Influence of Interlayer Materials on the Mechanical Properties and Thermal Stability of a CrAlN Coating on a Tungsten Carbide Substrate

Abstract

CrAlN coatings have earned significant attention for use in cutting tool coating applications due to their excellent properties such as high hardness and superb oxidation resistance. It is well known that the interlayer between the coating and the substrate can influence the mechanical properties of the coatings. In this work, three interlayers—CrN, CrZrN, and CrN/CrZrSiN—were synthesized between a CrAlN coating and a tungsten carbide substrate to improve the mechanical properties and thermal stability of the CrAlN coating. All the CrAlN coatings with their respective interlayers showed high hardness values in the range of 34.5 to 35.1 GPa, and they were not significantly affected by the interlayer type. However, wear and scratch tests showed that the CrAlN coatings with CrN and CrN/CrZrSiN interlayers exhibited an improved friction coefficient of 0.33 and adhesion strength (Lc2) of 69 N compared to the CrAlN coating with the CrZrN interlayer. These improved wear properties were attributed to the H/E ratio of the interlayer between the coating and the substrate, in that the CrN and CrZrSiN interlayers effectively induced a smooth transition of the coating stress under a loading condition. During the thermal stability tests, the hardness of the CrAlN coating with the CrN/CrZrSiN interlayer was maintained up to 1000 °C due to the excellent oxidation resistance of the CrZrSiN layer, which contained an amorphous Si_xN_y phase.

1. Introduction

For the past decade, Cr-based ternary hard coatings such as CrAlN have been applied in industrial fields due to their superior wear and oxidation resistance at high temperatures compared to binary CrN coatings [1,2,3]. Their excellent thermal stability has been theoretically and experimentally reported in many publications, suggesting that both chromium and aluminum in CrAlN coatings are able to form protective oxide layers to inhibit oxygen diffusion [2,3]. Typically, the microstructure and phase of CrAlN coatings are correlated with their hardness and oxidation resistance, which can be controlled by precisely tuning the elemental composition [4,5,6]. It is well established that increasing the Al content up to the solid solubility limitation for a cubic B1 structure leads to enhanced mechanical and tribological properties via a strengthening effect. However, in the case of excess Al content over the cubic solubility of Cr_1−xAl_xN (x_max, ~0.7), a mixed structure of cubic B1 and hexagonal B4 (wurzite AlN) is formed, resulting in reduced hardness and wear performance [4]. Importantly, improvement in CrAlN coatings is still required in order to meet the strict criteria of a fast-changing industry.

In general, it is well known that interlayers are very effective in enhancing the adhesive strength between substrates and materials to be deposited. Recent research has demonstrated that the application of interlayers to hard coatings enables the modification of the structure and internal stress to enhance the mechanical and tribological properties [7,8,9,10,11,12]. In particular, the modulation ratio, i.e., H/E or H³/E², is a key factor in the enhanced mechanical and adhesive strength in the presence of an interlayer, since it plays a simultaneous role of hindering dislocation glide to some extent and relieving a high stress gradient between the substrates and coatings. Therefore, it is believed that the mechanical and tribological properties of CrAlN coatings could be enhanced by the introduction of appropriate interlayer materials. Nevertheless, the effects of various interlayers for CrAlN coatings have never been reported in depth.

In our previous study, CrAlN monolithic layer coatings showed great mechanical properties such as high hardness (35 GPa) and high adhesion (46 N for SKD61 steel), but the coatings also showed relatively poor wear resistance and thermal stability [4]. In addition, other monolithic layers such as CrZrN and CrZrSiN coatings with a low friction coefficient of approximately 0.25 and excellent thermal stability up to 800 °C have been reported [13,14,15,16,17]. To deposit a CrAlN coating on a very hard WC (≈20 GPa) substrate without delamination, the use of an interlayer of these Cr-based nitride compounds could be advantageous for adhesive strength by reducing the gap of hardness and Young’s modulus between the CrAlN coating and the WC substrate. Although the formation of Si_xN_y compounds in the coating shows a strong oxidation resistance, nitride interlayers containing Si have not been researched yet. Accordingly, in this work, CrAlN coatings with three interlayer materials of CrN, CrZrN, and CrZrSiN were designed and synthesized in order to obtain coatings with excellent hardness values, friction coefficients, adhesion properties, and high-temperature properties. This work clarifies the importance of structuring an interlayer between the CrAlN top coating and the substrate, which will be a significant benefit for practical applications in the cutting tool industry.

