Combined technology of electrospark and cathodic-arc formation of wear- and oxidation resistant coatings

Introduction

Over the past decades, considerable attention has been devoted to developing methods for extending the service life of superalloys used in high-temperature applications [1; 2], which are essential for the advancement of aerospace, chemical, and power engineering. One of the key limitations of superalloys in the manufacture of critical components is their insufficient wear and oxidation resistance, necessitating the development of effective surface modification techniques [3].

Electrospark treatment (EST) is a promising technology for strengthening and restoring working surfaces, as well as for producing coatings with enhanced wear and oxidation resistance [4–6]. An important advantage of EST is its localized action of electrical discharges, which makes it possible to treat specific areas of critical components without overheating. A distinctive feature of the technology is the need for careful selection and control of processing modes to ensure the formation of crack-free coatings.

The versatility of EST is determined by the wide range of developed electrode compositions, which allows selecting specific alloys to improve surface-sensitive properties of components. In particular, the use of fusible electrodes Al–Si, Al–Ca–Si, and Al–Ca–Mn in [7; 8] or EST of EP741NP alloy samples fabricated by selective laser melting (SLM) resulted in the formation of coatings containing intermetallic Ni_xAl_y particles synthesized during EST. This not only increased wear resistance by a factor of 4.5 through intermetallic strengthening but also reduced surface roughness to R_a = 3 µm due to the healing of surface defects typical of SLM. In addition, the in-situ synthesis of (Al, Ca)O nanoparticles during oxidative annealing of the coatings at 1000 °C provided a fourfold increase in oxidation resistance of the EP741NP alloy [8]. The oxidation rate can be further reduced [9] by improving the crack resistance and adhesion of the oxide layer to the substrate through microalloying the coatings with rare-earth (RE) metals. In this context, particular interest lies in studying the effect of Al–Ca–RE electrodes, where RE = Ce or Er. Such additives are especially relevant for coatings applied to superalloys in order to extend their operating temperature range.

A promising approach is to combine, within a single technological cycle, electrospark and cathodic-arc treatment (ESCAT) [10]. The clear advantages of computer-controlled combined processing include improved reproducibility of the technology and intensified mass transfer during cathodic arc evaporation of the electrode.

The aim of this study was to investigate the conditions for forming wear- and oxidation-resistant coatings by combining electrospark and cathodic-arc treatment (ESCAT) of the AZhK superalloy.

Methods

Rod electrodes 4 mm in diameter, made of near-eutectic Al–Ca–Ce and Al–Ca–Er alloys prepared in accordance with TU 24.45.30–042–11301236–2024, were used for the combined process of ESCAT. The electrodes were produced by a two-stage process: (1) remelting of the charge components into an ingot in a resistance furnace; (2) induction melting of the ingot followed by casting of the melt into a copper mold in a vacuum chamber filled with argon at a pressure of 0.2 atm.

Substrates were made of the AZhK nickel-based superalloy, fabricated by selective laser melting (SLM) at JSC Komposit (Russia). The chemical composition of the AZhK alloy substrates is given below [11], wt. %:

ESCAT were carried out in a single cycle using a modified cap-type UVN-2M unit (Russia), equipped with a CNC system for programmed lateral movement of the sample during processing under specified conditions. After reaching forevacuum, the vacuum chamber was filled with argon to a pressure of 20 Pa, in accordance with TI 65–11301236–2024. This pressure ensures the simultaneous initiation of an arc discharge and electrospark discharges [10]. A schematic of the ESCAT process is shown in Fig. 1.

The microstructure and composition of the samples were examined using a Hitachi S-3400N scanning electron microscope equipped with a NORAN System 7 X-ray microanalysis system (Thermo Scientific, USA). For detailed microstructural studies, cross-sectional samples were prepared using standard metallographic methods (grinding, polishing), followed by electrochemical etching at a constant voltage of 10 V in an aqueous 10 % H₂SO₄ solution.

