Structure and Properties of Multicomponent Arc-PVD Ti–Cr–Mo–Ni–N Coatings on Carbide Alloys HG20, HS123, HV420 and Tool Steel C80W1

1. Introduction

Coatings based on Cr, Mo and Ti nitrides exhibit high physico-mechanical properties, which ensure their relatively high wear resistance [1,2,3,4,5]. However, single-phase coatings of these systems have significant drawbacks that limit their application:

• The low heat resistance of titanium and molybdenum nitrides places severe restrictions on the permissible operating temperature ranges [6,7,8].
• The relatively low hardness of Cr–N-based coatings, combined with a high friction coefficient, promotes greater heat generation in the contact zone, thereby reducing the performance efficiency and service life of components [9].
• These coatings are characterised by a columnar structure. Along the boundaries of columnar crystallites, oxygen rapidly diffuses towards the substrate surface, lowering their heat resistance. Moreover, the columnar morphology is more conducive to crack propagation into the coating depth than an equiaxed (isomorphic) structure, thus reducing the adhesion strength of the substrate–coating composite [10,11,12].

The use of ternary Me–Me′–N systems, in which both metallic elements form nitrides with a face-centered cubic (FCC) lattice that are mutually soluble, significantly improves the coating properties. For example, reference [6] demonstrated that ion-plasma vacuum-arc Ti–Cr–N coatings with chromium concentrations ranging from 10.1 to 25.2 at.% exhibit higher hardness than CrN coatings. At the same time, chromium-rich coatings show better wear resistance in aggressive environments. According to the findings in [13], alloying with chromium raises the oxidation resistance temperature of TiN coatings by over 100 °C. Another example of ternary coatings based on mutually soluble nitride phases is the Cr–Mo–N system. According to [14], the maximum molybdenum concentration at which a (Cr,Mo)N solid solution forms is 30.4 at.%; upon dissolution of molybdenum in chromium nitride, hardness increases due to solid-solution strengthening and an increase in residual macrostresses in the coating, reaching a maximum of 34 GPa at 21 at.% Mo. The system attracts considerable attention due to the excellent tribological properties of molybdenum nitride. Indeed, when a (Cr,Mo)N solid solution forms, the friction coefficient declines as the molybdenum fraction increases. For CrN/Mo₂N multilayers, raising the temperature from ambient to 300 °C causes the friction coefficient to approximately double, a phenomenon attributed to the generation of hard wear debris, including CrN, Mo₂N, and Cr₂O₃ particles. At 600 °C, however, the friction coefficient drops to 0.53, which is ascribed to the appearance of a Magnéli phase (MoO₃) that provides solid-lubricant action. Moreover, these coatings exhibit a 300 °C improvement in thermal resistance relative to single-phase Mo₂N coatings, thanks to the protective CrN layers that act as an oxidation barrier for molybdenum nitride. In some instances, the wear resistance of CrN/Mo₂N coatings is six-fold greater than that of CrN and double that of Mo₂N.

