2019 年 60 巻 11 号 p. 2305-2310
In order to understand the evolution of texture during the aerosol deposition (AD) method, coatings were deposited under various conditions using pure nickel powder particles. These particles are thought to undergo plastic deformation during deposition. X-ray diffraction analyses of the surfaces of the coatings obtained revealed specific textures. A {101} fiber texture was observed in as-deposited specimens. Texture development was observed with increasing gas flow rate. The strain in the nickel coating was estimated to be about 1.0–1.4 in true strain. This strain is thought to be induced by the nickel particles impacting the substrate during deposition.
Fig. 10 Relationship between maximum pole density and absolute value of true strain after uniaxial compression.
Aerosol deposition (AD) is a coating process that can produce dense coatings at room temperature by impacting a substrate with solid particles. This process is also referred to as aerosol gas deposition, aerosol-type jet printing, or kinetic spray. The basic method was developed by Hayashi in the 1980s.1) After that, the AD method has spread worldwide through the work of Akedo involving the deposition of ceramic coatings.2–5)
This coating method has attracted attention because dense and crystalline coatings can be formed without changing the composition of particles. Many powders such as ZrO2,6–9) α-Al2O3,2,3,10–15) Mullite,16) Pb(Zr52,Ti48)O3,3,4) Ni–Zn–Fe3O4,3) Li4Ti5O12,17) AlN,3) TiN,18) MgB2,3) Ti3SiC219) and Cu5) have been used to produce coatings. It has been reported that the deposited coatings have a dense and crystalline microstructure with no preferential orientation distribution.2)
One of the advantages of the AD method is its ability to form dense coatings up to several micrometers thick without heating. The resulting coatings are considered to form by fracture and/or plastic deformation of the particles due to the impact of particles on the substrate. This coating principle is called room-temperature impact consolidation.2–4) It has also been proposed that a coating is formed through a plasma that is generated as follows.8,9) The particles ejected from the nozzle tip onto the substrate are positively charged owing to the friction between the particles and the nozzle wall. Right before these particles reach the substrate, electrons fly out of the substrate and to the particles. The plasma may be generated by the impact of electrons on the deposition gas. The plasma presumably forms an active surface on the particles, bonding them together to produce a coating.
Generally, when the deformation occurs in polycrystalline metals, alloys, intermetallic compounds and ceramics, a deformation texture forms by the activation of several slip and twin systems.20–28) Therefore, if the plastic deformation during the AD method occurs as a result of the collision of particles with the substrate, the formation of textures may occur. However, there are only a few reports that indicate the formation of texture by the AD method. Fuchita et al. report the possibility of texture formation in ZrO2 coatings.8) Furthermore, texture formation by the deposition of Al2O3 particles has been reported.13,14)
In the present study, coatings are formed under various deposition conditions by the AD method using pure nickel metal powders having a face-centered cubic lattice (fcc), whose slip and twin systems are well known. The effects of deposition conditions on texture are examined experimentally. As will be described later, this study found that the texture of the nickel coating formed by the AD method has the same characteristics as the texture formed by uniaxial compression deformation of nickel at room temperature. It is thought that since plastic strain and texture development are correlated, the plastic strain in the coating can be estimated as the strain under uniaxial compression deformation by comparing the textures. When the plastic strain in the coating is due to the deformation caused by the collision of particles to the substrate, it has a possibility to estimate the impact energy and velocity of the particles from the strain. From this perspective, it is useful to establish a method for estimating the true strain value in the coating. In addition, it is important and useful to estimate the true strain in the coating, since the plastic strain in the coating affects the mechanical properties of the coating and grain growth when the coating is heated after coating formation. Therefore, the plastic strain induced in the coating during the AD method was estimated from the resulting texture of the coating.
Pure nickel powder was used to form the coatings. Figure 1 shows the nickel powder used for the deposition. The average particle size was about 2–3 µm. Although the powder particles were somewhat angular, they had roughly equiaxed shapes. The purity of the powder was 99.9 mass%. For each coating, the powder was dried at 523 K and then loaded into the aerosol chamber. The coatings were deposited at room temperature by the AD method (Type GD-AE04/SS2, Fuchita Nanotechnology, Ltd., Tsukuba, Japan). A Mo disc having a diameter of 15 mm and a thickness of 3 mm was used as the substrate. Prior to deposition, the substrate surface was polished with diamond paste. N2 gas was used as a carrier gas for AD. The dimensions (width × thickness) of the nozzle port was 5 × 0.5 mm. During the AD process, the gas flow rate, nozzle-substrate distance, angle between the substrate and the direction of flow from the nozzle (nozzle angle), scanning speed, and number of scans were 8–15 L/min, 7 mm, 90°, 150 mm/min, and 30 times, respectively. Figure 2 shows the specimen where nickel was deposited on the Mo substrate. Nickel powder was deposited on the entire substrate except on its sides, which were masked. The uncoated regions were formed to enable measurement of the coating thickness by a surface roughness meter. Visual inspection revealed no large cracks on the coating. If the coating thickness was less than 1 µm, the specimen was recoated until it was 1 µm thick.
