Study of Microstructure Regulation and In Situ Tensile Performance of Ni-Al Films

Abstract

In this paper, Ni-Al films were prepared using magnetron sputtering technology. The microstructure of the films and the relationship between the residual stress and the adhesion strength were studied. More importantly, the mechanical strength–ductility properties of Ni-Al films were evaluated by in situ tensile testing. The results showed that the film mainly consisted of Ni₃Al phase and Ni-based solid solution at the low power of Al target. The phase transition occurred, and the NiAl phase appeared when the Al sputtering power was increased. The complex structure, with the coexistence of Ni-based solid solution, Ni₃Al, and NiAl phases, possessed increased residual stress and reduced adhesion strength. Meanwhile, the crack was easily penetrated through the inside and outside of the film, falling off during in situ tensile testing. While the small residual stress and large adhesion strength were obtained, only Ni₃Al and NiAl coexisted, and the film fell off with difficulty. When the Al sputtering power was 400 W, the film showed the largest adhesion strength and the smallest residual stress. The best comprehensive performance was achieved with a tensile strength of 854 MPa and a yield strength of 90 MPa. The Al content of the film was up to 23.03 at.%, which was beneficial to the application and performance improvement of the film in molten salt corrosion resistance.

1. Introduction

In recent years, concentrated solar power technology (CSP) has developed rapidly and become the focus of the new energy industry [1]. Due to its low melting point (~380 °C) and high boiling point (~900 °C), chloride-based molten salt is expected to be the next generation of heating or heat storage medium for CSP [2,3]. However, chloride-based high-temperature molten salt creates higher requirements for high-temperature corrosion resistance of the critical components (mainly stainless steel) of CSP, and the commonly used stainless steel, such as 347H, and it is challenging to meet the demand. Inconel 625 and other superalloys have good corrosion resistance to molten salt, but high cost limits their widespread use in CSP systems. Therefore, developing a new type of protective film which can be deposited on the surface of 347H stainless steel and withstand the erosion of chloride-based molten salt has become an important research topic [4,5,6].

Owing to its high melting point and good high-temperature corrosion resistance, Ni-Al films have been widely used as a protective coating for turbine blades of aero-engines [7,8,9,10,11,12]. At present, there are relatively few reports concerning the application of Ni-Al films as a protective coating against chloride-based molten salt erosion on the surface of critical components in the CSP system. The film must be uniform and dense, with good adhesion strength and enough film thickness. In addition, the low tensile strength and toughness of this material greatly limit its application [13,14,15]. Magnetron sputtering technology to prepare the film has the advantages of uniform, compact, and good adhesion, conducive to obtaining high-quality Ni-Al films. Zhang et al. used magnetron sputtering to prepare the β-NiAl coating and found that the oxide film generated by the microcrystalline β-NiAl phase had better adhesion [2]. Dai Bo et al. concluded that Ni₃Al coating prepared by magnetron sputtering at 400 °C had the best corrosion resistance and mechanical properties [16]. Zhong et al. prepared NiAl coating by magnetron sputtering with excellent high-temperature oxidation resistance [17]. However, the structure-activity relationship between the microstructure of the films and the mechanical properties is still unclear. Its tensile properties need further study concerning tensile strength and toughness.

Therefore, in this paper, Ni-Al films were prepared by magnetron sputtering technology combining radio frequency (RF) and direct current (DC) power. The microstructure, adhesion, residual stress, and mechanical properties of the prepared films were evaluated. In particular, the tensile strength of the films was studied by in situ tensile. An optimized Ni-Al coating film with good adhesion and tensile strength was obtained by adjusting parameters.

2. Experimental

Figure 1 shows a schematic diagram of the magnetron sputtering system. Ni-Al films were deposited using the radio frequency (RF) and direct current (DC) magnetron sputtering technique, and 347H stainless steel with a thickness of 1000 ± 25 μm was used as substrates. The 347H stainless steel sheet with a 10:1 aspect ratio was used to accurately measure the residual stress. The target materials were pure 99.99% nickel and aluminum from Dream (Beijing, China) Material Technology Co. Ltd. The diameter and thickness of the circular target were 101.60 ± 0.05 mm and 6.35 ± 0.05 mm, respectively. The Ni and Al targets were connected to DC and RF power, respectively. Before deposition, the substrates were ultrasonically cleaned in acetone for 10 min, rinsed with deionized water, and dried with nitrogen. The background vacuum reached 6.0 × 10⁻⁴ Pa and then introduced 80 sccm of argon. The current of DC power and sputtering power of RF power was set as 1.0 A and 400 W to remove the oxidized layer of the target, respectively. During deposition, the substrates were heated to 150 °C, and the rotation speed of the sample platform was set at 5.0 r/min. The films were deposited at Al target with RF power of 100, 200, 300, and 400 W, and deposition time was 140, 125, 100, and 85 min at a bias voltage of −50 V. The DC of the Ni target was fixed at 0.5 A, and 10 sccm argon was introduced. The chamber pressure reached 0.1 Pa. The film thickness was kept at around 2.3 ± 0.2 μm.