One effective alternative for enhancing surface functionality is to replace the TiN system with Ti–Mo–N. It has been shown [15] that they have a lower friction coefficient and wear rate than single-phase TiN. The increase in hardness is associated with the refinement of nitride grains – it is known that titanium and molybdenum nitrides are practically insoluble in each other [16]; accordingly, during coating deposition, competitive growth of TiN and Mo₂N grains occurs, which block each other. Introducing into the system a nitride-forming metal that does not tend to form nitrides is one of the effective and well-known methods for synthesising “superhard” coatings [17,18,19,20,21]. Such coatings include the ZrN–Ni(Cu), Cr₂N–Ni, TiN–Ni(Cu) systems, etc. The presence of a metallic component in such systems leads not only to an increase in hardness due to the refinement of nitride grains down to several nanometres in size, but also to an increase in impact toughness. As the concentration of the metallic component increases, the columnar structure characteristic of nitride coatings becomes finer, up to the formation of an equiaxed structure – with a thin metallic or amorphous interlayer formed around the nitrides. Thereby, a dispersion-strengthened structure is realised, where the matrix is a soft and ductile metal. In the Ti-N-Ni system, the hardness increases by more than a factor of two at 12 at.% Ni, with a grain size of 15–18 nm, compared with single-phase TiN coatings [22]. A significant synergistic effect in improving coating properties is observed for five-component coatings based on the Ti–Cr–Ni–Mo–N system. Thus, the study found a substantial increase in the wear resistance of coatings of this system on the nickel alloy ÉK61 and on silicon, compared with Ti–Cr–N and Ti–Cr–N–Ni coatings [23]. [24] showed that Cr–Ti–Al–Mo–N coatings exhibited higher hardness (29.3 GPa) and better wear resistance (1.2 × 10⁻¹⁶ m³/(N·m)) than CrTiMoN, TiMoN and CrTiAlN coatings. In work [25], the authors, when investigating Al–Ti–N and Al–Ti–Cr–Mo–N coatings deposited by cathodic arc evaporation (CAE) on carbide alloy (WC–Co, 7 wt.% Co), found that at high hardness values (35.6 ± 1.5 GPa), the coatings of the first system had significantly lower H/E and H³/E² parameters (0.059 and 0.124 GPa, respectively) than the coatings of the second system (0.071 and 0.157 GPa, respectively). These differences resulted in an almost three-fold increase in the wear resistance of the Al–Ti–Cr–Mo–N coatings compared with the Al–Ti–N coatings. The above indicates a considerable effectiveness of five-component coatings based on the Ti–Cr–Mo–N system with the addition of Ni and Al as strengthening elements under various friction conditions. At the same time, when choosing strengthening coatings for different materials, as individual studies show, one should take into account that the nature of the substrate can significantly affect the properties of both the coatings and the coating–substrate composite as a whole. [26] demonstrated the influence of different grades of carbide alloys with different cobalt contents on the processes of crack initiation and propagation was established, due to different macro-stress values arising in them. The effect of carbide alloy compositions on abrasive erosion resistance and on coating wear resistance was confirmed by the authors of [27, 28]. The aim of this study is to determine the influence of substrate materials made of different grades of carbide alloy and tool steel on the structure and properties of Ti–Cr–Ni–Mo–N coatings.

2. Materials and Experiments

The Ti–Cr–Mo–Ni–N coatings were produced by ion-plasma vacuum-arc evaporation using a three-cathode system. The cathodes were prepared from a Ti–Ni alloy (50 at.%), Mo, and Cr, which were installed in evaporators C1, C2, and PI (Figure 1), respectively (C1 and C2 are evaporable cathodes with droplet phase separators, while PI is a direct-flow evaporable cathode).

The deposition time was 90 minutes. Nitrogen served as the reactive gas. Substrate heating to 400 °C was performed indirectly. The coated substrates rotated around both the evaporating cathodes and their own axes. The reactive gas pressure (P_N2 was approximately 3 · 10⁻¹ Pa. The current on the Ti–Ni and Mo cathodes (Ia) was about 100 A, and about 90 A on the Cr cathode. The bias voltage applied to the substrates (Uc) was about 120 V. Prior to deposition, the substrate materials were ion-etched. The resulting coating thickness was 4 ± 0.1 μm. Deposition parameters for all substrates were selected to maximise the deposition rate while minimising the droplet phase content in the coatings

Three carbide alloys grades were selected as substrate materials: WC–Co (6 wt.%) (HG20); WC–TiC (15 wt.%)–Co (6 wt.%) (HS123); and WC–TiC (4 wt.%)–TaC (6 wt.%)–Co (8 wt.%) (HV420) – the grades most widely used for tool and structural applications – together with carbon tool steel C80W1 (C: 0.81–0.89; Si: 0.17–0.33; Mn: 0.33–0.58; Ni: up to 0.25; S: up to 0.028; P: up to 0.03; Cr: up to 0.2; Cu: up to 0.25 wt.%).

Coating thicknesses on various substrates ranged between 3.9 and 4.1 µm.

Morphological features were assessed using a Vega 3SB scanning electron microscope (Tescan, Czech Republic) equipped with an X-ACT SDD energy-dispersive X-ray spectrometer (Oxford Instruments, UK). Microstructure was examined with a JEM-2100 transmission electron microscope (JEOL, Japan) operating in microdiffraction mode, which provided electron diffraction patterns for phase identification. For this purpose, lamellae were prepared from the coatings by ion thinning. Chemical states and elemental compositions were determined by X-ray photoelectron spectroscopy on a Versa Probe II spectrometer (Physical Electronics, USA). Phase composition was further verified by X-ray diffraction on an Ultima IV diffractometer (Rigaku Corporation, Japan) with Cu Kα radiation and a graphite monochromator, using an asymmetric scanning geometry (5° incidence angle). Coherent scattering domain sizes (D), lattice microstrains (ε), and the complex nitride lattice parameter were calculated from (111) and (222) reflections recorded in symmetric geometry. Physical peak broadening was extracted by jointly fitting the specimen against a reference germanium powder (10–20 µm) using pseudo-Voigt functions.