2. Experimental Details

CrAlN coatings with the selected interlayers were synthesized on two types of substrates—Si (100) wafers and WC-6 wt.% Co substrates—using a closed-field unbalanced magnetron sputtering system. The CrN, CrZrN, and CrN/CrZrSiN interlayers were deposited between the CrAlN coating and the substrate. The thickness of each interlayer was fixed at 300 nm, which was approximately 10% of the total thickness, for avoiding unexpected fracture generation due to shear stress [8]. Pure Cr single and CrAl, CrZr, and CrZrSi segment targets were used as target materials. Before the deposition process, the Si (100) wafers and WC substrates were ultrasonicated in acetone for 30 min and then dried well. After evacuating a base pressure below 2.8 × 10⁻³ Pa, the substrates were further cleaned by Ar plasma etching for 20 min at an Ar pressure of 0.4 Pa with a DC power of 0.5 kW. The deposition of the coatings was performed in an Ar–N₂ mixed atmosphere, and the working pressure was maintained at 0.52 Pa. All the layers were deposited using pulsed DC power (freq.: 25 kHz, duty ratio: 70%). The distance of applied DC bias and the deposition temperature during deposition were fixed at −100 V and 400 °C, respectively.

The microstructures of the CrAlN coatings with various interlayers were examined via a field-emission scanning electron microscope (FE-SEM, JEOL, JSM-7100F, Tokyo, Japan) operating at an accelerated voltage of 15 kV, and the chemical compositions of each layer were determined using energy dispersive spectroscopy (EDS, JEOL, JED-2300). The hardness and elastic modulus of the individual layers—e.g., the CrN, CrZrN, and CrN/CrZrSiN interlayers—and the CrAlN coating were measured using nanoindentation (Helmut Fischer, HM2000, Sindelfingen, Germany) with a load of 25 mN and a dwell time of 30 s. These measurements were carried out 12 times. In order to alleviate the thickness effect, the indentation depth did not exceed approximately 0.24 μm and was kept to less than 10% of the total coating thickness [18]. To obtain reliable results, nanoindentation tests were carried out 10 times on each sample. The friction coefficients of the coatings were measured using a ball-on-disk-type wear tester with an alumina counter ball (Al₂O₃, Ø = 9.25 mm) at room temperature after annealing at 500 °C. The sliding velocity was 0.25 m/s, and the total sliding distance was 1000 m with an applied load of 5 N. Tribological tests on each sample were carried out 4 times to obtain precise and reproducible results, and the representative average results were used. A scratch tester (CSM, Revetest, Corcelles, Switzerland), with a Rockwell C diamond stylus with a radius 200 μm, was used to determine the quantitative value of the adhesion strength by measuring the place to be initially delaminated, denoted as critical load (Lc2), during a progressive loading up to 100 N. The load rate and sliding speed were maintained at 100 N/min and 10 mm/min, respectively. The track of the coatings was investigated using optical microscopy (OM, Olympus, BX51M, Tokyo, Japan), and the failure mode was determined. The surface morphologies were examined using atomic force microscopy (AFM, Park Systems, XE-100, Suwon, Korea), and the scan area was fixed at 10 μm × 10 μm. All the coatings were annealed at temperatures ranging from 500 to 1000 °C in air for 30 min, and the hardness values were investigated using nanoindentation (Helmut Fischer, HM2000, Hünenberg, Switzerland) to evaluate the thermal stability.

3. Results and Discussion

3.1. Characteristics of the CrAlN Coatings

The CrAlN coatings with the three interlayers were synthesized to enhance their hardness values, friction coefficients, adhesion properties, and thermal stability. Cross-sectional images of the CrAlN coatings with different interlayers are shown in Figure 1. The total thickness of all the CrAlN coatings was fixed at 3 μm and was controlled by the deposition duration based on the sputtering yield. Most of the layers, e.g., the CrN, CrZrN, and CrAlN layers, exhibited a columnar structure, while the CrZrSiN layer showed a featureless structure due to high Si content (

Table 1. Chemical composition of CrZrN, CrN, and CrZrSiN interlayers and the CrAlN coating.