X-ray diffraction (XRD) phase analysis was performed using diffractograms obtained on a DRON-4 diffractometer (Burevestnik, Russia) with monochromatic CuK_α radiation (1.5418 Å), and in the case of oxide layers, CoK_α radiation (1.7902 Å), in the 2θ range of 20–110°. For detailed analysis of coating substructures, a JEM-2100 transmission electron microscope (Jeol, Japan) equipped with an Oxford Instruments X-Max 80 energy-dispersive detector was employed. Lamellae were cut from the coating surface layer using a focused ion beam system (Quanta 200 3D FIB, FEI Company, USA). TEM foils were additionally thinned by ion etching on a PIPS II system (Gatan, Inc., USA). In-situ studies of structural–phase transformations in the coatings under thermal exposure at 400, 600, and 700 °C were carried out directly in the microscope column. The heating rate was 50 °C/min. Bright-field images and electron diffraction patterns were obtained with a dwell time of about 15 min at each heating step.

Mechanical properties (hardness H and elastic modulus E) of the coatings were measured on cross-sectional samples using a Nano-Hardness Tester (CSM Instruments, Switzerland) at a maximum load of 10 mN. Tribological tests were performed on a CSM Tribometer (Switzerland) in accordance with ASTM G133-22 under reciprocating sliding at room temperature in air. A 3-mm diameter 100Cr6 steel ball (analog of ShKh15) was used as the stationary counterbody. The test parameters were: track length – 4 mm, applied load – 2 N, maximum speed – 5 cm/s. Wear tracks were examined using a WYKO NT1100 optical profilometer (Veeco, USA).

Oxidation resistance tests at high temperature were carried out under cyclic heating–cooling in a SNOL 7.2/1200 muffle furnace (Lithuania), exposing the samples in air at 1000 °C. The total isothermal holding time in air was 30 h. After each cycle of “heating – isothermal hold – cooling”, the samples were weighed on an ALC-210d4 analytical balance (Acculab, USA) with an accuracy of 10^–5 g. The specific mass gain (K) was calculated as:

K = ∆m/S₀,

where ∆m is the mass difference before and after oxidation, mg; S₀ is the total surface area of the sample before testing, cm².

Results

Structure of fusible
Al–Ca–RE (Ce, Er) electrodes

Figs. 2, a and b show the microstructures of the fusible electrodes, while the diffraction spectrum (Fig. 2, c) displays the lines corresponding to their phase constituents. It can be seen that the Al–Ca–Ce electrodes have a near-eutectic structure, which is consistent with the results of [12]. As shown in Table 1, this structure consists of fine two-phase eutectics e₁ [(Al) + CaAl₄] and three-phase eutectics E₁ [(Al) + CaAl₄ + (Ca,Ce)Al₄], as well as dendrites of a solid solution based on aluminum (Al).

As shown in Fig. 2, b, the addition of Er to the Al–Ca–Er electrode promotes the formation of bright, faceted primary Al₃(Er,Ca) crystals. Consequently, the principal structural constituents of the Al–Ca–Er electrode are fine two-phase eutectics [(Al) + CaAl₄], three-phase eutectics [(Al) + CaAl₄ + (Er,Ca)Al₃], primary Al₃(Er,Ca) crystals, and (Al) dendrites. The elemental distribution among the structural constituents of the Al–Ca–Ce and Al–Ca–Er electrodes is summarized in Table 1.

Kinetics of coating formation
and structure

The kinetic curves of mass transfer as a function of electrode polarity are shown in Fig. 3. The maximum specific mass gain of the substrate, corresponding to the greatest coating thickness, is observed after 5 min of treatment, regardless of electrode polarity. However, the specific erosion after 5 min of treatment under cathodic electrode connection (ΔА₅ = −94.9·10⁻⁴ g) is an order of magnitude higher than that under anodic polarity (ΔА₅ = −7.1·10⁻⁴ g). The mass gain curve obtained under anodic polarity (ACe) indicates a low contribution of mass transfer (ΔK₅ = 4.8·10⁻⁴ g). In this case, an increase in mass is observed only after 3 min of treatment.

