Studies on the Morphological and Mechanical Properties of TiZrHfNiCuCo Metallic Coatings Deposited by DC Magnetron Sputtering

Abstract

This research aimed to investigate the effects of input power and gas flow rate on the composition, microstructure, and mechanical properties of TiZrHfNiCuCo metallic coatings. These metallic coatings were deposited on a p-type Si wafer using a direct current magnetron sputtering system with varying input powers (100–300 W) and Ar flow rates (5–20 sccm). It was observed that increasing input power and decreasing Ar flow rate led to TiZrHfNiCuCo metallic coatings with higher hardness and smoother surfaces. The experiment resulted in the formation of a metallic amorphous coating. This study describes the mechanism by which the mechanical properties of the TiZrHfNiCuCo coating change according to sputtering parameters. Based on these results, the effects of sputter variables, such as input power and gas flow rate properties, on the properties of coatings are discussed.

1. Introduction

High-entropy alloys (HEAs) represent a novel approach in alloy design, distinct from traditional alloys, by incorporating multi-principal elements. Typically composed of five or more elements in nearly equiatomic ratios, HEAs were first introduced by Cantor et al. [1] and Yeh et al. [2]. Remarkably, even with a high concentration of many elements, HEAs predominantly form simple phases, such as solid solutions and nanocrystalline or occasionally amorphous alloys, rather than complex intermetallic compounds. This phenomenon results in lower free energy and enhanced phase stability [3,4]. The variation in atomic sizes among the constituent elements generally leads to lattice distortion and sluggish diffusion [4,5,6]. Consequently, HEAs exhibit exceptional properties, including superior hardness, oxidation resistance, irradiation resistance, good corrosion resistance, and wear resistance [7,8,9,10,11,12,13,14,15]. These characteristics have spurred extensive research into their applications across diverse fields, such as structural materials, hard coatings, diffusion barriers, and energy materials [10,12,15,16,17,18,19].

Enhancing the durability of structural components can be effectively achieved through surface coating, which modifies and strengthens their mechanical and functional properties. In addition to their use as bulk structural materials, HEAs have become attractive options for coatings, thanks to their superior properties [6,16,17,18,19,20,21]. When deposited as layers on substrate surfaces, HEA coatings can significantly improve the mechanical performance of the surface, provided that the compositions and deposition parameters are appropriately selected.

In the surface coating technology, crystalline growth and even nuclear growth are inhibited, so they can be more easily reflected in the formation of an amorphous phase [2]. One of the widely used techniques for forming HEA coatings is sputtering deposition [22,23,24].

Sputtering deposition is recognized for its ability to easily control the stoichiometry of coatings by adjusting the chemical composition of the target. It also enhances the quality of coatings by regulating variables such as input power, gas flow rate, reactive gases, substrate temperature, and deposition time [25,26,27,28,29]. In addition, the atoms with momentum can quench condensation rapidly for coatings, hence minimizing concentration fluctuations in the coatings [19]. Due to the special ‘rapid quenching’ effect of the preparation method of the HEA coatings, the FCC and BCC solid-solution phases or amorphous phases are commonly fabricated [19]. Moreover, the grain size of the HEA coating is usually restrained by their thickness, so the grain size of the HEA coating is generally smaller than the HEA bulk counterparts, which results in the grain refinement strengthening, which can result in the strengthening of the mechanical properties [19,30].

In our previous study, we developed M-containing HEAs using an equiatomic substitution strategy (M=Co, Nb and Fe) [20,31]. Additionally, we achieved TiZrHfNiCuCo HEA nitride coatings with a nano-composite structure using direct current magnetron sputtering according to various deposition parameters [25]. It was found that the HEA metallic coating exhibits an amorphous structure, whereas the addition of N₂ reactive gas leads to the formation of a single-phase FCC-type nanocrystalline structure in the HEA nitride coatings. Based on these results, the present study focuses on the deposition of HEA metallic coatings without the incorporation of reactive gases, such as N or O. Additionally, this study aims to expand the scope of research on TiZrHfNiCo HEA coatings by investigating the effects of magnetron sputtering parameters on the microstructure and properties.