Layer	Chemical Composition (at%)
CrZrN	35.4Cr-14.3Zr-50.3N
CrN	49.2Cr-50.8N
CrZrSiN	18.8Cr-9.3Zr-18.4Si-53.5N
CrAlN	23.8Cr-25.2Al-51.0N

). This phenomenon could be attributed to the fact that a nanocrystalline or amorphous phase with high Si content (Si > 7%) was formed, which is in good agreement with previous work [19]. The chemical composition of the CrAlN top-layer coating was controlled to be 1:1 to obtain the optimal solid solution effect [4]. As can be seen in Figure 2, no significant difference was observed in the hardness and elastic modulus dependent on the interlayer material, and they were measured to have a hardness and elastic modulus ranging from 34.5 to 35.8 GPa and from 423.6 to 425.0 GPa, respectively.

3.2. Tribological Properties of the CrAlN Coatings

The tribological properties of the CrAlN coatings were evaluated by carrying out ball-on-disk-type wear tests, the results of which are shown in Figure 3a. The friction coefficient curves were relatively stable with little scattering. The CrAlN coating with the CrZrN interlayer exhibited a high friction coefficient of 0.41; this value was similar to that of the CrAlN coating without an interlayer. However, the friction coefficient values of the CrAlN coatings with CrN and CrN/CrZrSiN interlayers drastically decreased, and the coating with the CrN/CrZrSiN interlayer produced the lowest friction coefficient of 0.33. To further investigate the tribological properties, the wear rate (K) was calculated [20].

K = V/F∙d

(1)

where K is the wear rate, V is the wear volume, F the normal load, and d the sliding distance. As shown in Figure 3b, it was revealed that the wear rate was affected by the interlayer material, and the results showed an identical trend to that of the friction coefficient. These wear resistance differences in the CrAlN coatings dependent on the interlayer could mainly be associated with the gradient of the modulation ratio (H/E) of the substrate/interlayer/coating. As reported, an interlayer with an intermediate H/E value between individual values of the coating and the substrate can increase the coating toughness improvements involving the wear resistance of the coating [21,22]. In this work, the H/E ratio of the CrZrN interlayer exhibited a much higher value (0.119) than that of the CrAlN (0.089), as listed in

Table 2. Hardness, elastic modulus, and H/E ratio of the WC substrate; the CrZrN, CrN, and CrZrSiN interlayers; and the CrAlN coating.

Layer	Hardness (GPa)	Elastic Modulus (GPa)	H/E Ratio
WC (Substrate)	19.6	510.6	0.040
CrZrN	35.1	293.1	0.119
CrN	23.3	305.6	0.076
CrZrSiN	24.5	288.6	0.083
CrAlN	35.8	425.0	0.089

. This could be explained by the fact that a large difference in the H/E ratio between the CrAlN and CrZrN interlayers exerted a non-uniform distribution of loading stress, which resulted in the poor wear properties of the coating [7]. In view of the CrAlN coating with the CrN/CrZrSiN interlayer structure, however, there was a gradual increase in the H/E ratio from the WC substrate (H/E, 0.040) to the CrN interlayer (H/E, 0.076), CrZrSiN interlayer (H/E, 0.083), and the CrAlN coating (H/E, 0.089). Therefore, the enhanced wear properties could be attributed to the fact that the CrN and CrZrSiN interlayers effectively induced the stress relief between the interfaces by the sequential H/E ratio (WC < CrN or CrZrSiN < CrAlN) during the wear test.

3.3. Adhesion Properties of the CrAlN Coatings

The scratch test is an informative method for evaluating the cohesion and adhesion properties of coatings. During the scratch test with progressive loading on a coating, three failure modes were identified, which could be as critical loads of Lc0, Lc1, and Lc2. We defined a semi-circular crack inside the scratch track as Lc0, a partial fragment at a track edge as Lc1, and an initial spallation or delamination of the entire coating surface as Lc2 [23,24]. To quantify the adhesion strength, the Lc2 was used, and the average values were calculated from the five measurements made for each coating. Figure 4 shows the entire scratch tracks of all the CrAlN coatings. The Lc2 of the CrAlN coating with the CrZrN interlayer exhibited a slightly increased value of 23.6 N compared to 17.8 N for the CrAlN coating without an interlayer. In the scratch tracks of the CrAlN coatings, chipping cracks were observed, and the coatings were totally delaminated over Lc2. It has been reported that chipping will often occur during a scratch test of a high-hardness coating on a hard and brittle substrate [23,25]. However, the Lc2 of the CrAlN coatings with CrN and CrN/CrZrSiN interlayers showed much improved adhesion strength (approximately 69 N), and there were no obvious chipping cracks or signs of delamination in any of the scratch tracks.