The thickness of the film was measured using an AlphaStep D-500 nanoscale Profilometer and Quanta 450 FEG scanning electron microscope (SEM). The element content of the film was measured by an EPMA-1600 electron probe (EPMA). An Empyrean S3 type X-ray diffractometer (XRD) and a JEM-F2100 type transmission electron microscope (TEM) were used to analyze the microstructure of the films. An MFT-4000 multifunctional material surface performance test instrument was used to measure the adhesion. The residual stress was calculated from the curvature of the film/substrate composite using Stoney’s equation. The in situ tensile test of the film was carried out using the MTEST2000 in situ tensile table loading scanning electron microscopy (SEM) technology and equipped with a 2000 N maximum load cell. The displacement rate control mode was adopted. The drawing speed was 0.1 mm/min. To prevent the sample from slipping during the stretching process, two round holes with a diameter of 4 mm were machined at both ends of the sample to insert wedges and fix the sample. To observe the generation and propagation of a crack in time, a small U-shaped notch was made on the sample as a preset crack source. To facilitate the observation of the microstructure changes during the tensile process, the samples were placed in CH₃COOH+HNO₃+HCl mixed solution with a volume ratio of 8:4:1 for 30 s.

3. Results and Discussion

3.1. Microstructure

Figure 2a shows the XRD patterns of Ni-Al films with different RF sputtering power. The tetragonal Ni₃Al phase was formed with an RF sputtering power of 100 W. The Ni peak was found, indicating the formation of a Ni-based solid solution. At the RF sputtering power of 200 and 300 W, the cubic NiAl phase appeared beside the Ni-based solid solution and Ni₃Al phase. When the RF sputtering power was further increased to 400 W, only Ni₃Al and NiAl phases coexisted. With the increase of RF sputtering power, the content of Al in Ni-Al films increased gradually, which promoted the phase transformation of the films (Ni→Ni₃Al→NiAl phases). The content of Al element was 10.26, 16.15, 21.90, and 23.03 (at.%), respectively, corresponding to RF sputtering power of 100 to 400 W. The phase shift analyses for the peaks in the XRD patterns are shown in Figure 2b. Additionally, the peaks slightly shifted towards higher angles with the increase in RF sputtering power of 100 to 300 W films. When the RF sputtering power was further increased to 400 W, the peaks slightly shifted towards a lower angle. The shift in the diffraction peaks was attributed to the change in the crystallite size. Hence, the stresses in the Ni-Al films were produced, resulting in the shift. The average crystallite size was determined using the Scherrer formula [18]. Consequently, crystallite size decreases first and then increases with increased RF sputtering power, as shown in Figure 2b. The film crystallite size was small and formed nanocrystals. Considering the Ni-Al binary phase diagram, the Al content of the NiAl phase in the thermodynamic equilibrium state was more than 25.00 at.% above 400 °C. The main reason was that the substrate temperature of magnetron sputtering in this paper was 150 °C. The Ni and Al atoms or atomic groups sputtered on the substrate formed compound films at a lower temperature, which was in a non-equilibrium state [19].

The XPS analysis was used to understand the valence states of Ni and Al in the specimen. Figure 3a shows the XPS spectra of Ni 2p. The fitting result in the Ni 2p XPS spectrum shows that Ni can be assigned to Ni 2p3/2 and Ni 2p1/2 at 852.3 ± 0.4 eV and 869.7 ± 0.4 eV, respectively. One satellite peak was detected for each primary peak, indicating that Ni coexisted in a mostly metal and little oxidation state, and had a chemical state of Ni⁰ 2p3/2 and Ni²⁺ 2p3/2. The main reason was the film’s contact with oxygen in the air. The peak intensity of Ni²⁺ was very weak, implying that oxygen content was low. Figure 3b shows the XPS spectra of Al 2p. The fitting result in the Al 2p XPS spectrum shows that Al can be assigned to Al 2p3/2 and Al₂O₃ at 72.0 ± 0.8 eV and 74.2 ± 0.3 eV, respectively. Another peak, at 65.9 ± 0.4 eV, can be assigned to Ni 3p. It shows that Al was highest in the metal and a little oxidized state [20].

Figure 4 shows the high-resolution electron microscope (HRTEM) image and selected area electron diffraction (SAED) at 100 W RF sputtering power, from which it can be seen that Ni₃Al and Ni appear in the Ni-Al film. The crystallite size was very small. Figure 5 shows the bright field TEM image, SAED, and HRTEM image at 400 W RF sputtering power, from which it can be seen that Ni₃Al and NiAl phases appear in the film, consistent with the XRD results. At the same time, the film formed nanocrystals, and the average crystallite size was 6.37 nm, according to the statistics of Image-Pro Plus software. The smaller crystallite size could promote the formation of the Al₂O₃ phase after the surface oxidization and enhance the adhesion and high-temperature corrosion resistance of the film [21].