Scanning electron micrograph of pure nickel raw powders used for the AD coating process.
Optical micrograph of a nickel coating deposited on a Mo substrate by AD method.
After deposition, the surface and cross-section of the coating were observed by scanning electron microscope (SEM) (JSM-7001F, JEOL Ltd., Tokyo, Japan). The thickness and surface roughness of the deposited coating were measured by a surface roughness meter (ET200, Kosaka Laboratory Ltd., Tokyo, Japan). The texture of the coating was determined through X-ray diffraction (XRD).
2.2 Uniaxial compression deformation of nickel produced by powder sinteringA nickel specimen having no preferential orientation was produced by nickel powder sintering. This specimen for uniaxial compression deformation was prepared by charging nickel powder into a cylindrical container made of stainless steel, compacting for 600 s under a load of 0.32 MPa, and sintered in a vacuum at 1273 K for 10 h. The average size and purity of the powder used were 45 µm and 99.7 mass%, respectively. After sintering, cylindrical specimens having a diameter of 12 mm and a height of 18 mm were prepared by machining with a lathe. In addition, to improve the lubrication on the compression plane, concentric grooves with depths of about 0.1 mm were introduced at intervals of 0.5 mm. MoS2-containing grease was used to lubricate the compression face. The tests were performed on a screw-driven type testing machine. The deformation was applied up to various true strains at a constant crosshead speed of 0.017 mm/s. After the compression tests, a midplane section of each specimens was prepared by mechanical and chemical mechanical polishing for texture characterization by XRD.
2.3 Texture characterizationTextures were characterized through the Schulz reflection method, with Cu-Kα radiation filtered by a monochromator (Ultima IV, Rigaku Co., Tokyo, Japan). Here, the X-ray tube voltage and tube current were 40 kV and 40 mA, respectively. To prevent an overlap between the nickel diffraction peaks and those from the substrate, a Mo substrate was used to characterize the coating. The intensities of the 100, 110, and 111 peaks of nickel, which has an fcc structure, were measured. From the pole figures obtained, the orientation distribution function (ODF) was calculated by the arbitrarily defined cell (ADC) method29) (TexTools Ver. 3.3, Resmat Co., QC, Canada). The main component and sharpness of the texture, which correspond to the position and value of the maximum pole density, were determined from the normalized pole figure and the inverse pole figure derived from the ODF. The normalized pole density, P(α, β), can be expressed as
| \begin{equation} \mathrm{P}(\alpha,\beta) = \cfrac{I(\alpha,\beta)}{\cfrac{1}{2\pi}\displaystyle\int_{\alpha = 0}^{\alpha = \pi/2}\int_{\beta = 0}^{\beta = 2\pi}I(\alpha,\beta)\sin\alpha d\alpha\,d\beta}, \end{equation} | (1) |
Figure 3 plots the deposition rate of the nickel coating versus the N2 gas flow rate. At gas flow rates of 8 and 10 L/min, the deposition rates were, respectively, 0.33 and 1.20 µm/min·cm2. However, when the gas flow rate increased further to 15 L/min, the deposition rate decreased to 0.34 µm/min·cm2. Here, the deposition rate in this experiment was evaluated in a region of 5 × 5 mm (nozzle width × scanning distance same as the nozzle width).
Relationship between gas flow rate and deposition rate.
Figure 4 shows surface and cross-sectional micrographs of a nickel coating. This coating was deposited under 10 L/min. The surface showed no cracks or chips, but rather, a flat section, where the nickel particles were crushed by the impact (Fig. 4(a)). This suggests that the impact of the particles on the substrate may be causing plastic deformation in the particles during deposition. Some regions that the particles were crushed were also observed in Fig. 4(a) (indicated by the arrows.). The surface roughness, Ra, of the coating was about 78 nm. Because the substrate surface before deposition was very smooth (Ra = 7 nm), the resulting coating was also relatively smooth. Figure 4(b) shows the cross-section of the same coating. Interface between coating and substrate was indicated by the dashed line. It can be seen that a dense coating was produced. Furthermore, no delamination is observed between the substrate and the coating. The interface between the coating and substrate was flat and smooth. It is indicating that the substrate is not largely plastic deformed. Therefore, it can be judged that the texture caused by plastic deformation is not formed on the substrate surface. Thus, in section 3.2, it was decided to evaluate the texture only of the formed coating.