On the surface of the ACe series coatings, elongated solidified droplets can be observed (Fig. 3, c), some of which are marked with yellow arrows. Their formation under anodic polarity can be attributed to local melting of fusible structural constituents within cathode spot [12]. In addition, oxide particles approximately 15 µm in size were found on the surface of these coatings (highlighted with orange arrows). According to EDS data, these particles, in addition to oxygen, contain high concentrations of Al (28–33 at. %) and Ca (6–10 at. %). They also contain about 1 at. % Ce and Er, which is expected given their high affinity for oxygen. Under cathodic polarity, the surfaces of the coatings (KEr and KCe) exhibit overlapping solidified melt droplets, but they are less homogeneous in composition. It is worth noting that, unlike the A-series, cracks were observed on the surfaces of the K-series coatings.

XRD patterns (Fig. 4, a) of the initial and ESCAT -treated substrates revealed differences in the phase composition of the coatings depending on electrode polarity. Under cathodic polarity (sample KCe), the coating consists of β-NiAl, which has an ordered cubic B2 crystal structure, and γ′-Ni₃(AlCr) with an ordered L1₂ superstructure. The β-phase fraction is dominant (Table 2), amounting to 84.6 wt. %.

Anodic polarity (sample ACe) results in the formation of a coating predominantly composed of γ′-Ni₃(AlCr) (97.6 wt. %). Diffraction peaks of the β-phase were not detected in the diffractogram. This indicates an insufficient amount of free Al in the EST melt to form the equiatomic NiAl intermetallic.

As shown in Figs. 4, b and c, typical cross-sectional images of the coatings are presented. The ACe and AEr coatings do not display a distinct interface with the substrate, which can be attributed to their similar elemental concentrations (Table 3). According to EDS data (regions 1 and 3), however, the Al content in the coatings (15 at. %) is nearly twice that of the AZhK alloy (9 at. %). No Ca was detected in the bulk of the coating, and the Ce content did not exceed 0.2 at. %. The coating structure nevertheless contains small black inclusions identified by EDS as (AlCaCe)O (Table 3, region 2). Larger particles of the same composition were also found on the surfaces of these coatings (see Fig. 3, c).

The thickness of the coatings formed under cathodic polarity (18–20 µm) is greater than that of the anodic ACe and AEr coatings (15 µm). The microstructure of the cathodic coatings (Fig. 4, c) exhibits pronounced differences due to the increased incorporation of elements from the Al–Ca–RE electrodes. As shown in Table 3, the Al concentration in the K-series coatings reaches 74 at. %, while the Ni content decreases to 15 at. %. A distinct interface between the AZhK substrate and the coatings is evident; it is heterogeneous in both structure and composition, with the Ni concentration in this region at about 30 at. %. It should also be noted that under cathodic polarity, Ca and RE metals are distributed more uniformly throughout the coatings compared with the ACe and AEr samples, although the total RE metal content does not exceed 0.6 at. %.

The microstructure of the A-series coatings (Figs. 5, a and b) consists of uniform fine columnar crystals, which differ entirely from the microstructures of both the electrodes and the substrate. Yellow dashed lines indicate the boundaries of solidified melt droplets. The crystallite orientation coincides with the growth direction of the AZhK alloy, which is also reflected in the increased intensity of the 200 γ′ peak at 2θ ≈ 50° (Fig. 4, a). During crystallization of the melt droplets, the columnar crystallites grow from the interface toward the surface.

The coatings formed under cathodic polarity (Figs. 5, c and d) exhibit a different structure, characterized by a less pronounced metallographic texture. Compared with the substrate, these coatings are less susceptible to chemical etching, which indicates higher corrosion resistance.