Hardness and elastic modulus for the Ti–Cr–Ni–Mo–N coatings were determined by the Oliver–Pharr method [29] using a Micro-Hardness Tester (CSM Instruments, Switzerland) at a 25 mN load. The plastic deformation work fraction, derived from load–displacement curves, served as a fracture toughness indicator [30]. Indentation depth never exceeded 10 % coating thickness to avoid substrate interference, in line with ISO/CD 14577.

Tribological properties were studied with a Tribometer (Nanovea, USA) in a pin-on-disk configuration at 20 °C and 500 °C in air. The counterbody was an alumina ball; the normal load was 5 N, sliding speed 10 cm/s, sliding distance 100 m, and track radius 3 mm. Wear rates were measured by optical profilometry using a WYKO NT1100 instrument (Veeco, USA).

Coating adhesion and cohesive strength were evaluated by progressive-load scratch testing up to 100 N on a Revetest instrument (CSM Instruments, Switzerland). Acoustic emission, tangential force, and friction coefficient were recorded to determine critical loads L_c1, L_c2, and L_c3 – corresponding to first cracking, adhesive delamination, and substrate exposure under cohesive failure, respectively [31]. Substrate exposure was additionally confirmed by colour changes at the scratch-track bottom and by a sudden friction-force jump; failure morphology was examined by optical microscopy.

Macrostresses arising from differential thermal expansion and structural inhomogeneities were calculated using the Stoney equation [32,33,34]. Substrate curvature radii before and after deposition were measured with an optical profilometer. Layer thicknesses were determined from cross-sectional SEM images. The curvature radius was computed as:

$$\mathrm{R}={\displaystyle \frac{{\mathrm{L}}^2}{8\mathrm{b}}},$$

where L – the specimen length and b – the deflection sagitta.

3. Results and Discussion

1. Structure and substructure characteristics, elemental and phase composition – for coatings deposited on carbide alloys and tool steel substrates.

Figure 2 presents Ti–Cr–Mo–Ni–N coating micrographs grown on various substrate compositions.

Morphological analysis shows that coatings deposited on carbide alloy and steel substrates exhibit a reticulated surface texture. The so-called reticulated texture consists of numerous cells. This texture results from the deposited layer replicating substrate roughness created during sandblasting prior to deposition. On steel substrates (Figure 2d), however, globular formations appear along cell ridges, which can be explained by steel’s much lower thermal diffusivity (13 m²·s⁻¹·10⁶) versus that of carbide alloy substrates (32, 23, and 19 m²·s⁻¹·10⁶ for HG20, HS123, and HV420, respectively) [35,36,37,38]. This leads to a lower cooling rate for the growing coating on steel and thus a possible vapour–liquid–solid phase transition, whereas on carbide alloy substrates coating formation proceeds via a vapour–solid transition. The globule formation on cell ridges is predetermined by more restricted heat removal from these surface regions compared with that from cell sidewalls. Additionally, step-like surface areas appear on steel substrates, indicating a macro-layered coating structure. This feature can be attributed to the steel substrate’s coefficient of linear thermal expansion (CLTE) (14.3·10⁻⁶ K⁻¹) [39] being roughly twice that of the Ti–Cr–Mo–Ni–N coating (8.3·10⁻⁶ K⁻¹, estimated by the rule of mixtures [40,41]). Notably, the differences between the coating’s CLTE and those of the carbide substrates are much smaller: 3.6–5.0, 5.5–6.0, and 5.5–6.0·10⁻⁶ K⁻¹ for HG20, HS123, and HV420, respectively [42,43]. This contrasts with the carbide substrates, where the coating’s CLTE exceeds the substrate’s.