2. Materials and Methods

The TiZrHfNiCuCo metallic coatings were deposited on p-type Si (100) single-side polished wafer substrates with resistivities of 1–10 ohm-cm using a direct current magnetron sputtering system (KCMC, Suwon-si, Republic of Korea) under high vacuum at room temperature. The Si wafer was chosen for the investigation of the morphological and mechanical properties of the metallic coatings due to its thin thickness and flat surface. The TiZrHfNiCuCo (Plansee SE, 16.6/16.6/16.6/16.6/16.6/16.7 at.%) powder target was prepared using high-purity metallic powders of Ti, Zr, Hf, Ni, Cu, and Co. The processes involved weighing, mixing, and finally cold pressing to achieve a diameter of 50.8 mm and a thickness of 6.35 mm. As a preliminary step, the p-type Si (100) substrates were ultrasonically rinsed with acetone, ethanol, and distilled water. The rotational speed of the substrate was fixed at 10 RPM to ensure the uniform deposition of the coating layer, and the initial chamber pressure reached below $8\times {10}^{-3}$ Pa before proceeding to the next critical phase of argon injection. Afterwards, the flow rate of high-purity argon gas (99.999%) was controlled using a mass flow controller (MFC) (Gencoa, Liverpool, UK). The applied power was controlled from 100 W to 300 W at a fixed Ar flow rate of 20 sccm. Then, the Ar flow rate was controlled from 5 sccm to 20 sccm at a fixed applied power of 300 W. In order to minimize any contamination from the target surface and ensure a reproducible target surface state, the pre-sputtering was conducted for 10 minutes at 300 W in 20 sccm before starting the deposition process. Also, specific deposition conditions, including the applied power, deposition time, and reactive gas flow rate, are listed in

Table 1. The deposition conditions of the TiZrHfNiCuCo metallic coatings.

Sample Number	Substrate Temperature	Rotation Speed (RPM)	DC Power (W)	Deposition Time (min)	Ar Flow Ratio (sccm)
1	RT	10	100	30	20
2			200	30	20
3			300	30	20
4			300	30	15
5			300	30	10
6			300	30	5

.

An X-ray diffractometer (PANalytical/Empyreal/PC, Malvern, UK) with Cu Kα1 radiation ($\mathsf{\lambda }=1.5406\mathrm{Å})$ (Malvern Panalytical Ltd., Malvern, UK) and a transmission electron microscope (TEM, Tecnai F20 G2, FEI) were used for phase analysis. When measuring XRD, the scan range was from 2$0°$ to $80°$. The specimen for TEM analysis was prepared via focused ion-beam thinning (FIB, Helios Nanolab 600, FEI). The thickness, microstructure, and chemical composition of the coating were analyzed using a field-emission scanning electron microscope (FE-SEM, SU-8010, Hitachi, Tokyo, Japan) with an acceleration voltage of 15 keV. Analysis of the chemical composition was conducted using an energy-dispersive X-ray spectroscope (EDS) with an acceleration voltage of 20 keV.

The surface roughness of the coatings was analyzed using the atomic force microscope (AFM, hpAFM, Nanomagnetics Instruments, Oxford, UK) in contact mode with a standard cantilever tip with a radius curvature of less than 50 nm.

Hardness measurements were conducted on a nano-indentation (NHT-X, CSM, Needham, MA, USA) system equipped with a Berkovich diamond tip. The loading window, ranging from 5 to 10mN, depending on the coating thickness, was applied. Load displacement curves were evaluated according to Oliver and Pharr [32]. The analysis process was performed several times for each sample for the reliability of the results.

3. Results and Discussion

The analyses results of the process variables of DC power at 100 W, 200 W, and 300 W are shown in Figure 1 and Figure 2. Figure 1 indicates that the deposition rates ranged from 100 W to 300 W, with a constant Ar flow rate of 20 sccm. The deposition rate of TiZrHfNiCuCo coatings was calculated by dividing the thickness of the coating by the deposition time. The deposition time was 30 min, and coatings with thicknesses of approximately 472 nm, 853 nm, and 1230 nm were achieved at input powers of 100 W, 200 W, and 300 W, respectively; the deposition rates were thus about 15.7 nm/min, 28.4 nm/min, and 41 nm/min.