To investigate the structure of crack-free coatings obtained under anodic polarity in greater detail, the fine structure of the ACe sample was analyzed. A bright-field TEM image of a lamella cut from its surface is shown in Fig. 6, a (see also Fig. 3, c). The coating consists of columnar crystals with a strongly oriented growth direction along the [01–1] zone axis. An oxide particle located above the columnar crystals on the coating surface exhibits an amorphous structure, confirmed by electron diffraction showing an amorphous halo (inset in Fig. 6, a).

The columnar crystals, ~300 nm in cross-section, consist of the γ′ phase with an ordered L1₂ crystal structure, as confirmed by electron diffraction. Analysis of the corresponding diffraction pattern (Fig. 6, d) revealed displacements of atomic planes, which may result from dislocation motion in the γ′ matrix.

Spherical nanoparticles up to 30 nm in size are mainly located along the boundaries of the columnar crystals (Fig. 6, c). Analysis of these particles (Figs. 6, b and g) showed that they are enriched in calcium and oxygen, while the concentrations of all other elements are significantly reduced. According to Fig. 6, d, these oxides have an orthorhombic lattice with a CaO-type structure.

In addition, the γ′ grain boundaries contain a highly distorted secondary phase (Fig. 6, b). EDS analysis determined its composition as Ni_53.21Al₂₀Cr_15.85Co_6.5Mo_4.45. The large interplanar spacing (d = 5.7 Å), obtained from the Fourier transform (Fig. 6, e), suggests that the diffraction originates from the [10–10] plane of a hexagonal Laves phase (AB₂) of the C36 structure type. The calculated lattice parameters for this phase are а = 6.63 Å and с = 11.05 Å. Based on the elemental ratio, it can be assumed that the Laves phase has the composition (Ni,Co)₂(Al,Cr,Mo).

Fig. 7 shows the overall appearance of the lamella and enlarged images of the selected region before and after vacuum annealing at 700 °C. A distinct change in contrast in some areas indicates relaxation of internal stresses (Fig. 7, b). The series of electron diffraction patterns presented in Fig. 7, c, obtained from γ′ columnar crystals during lamella heating up to 700 °C, confirm the high thermal stability of this phase. At the same time, at 600 °C nanocrystalline particles appeared in the amorphous (AlCaCe)O particle located on the coating surface (Fig. 7, d). Their interplanar spacings (3.05 Å, 2.90 Å) correspond to a monoclinic CaAl₂O₄ mixed oxide.

Mechanical and tribological properties of the coatings

The results of instrumented nanoindentation (Fig. 8, a) showed that ESCAT increases the hardness and decreases the Young’s modulus (E) of the AZhK alloy substrate. The presence of Er in the coatings has a more pronounced effect on hardness compared with Ce alloying of the electrode. The maximum hardness values (12.3 ± 0.3 and 10.2 ± 0.3 GPa, respectively) were recorded when cathodic polarity was applied. Fig. 8, b also presents the H/E ratio, which is commonly used as an indicator of the damage tolerance of coatings. The coatings exhibit relatively low Young’s modulus values (E ≤ 160 GPa), which is atypical for intermetallics: 178 GPa for Ni₃Al and 284 GPa for NiAl [14].

As shown in Figs. 8, b and 9, b, ESCAT significantly improves the wear resistance of the nickel alloy, with the effect being stronger under cathodic polarity (up to a sixfold increase) compared with anodic polarity (up to a twofold increase). Erbium in the coating has a stronger influence on enhancing these properties. The effects of RE additions and electrode polarity on the wear resistance of the coatings correlate with the formation of a dual-intermetallic structure. In particular, coatings with a (β-NiAl + γ′-Ni₃Al) structure, which exhibit the highest hardness, are characterized by excellent wear resistance, with values ranging from 6.0 to 7.5·10^–5 mm³/(N·m).

The friction coefficient versus cycle number curves shown in Fig. 9, a demonstrate the higher amplitude of friction fluctuations in contact with the steel ball for both the untreated AZhK alloy and after anodic ESCAT. The average friction coefficient values range from 0.45 to 0.35, while a further reduction to ~0.3 is observed for coatings produced under cathodic polarity (K-series).