Such CLTE mismatch between the coating and the steel substrate can generate tensile thermal stresses [44], which, at a certain growth stage, rupture the growing layer. This rupture exposes fresh surface where new coating phase nuclei form and grow, producing a new layer.

Table 1 presents EDX chemical composition data for the Ti–Cr–Mo–Ni–N coatings. The results indicate that the coating elemental composition is essentially substrate-independent, within the experimental error.

X-ray phase analysis data (Figure 3, refined by XPS, indicate that the Ti–Cr–Mo–Ni–N coating composition includes Ti₁₋ₓCrₓN nitride, γ-Mo₂N, and Ni.

Molybdenum in the γ-Mo₂N form is confirmed by two independent methods. First, the diffraction pattern (Figure 3) shows peak asymmetry indicating this compound. Second, XPS (Figure 4) places Mo3d₅/₂ and N1s binding energies at 228.1 and 397.4 eV, characteristic of the γ-Mo₂N nitride phase [45].

Titanium and chromium nitride sublayers with cubic lattice, deposited prior to the molybdenum-containing layer, stabilise the γ-Mo₂N phase. The Ti2p₃/₂ binding energy (Figure 5) across the ~170 nm profile remains at 455.3 eV, which is 1.1 eV below the tabulated value for stoichiometric TiN (456.4 eV) [46].

Redistributed ionic–covalent–metallic bonds point to a complex (Ti,Cr)N nitride with B1 structure, typical for TiN and CrN. XPS spectra (Figure 5) exhibit two Cr2p lines: Cr2p3/2 at 574.9 eV and Cr2p1/2 at 584.4 eV. This confirms CrN phase [45,46] – formed as substrates pass the chromium cathode – and (Ti,Cr)N, which forms as overlapping vapor fluxes from TiNi and Cr cathodes condense.

Although X-ray diffraction detects no Ni-bearing phases in Ti–Cr–Mo–N–Ni coatings, XPS (Figure 6) shows a Ni binding energy near 852.4 eV, characteristic solely of metallic Ni [47].

The observed features are attributed to nickel’s much lower thermodynamic probability for bonding with nitrogen compared with the other coating elements [48,49,50,51]. Combined with nickel’s extremely low solubility in nitrides and the absence of clear diffraction peaks, this supports the view that nickel exists as an X-ray amorphous component – consistent with results reported in [22].

Ti–Cr–Mo–Ni–N coating substructure characteristics for the titanium-based nitride phase are listed in Table 2. A notable lattice parameter decrease is observed for the titanium nitride-based phase on all substrates, compared with stoichiometric TiN [52].

This confirms solid solution formation – chromium in titanium nitride – during coating deposition, where the lattice parameter should decrease given the Ti and Cr atomic radii ratio. Moreover, the determined average nitride phase lattice parameters match literature data [53] for Cr-in-TiN solid solutions at concentrations corresponding to those elements in our coatings. Simultaneously, the titanium-nitride-based phase exhibits large microstrains, whose origin lies in structural defects and compositional inhomogeneity. Their lowest values occur for coatings deposited on HG20 carbide alloy. As Table 2 shows, microstrain growth accompanies lattice parameter increases for the corresponding phases, consistent with published data. Significant microstrains alongside relatively small CSD sizes rather suggest that most strain arises from nonuniform component distribution. If crystal structure defects were the sole strain source, their density would have to exceed 10¹² cm⁻², which is physically improbable.

TEM structure studies reveal that Ti–Cr–Mo–Ni–N coatings on all tool material substrates form multilayer structures with a ~70 nm modulation period (Figure 7). Individual crystallites grow through several layers.

Owing to the substrate–coating thermal expansion mismatch noted above, together with possible compositional and structural heterogeneities arising in growing PVD coatings [54,55], compressive or tensile macrostresses typically develop [56], noticeably affecting the overall coating–substrate composite properties [57,58,59,60]. Accordingly, their values were estimated by the Stoney method. Table 3 lists the obtained results.

σ_calc – calculated macrostress;

R₀ – substrate curvature radius before deposition;

R – coating–substrate composite curvature radius;

tₛ – substrate thickness;

t_f – coating thickness;

νₛ – substrate Poisson’s ratio;

Eₛ – substrate elastic modulus.