As input power increases while maintain a constant deposition time and Ar flow rate, the deposition rate also increases. Controlling input power affects the microstructure, particle size, and deposition structure. At low power, deposited particles act as seeds or centers of nuclei, leading to an increase in particle size, which results in a rougher surface and slower deposition rate. At a higher input power, however, target particles are physically ejected from the target and deposited quickly on the substrate, due to the acceleration of the Ar ions with the rising potential as the input power increases, resulting in faster deposition [33,34,35,36]. This phenomenon is demonstrated by the flow of the ionized particle (J_ion) in the Langmuir–Child relationship and the average kinetic energy (KE_av) equation in the glow discharge [37,38]. According to the Langmuir–Child relationship and the average kinetic energy equation for DC glow discharge, the flow of the ionized particle (J_ion) and the average kinetic energy (KE_av) impacting the substrate are dependent on the DC applied voltage. Consequently, raising the bias voltage leads to a substantial increase in the Ar ion current bombardment and the kinetic energy of the ions striking the coating surface. This enhanced bombardment significantly boosts the surface mobility of adatoms by increasing momentum transfer through collisions. These processes are directly proportional to the DC applied voltage and input power, thereby contributing to the observed increase in the sputter deposition rate [39,40].

J_ion ∝ Voltage^3/2,

(1)

KE_av ∝ Voltage,

(2)

Figure 2 displays the XRD patterns of the as-deposited TiZrHfNiCuCo coatings on Si (100) substrates with varying sputtering input powers. The XRD result of the Si substrate is also added because of the thin thickness of the film inducing the inclusion of the structural information of the Si substrate. The difference between the diffraction peak of each coating and that of the substrate is the formation of a halo pattern in the metallic coating, centered at approximately 39$°$. The XRD patterns suggest that each coating has an amorphous phase. Additionally, the intensity of the halo pattern increases with higher input power, attributed to the increased diffraction intensity due to the growth of the metallic coating included in the diffraction process as the metallic coating thickness increases [41].

Figure 3 shows the cross-sectional images of the metallic coatings deposited under Ar flow rates ranging from 5 sccm to 20 sccm while maintaining an input power of 300 W.

In the absence of additional energy inputs, such as cooling rate, temperature gradient during sputtering, or post-sputtering heat treatment, vertical growth generates a columnar structure in the metallic coatings. Figure 3d clearly shows the metallic coating exhibiting a columnar structure. At higher Ar flow rates, the interactions between the sputtered atoms and the Ar ions decrease the kinetic energy of the sputtered atoms, leading to the relative creation of porous films [42]. The less distinct columnar structures in Figure 3a to Figure 3c are likely due to the increased atomic peening effect caused by the increased mean free path resulting from reduced working pressure. Therefore, the cross-sectional images of the metallic coatings can be used to compare the densities of the metallic coatings.

Figure 4 shows the XRD patterns of metallic coatings deposited at Ar flow rates ranging from 5 sccm to 20 sccm, with a constant input power of 300 W. Similar to the XRD results for varying input powers, a halo pattern around 39$°$ indicates that the TiZrHfNiCuCo metallic coating exhibits a distinct tendency to adopt an amorphous structure under conditions of varying input powers and Ar flow rates.

The formation of this amorphous phase in the metallic coating is generally attributed to the effects of high mixing entropy and kinetic factors in high-entropy coatings. The high mixing entropy enhances mutual solubility among the constituent elements, thereby inhibiting the development of a crystalline structure. Additionally, the rapid quenching during sputtering deposition significantly limits chemical fluctuations and suppresses crystallization [43]. For the detailed microstructure characterization, TEM analysis was performed on the metallic coating deposited at an input power of 300 W and an Ar flow rate of 5 sccm.