To investigate the wear mechanisms in greater detail, SEM images of the wear tracks were obtained (Fig. 10), and EDS analysis was performed. The results suggest two dominant wear mechanisms, which correlate with the coating structures determined by electrode polarity during ESCAT. In the case of A-series coatings, ball sliding is accompanied by the formation of deep grooves and numerous ridges, associated with local plastic deformation (Figs. 10, a and b).

In contrast, the wear track surfaces of the K-series coatings display a different morphology, characterized by compacted wear debris formed under load during testing (Figs. 10, c and d). According to EDS data, the bottoms of the wear tracks are depleted in aluminum (28.2 at. % Al and 47.9 at. % Ni). At the same time, the wear debris consists mainly of Al₂O₃, formed by selective oxidation of aluminum. Thus, wear of the K-series coatings is primarily governed by aluminum oxidation and compaction of the oxidized debris, processes which together promote the formation of a protective tribolayer.

High-temperature
oxidation resistance of the coatings

Fig. 11 and Table 4 show the results of qualitative and quantitative phase analysis of the samples after 30 h of oxidative annealing at 1000 °C. The XRD pattern of the AZhK alloy exhibits diffraction peaks corresponding to Cr₂O₃ and spinel phases NiCr₂O₄, NiMoO₄, and NiNb₂O₆.

Although the ESCAT coatings show a similar qualitative phase composition after long-term oxidative annealing, the quantitative phase ratios differ markedly (Table 4) and are dependent on electrode polarity. In the annealed coated samples, a large fraction of the γ′ phase is retained (30–72 wt. %). This suggests that the oxide layers formed on the coating surfaces are relatively thin and remain fully penetrable by X-rays under Bragg–Brentano geometry. Thus, after isothermal holding of the ACe and AEr coatings, the surface layer is dominated by α-Al₂O₃ and NiAl₂O₄ phases, while CaMoO₄ is detected only in small amounts (up to 3 wt. %). In contrast, oxidation of the KCe and KEr coatings results in an increase in CaMoO₄ content up to 15 wt. %, while the combined content of α-Al₂O₃ and NiAl₂O₄ reaches 25–56 wt. %.

Cross-sectional images (Fig. 12) of the oxidized samples confirm the XRD results. The total thickness of the oxide layers formed on the coatings reaches 2.5 µm, which is up to 10 times thinner than the oxide layer on the AZhK alloy (~25 µm). The minimum oxide layer thickness was recorded for coatings produced under anodic polarity.

The oxide layers exhibit a bilayer structure. The outer continuous NiAl₂O₄ layer has a spinel structure and contains finely dispersed CaMoO₄ particles. Beneath it lies a layer of α-Al₂O₃, whose boundaries (highlighted with blue arrows in the inset of Fig. 12, b) are enriched in Er and Ce. According to EDS (region 1, Table 5), the oxygen-free zone of the coating is depleted in Al (7.8 at. %). These results indicate that beneath the oxide layers there is a zone composed not of pure γ′ phase, as identified by XRD, but of γ′ + γ grains.

The oxide layers of the coatings produced under cathodic polarity (Figs. 12, c and d) also show a heterogeneous structure. CaMoO₄ grains are found on the surface, beneath which lies a continuous NiAl₂O₄ + α-Al₂O₃ layer, also containing dispersed CaMoO₄ particles. The oxygen-free zone of these coatings consists of γ′ grains, indicating a β → γ′ phase transformation driven by the interaction of aluminum with oxygen. Additionally, agglomerates of particles are present at the coating–substrate interface, a feature not observed in the A-series coatings. According to EDS (region 6, Table 5), these particles contain refractory elements consistent with topologically close-packed (TCP) phases.