The CrAlN coatings with the CrZrN interlayers showed poor crack resistance, as evidenced by a lower Lc2 relative to the CrN and CrN/CrZrSiN interlayers. This phenomenon is correlated with the resistance to plastic deformation of these coatings [25]. In general, once a crack is initiated during a scratch test, it propagates promptly and failure subsequently occurs due to the low toughness of the coating. Conversely, the CrAlN coating with the CrN/CrZrSiN interlayer exhibited greater toughness because the CrN and CrZrSiN interlayers effectively induced a smooth transition of the coating stress. Thus, crack propagation was inhibited during the wear test, and this led to an improvement in wear resistance for the CrAlN coating.

3.4. High Temperature Properties of the CrAlN Coatings

Figure 5 presents the hardness variation in the CrAlN coatings dependent on the interlayer after annealing in air at temperatures ranging from 500 to 1000 °C for 30 min. In the case of the CrAlN coating with the CrN/CrZrSiN interlayer, the hardness value was maintained above approximately 28 GPa up to 1000 °C, whereas the hardness values for the other coatings decreased drastically to 20 GPa over 800 °C. This revealed that the CrAlN coating with the CrN/CrZrSiN interlayer had an excellent thermal stability compared to the coatings with other interlayers. In previous studies, CrZrSiN monolithic coatings with a high Si content led to an improved thermal stability of the coatings compared to other hard coatings such as CrN and CrZrN [19]. To enhance the thermal stability of the CrAlN coating, a CrZrSiN monolithic layer with a high Si content (18 at%) was deposited between the coating and the substrate. The enhanced thermal stability of the CrZrSiN interlayer could be attributed to the outward diffusion of Si since the interlayer structurally existed below the CrAlN coatings. It is well known that the standard Gibbs free energy for oxide stability increased in the sequence of Cr₂O₃ (ΔG = −552.296 kJ·mol⁻¹ at 900 °C), SiO₂ (ΔG = −693.833 kJ·mol⁻¹ at 900 °C), Al₂O₃ (ΔG = −866.907 kJ·mol⁻¹ at 900 °C), and ZrO₂ (ΔG = −880.833 kJ·mol⁻¹ at 900 °C). Thus, the outmost surface could be mainly composed of a dense layer of Al₂O₃ and a coarse layer of ZrO₂ [26]. It might be postulated that the formation of nano-scaled Al₂O₃ dense film could block the oxygen diffusion, whereas ZrO₂ layers enable the coating to be coarse, providing a pathway for oxygen diffusion into the coating. However, more importantly, it is likely that the strong protective SiO₂ layer was competitively formed, which inhibited the diffusion of oxygen into the coatings [27]. As evidenced by EDS (see Figure 5), it was found that the oxygen content of the CrAlN coating with an interlayer containing the Si element is much lower than that of the coatings with CrN and CrZrN interlayers.

The friction coefficient values of all the CrAlN coatings with various interlayers at room temperature and at 500 °C in air are shown in Figure 6a. The friction coefficients of all the coatings at room temperature increased at 500 °C. It was noted that, despite the similar hardness after annealing at 500 °C (

Table 3. Hardness, H³/E² parameter, and H/E ratio of the CrAlN coatings with three interlayer types after an annealing test at 500 °C in air for 30 min.