The TEM images and the selected area diffraction (SAED) pattern of the metallic coating are shown in Figure 5. Considering the possibility of phase transformation as the metallic coating was deposited, the analysis was conducted in two areas. In both areas, the SAED patterns showed halo ring patterns, confirming the homogeneous amorphous phase throughout the coatings. This result strongly supports the XRD analysis. Additionally, many HEA metallic films, such as AlCrMoSiTi [44], TiAlCrSiV [45], and HfNbTaTiZr [46,47], are known to have amorphous phases.

The chemical compositions (at. %) of the HEA metallic coatings were determined by EDS and are shown in Figure 6 and

Table 2. The chemical composition of the TiZrHfNiCuCo metallic coatings deposited under each condition.

Experimental Condition	Elements (at%)
	Ti	Zr	Hf	Ni	Cu	Co	O
100 W, 20 sccm	11.13 ± 0.018	12.90 ± 0.020	14.50 ± 0.020	12.78 ± 0.051	16.73 ± 0.051	13.25 ± 0.036	18.73 ± 0.085
200 W, 20 sccm	12.78 ± 0.131	14.25 ± 0.104	16.80 ± 0.041	13.95 ± 0.247	17.15 ± 0.189	14.68 ± 0.175	10.38 ± 0.333
300 W, 20 sccm	12.88 ± 0.111	14.40 ± 0.082	16.70 ± 0.071	14.10 ± 0.183	18.03 ± 0.063	14.93 ± 0.095	9.03 ± 0.309
300 W, 15 sccm	15.00 ± 0.147	16.05 ± 0.104	16.03 ± 0.095	15.40 ± 0.135	15.93 ± 0.214	15.78 ± 0.048	5.90 ± 0.502
300 W, 10 sccm	14.98 ± 0.048	16.20 ± 0.071	16.13 ± 0.085	15.73 ± 0.103	15.50 ± 0.168	16.05 ± 0.189	5.40 ± 0.518
300 W, 05 sccm	15.15 ± 0.194	15.65 ± 0.104	15.78 ± 0.165	16.13 ± 0.103	15.83 ± 0.085	16.70 ± 0.091	4.73 ± 0.295

. For quantitative analysis, a scan area size of 2.5 μm was selected, and at least five different points were measured for statistical reliability. The composition of the specimens deposited at low input power (100 W and 200 W) and at 300 W with a gas flow rate of 20 sccm differed from the composition of the sputtering target, due to the varying degrees of scattering experienced by sputtered atoms with different masses during the sputtering process [35]. Firstly, the coating layers deposited at a low sputtering power, such as 100W or 200W, were not sufficiently thick and exhibited slight deviations from the stoichiometric ratio of the target material. In contrast, the composition of TiZrHfNiCuCo nitride coatings deposited at a low input power showed a significant difference, compared to the target composition, particularly for heavy atoms like Hf, which are scarcely detectable [25]. Nitride coatings that incorporate reactive gases showed more sensitive results to differences in input energy, compared to metallic coatings. This can be explained by the lower sputtering efficiency of reactive gases, such as N₂, compared to inert Ar ions [48]. Secondly, the metallic coating deposited at 300W with a higher gas flow rate (20 sccm) also showed slight deviations from the equiatomic ratio. Conversely, the composition of the metallic coatings closely matched the sputtering target as the gas flow rate decreased. This was because a lower working pressure and a longer mean free path allowed sputtered atoms to reach the substrate with minimal scattering caused by collisions between the sputtered atoms and the gas.

Additionally, it was observed that the oxygen content decreased as the input power increased and the gas flow rate decreased. The oxygen content was related to the density of the metallic coating. If the metallic coating is porous, more oxygen can be detected during component analysis, due to greater oxygen intrusion into the lattice spaces [35,39]. As previously described, denser coatings were produced with a higher input power and lower gas flow rates [35,39,49]. Thus, the density of coatings and the oxygen content are interrelated. In this study, the oxygen content was expected to affect the coating quality. Therefore, to optimize the microstructure and properties of various high-entropy coatings, it is essential to conduct systematic studies on the effects of process parameters.

The results of surface roughness measurements using AFM under various conditions are illustrated in Figure 7 and

Table 3. The roughness of the TiZrHfNiCuCo metallic coatings deposited under each condition.