Scanning electron micrographs of a nickel coating deposited on Mo substrate: (a) surface of as-deposited specimen, and (b) cross-section of as-deposited specimen.
Figure 5(a) and (b) show the XRD patterns obtained from the raw nickel powder and the nickel coating deposited on the Mo substrate, respectively. Diffraction signals from the Mo substrate can be seen in the case of the coating (Fig. 5(a)). However, the other diffraction signals occurred at the same diffraction angles as those from the raw nickel powder (Fig. 5(b)). This indicates that the nickel coating maintained the same crystal structure as the raw nickel powder. Furthermore, no broad patterns due to the amorphous phase or diffraction signals from the oxides of Mo and nickel were detected. These results demonstrate that the formation of a uniform and crystalline nickel coating was achieved without oxidation of the substrate.
X-ray diffraction spectra of (a) nickel deposited on Mo substrate and (b) nickel raw powder.
Figure 6 shows the texture of a nickel coating processed under a gas flow rate of 10 L/min. Figures 6(a), (b), and (c) show the {001}, {101}, and {111} pole figures, respectively. The pole densities are projected onto the coating plane. The mean pole density is taken as unity. In the {001} pole figure, the region of high pole density spread at approximately 45° and 90° away from the center (Fig. 6(a)). In the {101} pole figure, high pole density occurred at the center (Fig. 6(b)). As for the {111} pole figure, high pole density spread at approximately 35° and 90° away from the center (Fig. 6(c)). These results indicate that the angular relationship between the {001} and {101} planes, and the {111} and {101} planes satisfy 45° and 90°, and 35° and 90°, respectively. This relationship is the same as the angular relationship between crystal planes in the fcc lattice. Thus, it is possible to say that there are no serious errors in the analyzed textures. These pole figures show that the coating has a {101} fiber texture. Here, the {101} fiber texture obtained has the same characteristics as the texture obtained after uniaxial compression deformation of fcc metals at room temperature.20) Figure 7(a), (b), and (c) show inverse pole figures for nickel coatings from the normal direction (ND) after deposition at gas flow rates of 8, 10, and 15 L/min, respectively. The mean pole density is taken as unity. The main component in each pole figure is located in {101}. Figure 8 plots the maximum pole density versus the gas flow rate. When the deposition was performed at gas flow rates of 8, 10, and 15 L/min, the maximum pole densities were 3.7, 3.6, and 4.8. Thus, maximum pole density seems to increase with the gas flow rate.
Orientation distribution of nickel coating deposited at a gas flow rate of 10 L/min. (a) {001} pole figure, (b) {101} pole figure, and (c) {111} pole figure.
Inverse pole figures from the ND direction after deposition at different gas flow rates: (a) 8 L/min, (b) 10 L/min, and (c) 15 L/min.
Relationship between maximum pole density and gas flow rate.
The relationship between the gas flow rate and the deposition rate at a nozzle angle of 90° in the AD method has been reported for the deposition of TiN18) and Al2O315) powders. The same relationship has been reported at a nozzle angle of 60° for mullite powder.16) For both nozzle angles, as the gas flow rate increased, the deposition rate increased up to a maximum and then decreased with further increase in gas flow rate.15,16,18)
Coating formation by the AD method is thought to involve fracture and/or plastic deformation of the particles due to the impact of the particles on the substrate. This principle is called room-temperature impact consolidation.2,3) When the gas flow rate is low, many particles cannot reach the critical velocity for room-temperature impact consolidation, and the particles will be repelled by the substrate or flow away without reaching the substrate. Therefore, no coating can be formed on the substrate. Increasing the gas flow rate eventually brings the particles above the critical velocity required for coating formation. Thus, even in nickel metal powders, the formation of a coating on the substrate begins, and the deposition rate increases with the gas flow rate. On the other hand, too high a gas flow rate leads to a decrease in the deposition rate. This is due to wear of the coating by the increase in impact energy of the particles on the coating surface.18) Furthermore, although the particle velocity increases immediately after ejection from the nozzle, it has also a possibility that the formation of the coating was inhibited. This is because of the decrease in the velocity of the particles during the flight to the substrate due to the increase in gas reflectance from the substrate. These results suggest that the deposition rate is reduced in the nickel coating produced at a gas flow rate of 15 L/min.