3.2. Residual Stress and Adhesion Strength

Figure 6 shows the residual stress and adhesion strength of Ni-Al films as a function of RF sputtering power. It can be seen that the residual stress of the film increased first and then decreased with the increase of RF sputtering power. The maximum residual stress was 0.95 GPa at the RF sputtering power of 300 W, while the lowest residual stress of 0.62 GPa was obtained at 400 W. With the RF sputtering power increasing from 100 to 200 W, the film underwent phase transformation resulting in a significant change in the volume and stress of the film [22,23]. At 300 W, the phase transition was further aggravated with the increase of Al content, and the film volume further changed, leading to the rise in stress. At the same time, when the RF sputtering power gradually increased, the energy of the particles deposited on the substrate surface increased, and the lattice distortion occurred, leading to the increase of residual stress. During the deposition process, the incident energy of atoms releases to the surrounding atoms and forms a heat-effect zone. With increasing sputtering power to 400 W, the heat-effect zone became larger. The generated accumulation of atomic energy relaxed and rearranged the distorted structures, thus promoting a dramatic decrease in residual stress [24]. Figure 6 shows the minimum adhesion strength as 9.0 N at 300 W, and the best adhesion strength of 28 N is obtained at 400 W. The residual stress caused the Ni-Al coating and substrate delamination, and the adhesion strength changed with the stress [25,26,27]. The residual stress of the film affects the adhesion. The larger the residual stress, the smaller the adhesion strength. In conclusion, the complex structure of Ni-based solid solution, Ni₃Al, and NiAl phase coexistence led to the increase of residual stress and the reduction of adhesion strength. However, with the increase of Al content, the coexistence of Ni₃Al and NiAl phase was more conducive to improving the film residual stress and adhesion strength.

3.3. In Situ Tensile Research

Figure 7 shows the in situ tensile stress-strain curve of Ni-Al films with different RF sputtering power. The inset image of a flat dog bone specimen was used in the in situ test. The sharp point on the curve indicated that the loading stopped at this point, and the photographs were taken. It can be seen that when the RF sputtering power was 100, 200, and 300 W, the samples had strain hardening accompanied by obvious necking, and the tensile strength (σ_s) was 945, 536, and 474 MPa, respectively. The tensile strength of the sample reached 854 MPa, and there was no strain hardening and necking at 400 W. As the RF sputtering power increased from 100 to 400 W, the yield strength (σ_0.2) was 89, 75, 61, and 90 MPa, and the corresponding yield ratio was 0.094, 0.138, 0.129, and 0.105, respectively. It was found that the brittleness of the Ni-Al films varied with the change in RF sputtering power. When the sputtering power was 100 W, it showed the best tensile strength and the lowest yield ratio. While the sputtering power was 400 W, the best plasticity was achieved as well as good yield strength and tensile strength. Additionally, the higher Al content will be more conducive to forming corrosion-resistant alumina film. Combining the in situ tensile stress-strain curve and XRD data, the tensile strength change and main peak shift tendency with increased RF sputtering power were consistent. The shift in the diffraction peaks was attributed to the difference in the crystallite size. The change in crystallite size affected the tensile strength of Ni-Al films when the content of Al was less than 23.03%.

Figure 8 shows the surface microstructure of Ni-Al-100 W film during the in situ tensile testing. Figure 8b–f corresponds to b-f breakpoints in the stress-strain curve in Figure 7. Figure 8a shows the film’s surface as relatively intact without any cracks when no stress is applied. The small pits were caused by corrosion before stretching. When the stress was 470 MPa, and the strain was 5.56%, the crack initiation source was first observed in the red circle area (Figure 8b), where many pits existed. In Figure 8c, with the stress at 640 MPa and the strain at 7.69%, the crack further extends and appears in several places. The direction of the primary crack growth is perpendicular to the direction of the tensile stress. In Figure 8d, when the stress is 792 MPa and the strain is 10.82%, the cracks further expand and increase significantly, and part of the film falls off the substrate. In Figure 8e, when the stress of the sample increases to 874 MPa and the strain is 13.95%, the crack propagation intensifies. The crack spacing increases significantly, and the film spalling is aggravated. In Figure 8f, where the stress is 934 MPa and the strain is 17.00%, the crack spacing further increases and the film falls off, but no cracks are observed on the substrate [28,29,30,31,32]. When the crack initiation source in Figure 8b is enlarged-viewed in Figure 9a, it can be seen that the crack initiation source runs through the pits successively and expands along the large pits. The film defect becomes the origin of the crack. The crack is amplified, and the composition analysis is carried out, as shown in Figure 9b–d. It can be seen that the Fe element concentrated at crack B, the stainless steel matrix, and the Ni element mainly concentrated at coating A, indicating that the crack has penetrated inside and outside the Ni-Al film.