The results show tensile stresses developing in the coating on steel substrate. Compressive stresses, roughly equal in magnitude, occur in coatings on carbide alloy substrates. The calculated tensile macrostress values for the Cr–Ti–Ni–Mo–N coating on tool steel confirm the earlier inference that such stresses are likely here, given the substrate CLTE substantially exceeding the coating CLTE, and this factor’s effect on coating morphology – manifesting as macrolayering.

To assess how compositional inhomogeneity at the coating–substrate interface – arising from mutual diffusion between coating and substrate components – might affect structural stresses and their contribution to the integral macrostress value, a study was performed on the diffusion transition zone. TEM imaging of near-interface regions and local X-ray microanalysis were employed. Depth determination error corresponded to half the local analysis zone diameter. For the Cr–Ti–Ni–Mo–N coating on carbide alloys, the transition zone was defined by the diffusion depth for Cr, Ni, and Mo (coating elements) into the substrate and for W (substrate element) into the coating; on C80W1 steel, by the diffusion depth for Cr, Ni, and Mo into the substrate and for Fe into the coating. The resulting TEM images and local elemental microanalysis data for the coating–substrate transition zones are presented in Figure 8.

Estimated values for the coating–substrate diffusion transition zones were obtained by processing the experimental data. These are listed in Table 4. As the results show, the largest zone corresponds to C80W1 steel substrates

This is attributed both to the higher diffusion mobility (diffusion coefficients D at temperatures reached in the coating–substrate contact zone) for iron atoms in the carbide alloy phases and coating elements in iron, compared with the mutual diffusion mobility for the marker elements (listed above) in Ti–Cr–Mo–Ni–N coatings and carbide substrates [61,62,63,64,65], and to local interface heating during deposition. This is also due to more efficient heat accumulation in the growth zone, since steel removes heat less effectively than carbide alloy substrates. Evidence for this effect comes from the temperature field simulation results presented in [28] for analogous-composition substrates and growing coatings under high-energy particle flux during arc-PVD Ti–Al–Ni–N deposition. Those simulations showed that the highest surface and coating-interior temperatures occur for the lower-thermal-diffusivity material – C80W1 steel. For the Ti–Cr–Mo–Ni–N coating, this effect may be even stronger, given the presence of heavy molybdenum particles bombarding the substrate and the correspondingly greater energy transferred to the growing coating, according to:

E(U) = U∙Z∙e + Uo eV

(1)

where Z – the mean particle charge in the plasma flux; U – the substrate bias potential; e – the electron charge; U₀ (eV) – the initial ion energy in the arc-generated plasma used for cathode evaporation [66,67]. For Mo ions, U₀ and Z exceed those for Al and Ti – the main Ti–Al–Ni–N coating constituents [68,69].

2. Physical–mechanical, tribological, and adhesion to substrate properties for Ti–Cr–Mo–Ni–N coatings deposited on different substrates.

The values for hardness (H), elastic modulus (E), relative plastic deformation work (Wₚ), and the H/E and H³/E² ratios – which characterise material resistance to elastic and plastic deformation [70, 71] – are listed in Table 5.

The results show that coatings on carbide alloy substrates exhibited higher H and E values than those on C80W1 steel. At the same time, the H/E and H³/E² ratios for coatings on tool steel were maximal. The Wₚ value, which serves as a fracture toughness indicator [30], shows that this property is more pronounced for Ti–Cr–Mo–Ni–N coatings formed on carbide alloy substrates. The lower hardness on tool steel may stem from tensile macrostresses in this substrate – due to the substrate CLTE being nearly twice that of the coating, as noted above – as well as from compositional inhomogeneities in the substrate–coating near-interface regions, which arise depending on the deposited material and substrate nature [1, 57, 58].

Investigations into adhesion strength for Ti–Cr–Mo–Ni–N ceramic-metal coatings on various tool materials revealed a cohesive failure mode for all coating–substrate composites studied. Since the processes during abrasive particle penetration resemble those in controlled scratch testing with an indenter [72,73], this method is regarded as a model for hard abrasive sliding over a surface under increasing load. Figure 9 and Figure 10 present typical scratch-test results for Ti–Cr–Mo–Ni–N coatings on various tool substrates: friction force (Ff), friction coefficient (μ), and acoustic emission (AE) curves, along with scratch micrographs produced by indenter sliding under increasing load.