Fig. 13 shows the oxidation kinetics curves of the untreated and EST-treated samples during 30 h of isothermal exposure at 1000 °C. For comparison, the oxidation kinetics of a coating obtained under anodic polarity of ESCAT with an Al–7 at. % Ca–1 at. % Mn electrode (sample AMn) are also included. In Fig. 13, a, the experimental data are shown by solid lines, while the dashed lines represent approximations calculated by an empirical method using the equations given in Table 6. The relationship between mass gain and oxidation time for the AZhK alloy follows a linear law, as confirmed by approximation coefficients close to unity. For this sample (Fig. 14, b), after 30 h of isothermal exposure, a degraded green-colored surface is observed, which is most likely due to the formation of nickel or chromium oxides.

The oxidation kinetics curves confirm (Fig. 13) that the use of combined ESCAT with Al–Ca-based electrodes significantly improves the oxidation resistance of the AZhK superalloy at 1000 °C. While the mass gain of the AMn coating follows a parabolic law (Table 6), the oxidation of RE-containing coatings follows a logarithmic law. Overall, the oxidation curves of the coatings can be divided into two stages: an initial stage (0–5 h) and a steady stage (5–30 h). Despite the parabolic kinetics of the initial oxidation stage of the AMn sample, it shows the lowest slope of the curve (Fig. 13, b), indicating the slowest oxidation rate at this stage. In contrast, other samples with logarithmic oxidation show rapid mass gain during the initial stage, but in the steady stage the rate slows down as the oxide layer thickens. Thus, the transition from parabolic to logarithmic behavior is explained by the retardation of oxygen diffusion through the already formed oxide layer.

Discussion

The role of rare-earth metals in both coatings and nickel-based superalloys has been widely investigated [9; 16; 17]. However, a unified theory explaining the beneficial effects of such alloying has not yet been established. Alloying nickel-based superalloys with RE metals (Y, Ce, La, Hf, and Er), as well as applying RE-modified coatings, reduces oxidation rates by improving the crack resistance and adhesion of the oxide layer to the substrate. Small RE metals additions accelerate the nucleation of chromium and aluminum oxides [18], thereby promoting rapid formation of a protective layer. Owing to their large atomic radii and high chemical activity, RE form dispersed phases in coatings (e.g., Ni_xAl_y(Hf,Zr)_z [19], LaCrO₃ [20]) that hinder grain boundary diffusion of cations (Al³⁺, Cr³⁺, Ni²⁺, Ta⁵⁺ [21; 22]). For instance, [23] reported that the presence of Ce (0.038 wt. %) increases aluminum activity in the IC21 superalloy with an Al–Si coating after oxidative annealing at 1150 °C, reducing the Al content required for the formation of a protective α-Al₂O₃ layer. Ce also retards the polymorphic transformation of metastable θ-Al₂O₃ into stable α-Al₂O₃, thereby suppressing volumetric shrinkage of the oxide layer [24] and lowering the likelihood of crack formation.

Excessive RE additions, however, coarsen the microstructure and degrade the mechanical properties of alloys [25], and may also lead to degradation of the oxide scale. Therefore, electrode compositions were selected to be close to the eutectic point on the phase diagram, allowing the formation of a fluid reactive melt on the metal surface during ESCAT that actively interacts with substrate elements. Although the RE content in the coatings does not exceed 0.5 at. %, this level is sufficient to significantly influence their properties.

The polarity of the electrodes during ESCAT governs the structural and phase transformations in the coatings. Since this processing is carried out in vacuum [10; 12], spark breakdown during electrode retraction is accompanied not only by energy release at the anode, but also by pulsed arc evaporation of the cathode (Fig. 15). Under cathodic polarity, additional electrode sputtering increases the substrate mass gain and coating thickness, whereas anodic polarity leads to partial evaporation of the coating surface.

β-NiAl-based coatings can only be obtained under cathodic polarity, because intensive electrode evaporation enriches the coating with aluminum. Such coatings, however, are prone to cracking due to the mismatch between the thermal expansion coefficients of β-NiAl (15·10⁻⁶ K^–1 [26]) and Ni-based superalloys (12·10⁻⁶ K⁻¹ [27]). They also exhibit a weaker crystallographic texture (Figs. 6, c and d), resulting from stronger substrate heating.