Parameters	Interlayer
	CrZrN	CrN	CrN/CrZrSiN
Hardness [GPa]	35.5	35.4	35.6
H3/E2 [GPa]	0.255	0.250	0.248
H/E ratio	0.084	0.084	0.084

), the friction coefficient values of every CrAlN coating increased, and the CrAlN coating with the CrN/CrZrSiN interlayer showed the lowest increment from 0.33 to 0.42. Theoretically, it was revealed that the wear properties are closely associated with the hardness to the wear resistance of a surface. Hard coatings must have high resistance to plastic deformation as well as a low Young’s modulus (E) during a wear test. According to Johnson’s analysis, the load required to initiate plastic deformation is proportional to H³/E² when the contact event occurs [28]. Therefore, the H³/E² parameter controlling the resistance of materials to plastic deformation is closely related to the wear resistance of the coating. Moreover, the H/E ratio must be a strong indicator of a coating’s resistance to plastic deformation [25]. The wear resistance is related to the elastic strain to coating failure, i.e., the ability to recover from elastic deformation without plastic deformation. For the present study, the hardness, H³/E² parameters, and H/E ratios of the coatings with various interlayers after annealing at 500 °C for 30 min are shown in Table 3. However, there was not much difference between the values, and these results are not sufficient to explain the different friction coefficients at 500 °C. This suggested that another factor was affecting the friction behaviors of the coatings at 500 °C.

The surface morphologies and roughness (Rms, root mean square) values of all the CrAlN coatings with three interlayers after annealing at 500 °C were examined and compared to those at room temperature. The results are summarized in Figure 6b. The surface of all the coatings changed to a cone-like pebble structure at 500 °C, and the grains of the coatings were found to grow and coarsen. This phenomenon, known as a recovery process, corresponds to the deposition-induced lattice point defects caused by increased diffusivity at the annealing temperature, and this leads to grain growth. For these reasons, the surface roughness values increased compared to those at room temperature, and a morphology change was observed.

The surface roughness values of all the CrAlN coatings with various interlayers at 500 °C were examined and compared to those at room temperature. The results are summarized in Figure 6b. The surface roughness values of the CrAlN coatings with the CrZrN and CrN interlayers varied from 3.4 nm at room temperature to 25.4 nm after annealing at 500 °C. However, the surface roughness value of the annealed CrAlN coating with the CrN/CrZrSiN interlayer (7.8 nm) became only twice the value at room temperature (3.6 nm).

The reason there existed a large increase in the surface roughness of the coatings with the CrZrN and CrN interlayers could be attributed to the residual oxygen deposited into the films during the coating process. This oxygen could easily diffuse and react to form oxides with the CrZrN and CrN phases, resulting in an increased surface roughness in the coatings with the CrZrN and CrN interlayers. Conversely, the CrAlN coating with the CrN/CrZrSiN interlayer showed only a small increase in the surface roughness because the CrZrSiN interlayer, which consists of a Si_xN_y amorphous phase, is expected to work as an oxygen diffusion barrier to inhibit oxidation by the residual oxygen during the annealing process at 500 °C. These results indicated that the variation in the surface roughness of the coating due to the oxidation by the residual oxygen during the annealing process could affect the friction behaviors of the coating at 500 °C. The lowest increment in the friction coefficient variation in the CrAlN coating with the CrN/CrZrSiN interlayer, Δ0.09, could be attributed to the lowest Rms variation before and after the annealing process at 500 °C.

4. Conclusions

In this work, CrAlN coatings with three different interlayers were synthesized using a closed-field unbalanced magnetron sputtering system on WC-6 wt.% Co substrates to improve their mechanical properties. The hardness and elastic modulus of each coating showed similar values. However, the friction coefficient and adhesion strength of the CrAlN coating with the CrN/CrZrSiN interlayer showed greater values than those of the other coatings due to the smooth transition of the coating stress between the coating and the substrate via the median H/E ratio of the CrN and CrZrSiN interlayers. After the thermal stability test, the hardness value of the CrAlN coating with the CrN/CrZrSiN interlayer was approximately 28 GPa up to 1000 °C without drastic changes in hardness, and there were no significant increases in the surface roughness and friction coefficient at 500 °C. The variation in the surface roughness of the coating due to the oxidation by the residual oxygen could affect the friction behaviors of the coating, and the lowest increment in the friction coefficient variation in the CrAlN coating with the CrN/CrZrSiN interlayer could be attributed to the lowest Rms variation. These results indicated that the mechanical properties and thermal stability of the CrAlN hard coatings could be improved via optimal design and structuring interlayers. This work suggests that a CrAlN coating designed with a multi-interlayer will be a promising candidate as a protective hard coating for the cutting, milling, and machining tool industry.