At room temperature, all Ti–Cr–Mo–Ni–N coatings on every substrate exhibited a similar friction coefficient (~0.5), but they differed in wear rate. Coatings deposited on carbide alloys showed wear craters whose profile allowed wear depth estimation; this depth was comparable to the initial coating surface roughness (Figure 11a). In contrast, the coating on steel exhibited a substantially higher wear rate (Figure 11b).

The obtained tribological test results are somewhat at odds with several studies showing that high H/E and H³/E² values govern enhanced coating wear resistance [77,78,79]. Indeed, the highest wear rate belongs to the C80W1-coating composite, which exhibits maximum H/E and H³/E² values (Table 5). This apparent contradiction likely stems from different wear failure mechanisms for this composite versus the carbide alloy–Ti–Cr–Mo–Ni–N coating composites. For the coating on steel substrate, failure proceeds via delamination of the macrolayers present in the coating structure, as revealed by morphology studies. The likelihood of such a structure forming on this substrate was discussed above. Wear track examination for carbide alloys specimens with the Ti–Cr–Mo–Ni–N coating confirms that wear occurs by small fragment detachment from the coating within the track and along its edges (Figure 12a). In contrast, the wear track for the coating deposited on steel (12b) shows features indicative of wear through plastic deformation via macrolayer displacement within the coating structure.

At 500 °C, Ti–Cr–Mo–Ni–N coatings on carbide alloys show some increase and similar friction coefficients ~0.54–0.59 (Table 6), with noticeable wear debris transfer from the counterbody onto the track surface, hindering wear rate measurement. The coating on steel shows a slight friction coefficient decrease at 500 °C (to 0.47 vs. 0.53) accompanied by a higher wear rate (Figure 11b, d). The lower friction coefficient for Ti–Cr–Mo–Ni–N coatings on steel stems from their macrolayered structure, described earlier when discussing coating morphology on this substrate, and the resulting decrease in their shear strength due to poor interlayer adhesion, which reduces friction force [75]. The higher friction coefficients for Ti–Cr–Mo–Ni–N coatings on carbide alloy substrates at high temperature can be explained by chromium being the main coating element; its oxidation in air during high-temperature tribological tests forms Cr₂O₃-based wear debris, which acts as an abrasive in the contact zone and raises friction force [80]. Molybdenum present in the coating, which according to our and others’ studies should improve tribological performance at high temperature through MoO₃ lubricious phase formation [81,82,83,84,85], here has no significant effect due to its low content (~4 at.%) against high chromium (~35 at.%).

4. Conclusions

Analysis of structure and properties for Ti–Cr–Mo–Ni–N ceramic-metal coatings on various tool substrates revealed a difference in coating morphology on tool steel versus carbide-based substrates – namely, the appearance of globular formations. This is attributed to hindered heat removal from the ion flux bombarding the substrate, due to steel’s lower thermal diffusivity compared with the carbide alloys used. This leads to a lower cooling rate for the growing coating on steel and thus a possible vapour-liquid-solid phase transition, whereas on carbide alloy substrates coating formation proceeds via a vapour-solid transition. Additionally, these coatings show stepped surface areas, indicating a macrolayered structure. This difference likely results from tensile thermal stresses (in contrast to coatings on carbide alloys), which at a certain growth stage rupture the growing layer, exposing fresh surface where new coating phase nuclei form and grow, producing a new layer. The same reason may explain the relatively lower hardness for the coating on steel (20.3 ± 1.1 GPa) compared with coatings on carbide substrates (25.1 ± 2.2 – 21.7 ± 1.1 GPa).

Relatively low adhesion strength was found for the coating on tool steel versus coatings on carbide alloy substrates. This effect stems from tensile macrostresses in the former case and from the coating morphology on tool steel – its macromultilayer nature, which reduces shear strength. At the same time, the latter may account for the relatively lower friction coefficient for this coating compared with carbide-based coatings.

The studied coatings exhibit good tribological performance both at room and elevated temperatures – relatively low wear rates, with the highest wear rate for the coating on tool steel.

Considering the functional property study for Ti–Cr–Mo–Ni–N coatings on various tool substrates, it can be concluded that they may be effectively used as strengthening coatings on carbide alloys belonging to the HG20, HS123, and HV420 groups. Their effectiveness on carbon tool steel would be reduced due to lower hardness, weaker substrate adhesion, and higher wear rate.