Figure 10 shows the progressive fracture SEM images of Ni-Al-200 W film during the in situ tensile testing, and Figure 10a shows that the film’s surface is continuous and intact when no stress is applied. When the stress was 340 MPa, and the strain 4.95%, the crack initiation source was first observed in the red circle (Figure 10b). The crack continued to expand with the increased tensile stress. When the tensile stress was 490 MPa, and the strain was 7.78%, the crack in Figure 10d further extends and develops into a main crack. Meanwhile, multiple cracks appeared in other parts, and the growth direction of the main crack was perpendicular to the direction of the tensile stress. It differed from the 100 W sample because the cracks tended to be distributed in strips. In Figure 10e, the cracks further expand. The number of cracks increased significantly when the strain was 10.82%, and the stress was 530 MPa. The part of the film was separated from the substrate. When the sample continued to be stretched, the stress was 475 MPa, the strain was 12.21%, the crack width further increased, and the film fell off from the substrate (Figure 10f).

Figure 11 shows the progressive fracture SEM images of Ni-Al-300 W film during the in situ tensile testing, crack, and composition distribution diagram. It can be seen from Figure 11a that when no stress is applied, part films on the substrate surface fail due to corrosion and other reasons. When the stress of the sample was 353 MPa, and the strain was 2.98%, the crack initiation source was first observed in the film (red circle area) with no failure, and the main crack spread along the direction through the pit (Figure 11b). The crack was amplified 500 times, and the composition analysis was conducted (Figure 11d–f). The Fe content at the crack increased significantly, deriving from the stainless steel matrix and indicating that the crack had penetrated inside and outside the coating film. In Figure 11c, the crack further expands. Part of the film even fell off from the substrate when the stress of the sample was 441 MPa, and the strain was 5.35%. Compared with the RF sputtering power of 100 and 200 W, the film was easier to peel from the substrate and bear tensile stress was the smallest when the sputtering power was 300 W. The film was more likely to fall off the substrate due to the small adhesion strength when the tensile stress was constant.

Figure 12 shows the progressive fracture SEM images of Ni-Al-400 W film during the in situ tensile testing. Figure 12a shows the continuous integrity of the film before the stress is applied. When the stress was 543 MPa and the strain was 6.30%, the crack initiation source first appeared in the red circle area (Figure 12b). In the local magnifying image of Figure 12b, it can be observed that the crack expands along the corrosion line. The crack further extended and developed into a main crack when the stress was 597 MPa and the strain was 7.85% (Figure 12c). Multiple cracks appeared in other parts of the sample, and a part of the film was found to fall off in the red circle area. The main crack continued to grow, and the film detachment phenomenon increased when the stress was 647 MPa and the strain was 9.36% (Figure 12d). The direction of the main crack growth was perpendicular to the direction of the tensile stress, and the cracks tended to be distributed in a strip shape. Most of the film fell off the substrate when the stress was 841 MPa and the strain was 21.25% (Figure 12e). In Figure 12f, cracks appear in the substrate as the tensile stress increases.

When the RF sputtering power was increased to 400 W, only the Ni₃Al and NiAl phases coexisted. The average crystallite size was 6.37 nm. The size of the crystallite was very small. It was advantageous to enhance the adhesion strength. The film showed the largest adhesion strength and the smallest residual stress when the sputtering power was 400 W. The best comprehensive performance was achieved with a tensile strength of 854 MPa and a yield strength of 90 MPa. The increase of Al content affected the film phase structure. The present study has achieved balanced values of strength and toughness for Al content of 23.03 at.%, which was beneficial to the application.

4. Conclusions

In summary, Ni-Al films with different Al content were prepared using magnetron sputtering. The microstructure and mechanical properties of the films were studied. The yield strength and tensile strength of the films were further analyzed by in situ tensile. The formation of Ni₃Al phase and Ni-based solid solution when the RF sputtering power was small. The phase transition and NiAl phase appeared with the RF sputtering power increased. The film crystallite size was small and formed nanocrystals. It was helpful to improve the high-temperature corrosion resistance of Ni-Al film. During the preparation of Ni-Al films, the residual stress affected the adhesion strength of the films. In the in situ tensile process, the lower the residual stress, the higher the adhesion strength, and the film fell off less easily. By contrast, the higher the residual stress, the lower the adhesion strength and the coating fell off more easily and cracked through the film. The film showed the largest adhesion strength and the smallest residual stress when the sputtering power was 400 W. The balance of strength and toughness for an Al content of 23.03 at.% was achieved, which was helpful for the performance improvement of the Ni-Al film.