Anodic polarity promotes the formation of γ′-Ni₃(AlCr)-based coatings. Minimal substrate heating under these conditions enables efficient heat dissipation into the bulk [28; 29], supporting uniform grain growth with a dominant <100> crystallographic orientation. The reduced cross-section of the columnar crystallites compared with the AZhK alloy (inset in Fig. 5, a) is associated with the ultrahigh solidification rates typical of ESCAT [30; 31], which overall enhances hardness and wear resistance. In addition, in-situ formation of (Al–Ca–RE)O nanoparticles along grain boundaries restricts dislocation motion by generating dislocation loops. The strengthening role of oxide inclusions has also been noted in [32].

Regardless of electrode polarity, oxidation of the coatings follows a logarithmic law: a rapid mass gain during the initial stage is followed by a slower increase as the oxide layer thickens. The high oxidation rate at the early stage is associated with selective oxidation of Al, accompanied by Al₂O₃ formation (–890.4 kJ/mol) (Fig. 14, a). The presence of (Al–Ca–RE)O nanoparticles appears to accelerate Al₂O₃ formation. The solubility of RE in Al₂O₃ or Cr₂O₃ is very low [19], so their oxides segregate at the metal–oxide interface (Fig. 13, b). Furthermore, as shown in Fig. 7, vacuum annealing of the coatings results in the crystallization of CaAl₂O₄ nanoparticles, which are isostructural with alumina. Owing to their lower energy, these particles accelerate Al₂O₃ grain growth. At the boundaries of the thin α-Al₂O₃ layer, RE-enriched particles improve oxide adhesion and reduce stress concentrations, thereby preventing crack formation.

Prolonged isothermal exposure of ACe and AEr coatings in oxygen-containing environments leads to the formation of NiAl₂O₄ (–12.5 kJ/mol) through Ni diffusion across the thin Al₂O₃ film to the surface. Thus, in these coatings oxygen diffusion is strongly restricted by the heterogeneous NiAl₂O₄/α-Al₂O₃ layer, reinforced by CaMoO₄ particles. This interpretation is consistent with the low oxidation rate constants (K_p = 0.16–0.22) compared with those of the KCe and KEr coatings (Table 6).

Despite the presence of cracks, the KCe and KEr coatings exhibit lower mass gain than the AZhK alloy. Oxygen diffusion along cracks occurs during the first ~15 h of oxidation, after which the growth of CaMoO₄ grains within the cracks blocks further diffusion. The cracks thus act as channels for upward Mo diffusion, necessary for CaMoO₄ grain nucleation and subsequent coalescence into a continuous layer. According to [8], these grains with the tetragonal scheelite structure I4₁/a exhibit excellent thermal stability and ultralow thermal conductivity at 400–1000 K (0.6–1.2 W/(m·K)), values even lower than those of thermal barrier coatings such as YSZ (1.5–3.0 W/(m·K)).

Conclusion

The formation behavior of wear- and oxidation-resistant coatings produced by combined electrospark and cathodic arc treatment (ESCAT) on AZhK nickel-based superalloy samples was investigated. Optimal ESCAT modes with Al–Ca–Er and Al–Ca–Ce electrodes were identified, enabling the formation of 15–20 µm thick coatings composed of columnar crystallites with a cross-section below 300 nm, containing γ′-Ni₃Al and β-NiAl phases. These coatings increased substrate hardness from 5.2 to 12.3 GPa and improved wear resistance sixfold. The use of Al–Ca–Er electrodes provided enhanced oxidation resistance of the AZhK alloy at 1000 °C by changing the oxidation law from linear to logarithmic one due to the formation of a protective NiAl₂O₄/α-Al₂O₃ layer ~3 µm thick, reinforced by CaMoO₄ particles.