Condition	100 W, 20 sccm	200 W, 20 sccm	300 W, 20 sccm	300 W, 15 sccm	300 W, 10 sccm	300 W, 05 sccm
Roughness (Ra)	4.333 ± 0.609	1.990 ± 0.229	1.425 ± 0.115	0.870 ± 0.024	0.645 ± 0.018	0.603 ± 0.006

. The surface images of the coatings under several conditions are shown in Figure 7a–c. Figure 7d demonstrates the trend of surface roughness changes under each condition, and Table 3 quantitatively presents the roughness measurements. In Table 3, extremely smooth surfaces with Ra lower than 1 nm were observed in coatings fabricated with an Ar flow rate below 20 sccm. At low Ar flow rates, the metallic coating structure was dense and smooth, with low surface roughness. As the Ar flow rate increased, the metallic coating exhibited a clear columnar structure, but the surface became highly rougher. In addition, surface roughness increased as the input power decreased. A rougher surface indicates a more porous morphology, which is closely related to coating quality. Therefore, the roughness variations according to the sputtering process variable indicate that as the momentum of the emitted sputtering atoms decreases, a rough and porous coating is formed [35,39].

The influence of the Ar flow rate and input power on the hardness and Vickers hardness of the metallic coatings is illustrated in

Table 4. The hardness and Vickers hardness of TiZrHfNiCuCo metallic coatings and the TiZrHfNiCuCo melt-spun glassy ribbon.

		Mechanical Properties
Method	Condition	HIT (GPa)	HVIT (HV)
Magnetron Sputtering	100 W, 20 sccm	5.604 ± 0.055	519.0 ± 5.46
	200 W, 20 sccm	6.781 ± 0.124	628.0 ± 11.52
	300 W, 20 sccm	9.04 ± 0.214	837.4 ± 19.86
	300 W, 15 sccm	10.78 ± 0.180	998.2 ± 16.73
	300 W, 10 sccm	11.56 ± 0.247	1070 ± 22.84
	300 W, 05 sccm	11.64 ± 0.236	1078 ± 21.89
Melt-spinning	Cu wheel with a rotating speed of 40 m/s	7.88 ± 0.051	729.91 ± 4.711

. The TiZrHfNiCuCo metallic coating deposited at an Ar flow rate of 5 sccm and an input power of 300 W exhibited a high hardness value of 11.64 GPa, which is higher than that of its melt-spun glassy ribbon counterpart [50]. The higher hardness of the metallic coating, compared to the melt-spun glassy ribbon, is mainly attributed to the solid-solution strengthening, Hall-Petch effect, and large residual stress [35,51].

In the case of deposited metallic films with an amorphous structure, the structure can be characterized by the assembly of nanometer-sized atomic clusters. This is distinct from the amorphous structure of metallic glasses obtained through the rapid cooling of metallic liquids, which is based on the layering of individual atoms rather than atomic clusters [52].

High power leads to the rapid deposition of particles, and the nucleation rate decreases rapidly. The smaller particles in the films have a higher surface-to-volume ratio than the larger ones, affecting the hardness profile [36]. The grain size is greatly influenced by the gas pressure in the sputter chamber, with pressure reduction leading to grain size reduction, as observed in previous studies, which explains the hardening of materials due to the Hall-Petch effect [34,53]. In this research, we can see that the hardness of the coating is closely related to the density, as evidenced by the oxygen content and surface roughness of the coating.

4. Conclusions

The TiZrHfNiCuCo metallic coatings were deposited at room temperature using a magnetron sputtering system, with variations in input power and Ar flow rate. XRD and TEM analyses revealed that the microstructure of the TiZrHfNiCuCo metallic coatings consisted of an amorphous structure. Although phase transformations due to changes in input power and Ar flow rate were not observed, significant effects were noted in terms of oxygen content, surface roughness, and mechanical properties. The oxygen content, which can be indicative of the density of the coating, was found to be lower at a higher input power and lower Ar flow rate. Specifically, at an input power of 300 W and an Ar flow rate of 5 sccm, the films exhibited a roughness of 0.603 ± 0.006 nm and a Vickers hardness of 1078 ± 21.89 HV, demonstrating enhanced density and hardness.