4.2 Texture formation of nickel coating produced by AD methodIt has been proposed that the fabrication of a coating by the AD method involves the mechanisms of plasma generation8,9) and room-temperature impact consolidation.2,3) The former mechanism presumably occurs through the generation of plasma. The formation of an active surface of particles by plasma produces a coating by bonding the particles together. Thus, it is considered that no texture will form because of the adhesion of particles having no preferential orientation distribution. On the other hand, the latter mechanism involves fracture and/or plastic deformation induced by the impact of particles on the substrate. Thus, it is expected that if coating formation occurs preferentially by plastic deformation of the particles, a texture will form. Therefore, for coating formation by AD using nickel metal powder, which at room temperature is much more ductile than ceramics, a texture will presumably form in the coating owing to plastic deformation in the particles. Indeed, in the present study, the deposition of nickel particles by AD yielded a fiber texture with its {101} plane parallel to the substrate surface, regardless of the gas flow rate. Furthermore, it was found that the value of the maximum pole density, which is indicative of the evolution of texture, increased with increasing gas flow rate.
Uniaxial compression deformation in bulk metals with an fcc lattice typically yields a {101} fiber texture.20) The reason why the {101} fiber texture is formed is that cross slip occurs owing to the activation of multiple {111} ⟨110⟩ slip systems, which aligns the compression axis with the {101} plane normal, a stable orientation in uniaxial compression.31) It is also known that in bulk metals, the texture develops with increasing strain during deformation.21) Although the texture of the nickel coating has the same characteristics as a texture obtained by uniaxial compression of any fcc bulk metal at room temperature, the formation of the texture under deposition is thought to be attributable to the activity of the {111} ⟨110⟩ slip systems induced by uniaxial compression at the time of impact of the raw material powder with the substrate. Furthermore, the increase in gas flow rate increases the velocity of the particles and hence the kinetic energy at the time of impact. Thus, it is considered that the particles contributing to coating formation undergo greater plastic deformation, leading to a texture.
Thus, the strain induced in the coating can be predicted from the textures of nickel coatings by comparing the textures of nickel bulk metals uniaxially compressed at various strains. Figure 9(a), (b), (c), and (d) show the texture, obtained by uniaxial compression for cylindrical nickel specimens produced by powder sintering under true strains of 0.04, 0.54, 0.99, and 1.2, respectively. The mean pole density is taken as unity. In inverse pole figures observed from the ND direction, the position of the main component can be seen at {101} under all deformation conditions. The value of the maximum pole density tends to increase with increasing true strain. Figure 10 shows the relationship between maximum pole density and applied true strain. It can be seen that the maximum pole density increased linearly with true strain.
Inverse pole figures from the ND direction after uniaxial compression at different true strains: (a) ε = 0.04, (b) ε = 0.54, (c) ε = 0.99, and (d) ε = 1.2.
Relationship between maximum pole density and absolute value of true strain after uniaxial compression.
Estimates of the true strain induced in the nickel coating based on values of the maximum pole density are 0.98, 0.94, and 1.37, respectively, at gas flow rates of 8, 10, and 15 L/min. This means that the particles undergo a large plastic deformation (63–75%) during deposition. It is not known why the applied true strain in the coating is different for each deposition condition even though the deposition rate is roughly the same for gas flow rates of 8 and 15 L/min. At 8 and 15 L/min, the true strains in the nickel coating are about 1.0 and 1.4, respectively. The low deposition rate at 15 L/min, presumably because of the wear of the coating due to the impact of the particles. On the other hand, particles impacting to the substrate at high velocity may deform largely and develop texture. However, the relationship between the deposition rate and the texture in terms of the true strain induced in the coating remains unclear and necessitates further investigation.
A {101} fiber texture was obtained in nickel coatings formed by the aerosol deposition (AD) method. Fiber texture developed with increasing gas flow rate. Texture formation is thought to emerge as a result of the plastic deformation of nickel particles induced by the impact of nickel particles on the substrate during deposition. As in the case of the uniaxial compression of fcc bulk metals, activation of multiple {111} ⟨110⟩ slip systems form a {101} fiber texture. The plastic strain stored in the nickel coating is assumed to range from 1.0 to 1.4 in true strain.
This research was partially supported by the “Advanced Low Carbon Technology Research and Development Program” from the Japan Science and Technology Agency. Furthermore, this research was partially supported by a grant-in-aid for scientific research (C) (15K06501) and (B) (18H01745) from the Japan Society for the Promotion of Science. The authors greatly appreciate the grants.
