophic failure occurs, separately. Though the critical loads and CPR_s value of BCN-5 coating were the lowest as compared with other coatings, no fatigue cracks or delamination appeared on its wear tracks. That result was attributed to the high elasticity of wood which could effectively alleviate the stress concentration during friction testing. In terms of wood composition, some calcium magnesium potassium oxides are contained in wood, these oxides can play the role of abrasive particles during the friction process. As seen in Figures 6-9, smooth wear tracks were observed on BCN-5 and BCN-10 coating as sliding against acerbic balls, while serve abrasive wear was obtained for other two coatings, indicating that the wear resistance weakened with an increase of N₂ flows. However, the abrasive wear was alleviated as sliding against beech or lauan balls by comparison. It could be inferred that oxide particles mixed into tiny granular wood debris and then resulted in the loss of cutting function of oxides. Considering the XPS analysis, the reduced boron content weakened the B-N and B-C bonds as seen in Figure 2a. The absence of hard phase i.e. boron carbide phase in coatings was not conducive to improve the wear resistance of coatings. The excessively localwear of cutting edge resulted from chipping or abrasive wear was not conducive to improving the surface quality of wood processing [30]. As a result, it could be implied that a wide adaptability for BCN-5 and BCN-10 coatings as wood- working tools to cut hard or soft woods, whilst BCN-15 and BCN-20 coatings were suitable for cutting softwood.

Figure 9

Optical images of wear scars for BCN-20 coating sliding against different wood balls.