A coating is defined as a surface layer, which is capable of improving or, in some cases preventing the interaction between the nickel base superalloy and the surrounding environment. The oxidation resistance of a material depends on its ability to form a protective oxide scale. If the oxide scale is readily formed, is dense, creates a barrier to oxygen diffusion, is adherent and stable, it will act as a barrier between the metal and the environment, and will slow down, or even stop further oxidation. Aluminum has the best combination of properties to form a stable oxide; i.e. a high free energy of formation as shown in Table 1, and also exhibits a low growth rate as demonstrated by the rate constants in Figure 1. Aluminum is therefore chosen as the coating surface layer. (Ref. 1)

Aluminide coatings form a reservoir of aluminum that acts as a sacrificial layer for protecting the superalloy. The aluminum content of the coating decreases due to the continuous spallation, i.e. aluminum oxide mechanically removed from the surface, and reformation of the aluminum oxide scale. The coating degradation process involves a series of phase transformations until the coating, or aluminum reservoir is depleted, and the less protective nickel oxide formation occurs from the substrate. (Ref. 2)

Aluminum is deposited using a diffusion process. Diffusion aluminide coatings involve the interdiffusion between the aluminum deposited at the surface, and elements such as nickel, chromium, tungsten, and carbon from the nickel base substrate. Chemical Vapor Deposition (CVD) aluminizing has been used as a technique for forming aluminide coatings on gas turbine blades and nozzles. Gaseous reactants, typically aluminum chloride and hydrogen, are introduced from an external source into a pre-heated coating chamber where the reactions

1. AlCl₃(g) + 3/2H₂(g) + Ni(s) ® NiAl(s) + 3HCl(g)
2. AlCl₃(g) + 3/2H₂(g) ® Al(s) + 3HCl(g)

take place at relatively high temperature above 1000°C to form nickel-aluminum or aluminum at the surface. The CVD process, as shown in Figure 2, has the advantage of providing a low activity outward diffusion coating, thus generating a two-layer coating with a single-phase beta nickel-aluminum outer layer. As a measure to improve the aluminum oxide stability, a thin layer of platinum is electroplated onto the surface prior to aluminizing. The resulting coating forms a platinum rich beta nickel-aluminum outer layer. The platinum slows down the interdiffusion between the aluminum at the surface and elements from the nickel base substrate. Thus, platinum aluminide form a purer and slower growing aluminum oxide scale with increased scale adherence versus a simple aluminide. (Ref. 3)

CMSX-4 nickel base superalloy (Table 2) specimens, 1" diameter x 1/8" thick, were coated with a single-phase platinum aluminide coating applied by CVD. Oxidation testing was conducted on duplicate specimens in a still air conduction furnace at 982°C (1800°F) for 3000-hours and 1149°C (2100°F) for 100-hours. Half of the duplicate specimens were sectioned using an alumina cutting wheel, mounted in 1" diameter epoxy mounts, and polished for cross sectional analysis. The remaining specimens were kept intact to analyze the platinum aluminide’s top surface. Specimens were submitted for scanning electron microscope, energy dispersive spectroscopy, x-ray diffraction and microprobe analysis to determine the microstructure, composition, and phases present in the as-coated and oxidation conditions.

Microstructure: Coated specimens were evaluated for coating microstructure and composition using an ISI SR-50 Scanning Electron Microscope (SEM) equipped with a Princeton Gamma-Tech Energy Dispersive X-ray Analyzer (EDXA). Figure 3 is a scanning electron micrograph that displays the platinum aluminide coating microstructure on CMSX-4 alloy, and consists of a single phase, two layer coating. The outer 25 to 30 microns layer consists of a single-phase platinum rich beta nickel-aluminum matrix. The second layer is a 25-micron thick refractory enriched diffusion zone. The coating displays a high content of dark inclusions at the center of the coating. The inclusions were identified as alumina grit particles that were used in specimen preparation. Coating thickness was 50 to 55 microns.

The top layer of the aluminide coating, which was deposited using CVD, forms a columnar structure that grows outward from the original CMSX-4 alloy surface as shown in Figure 4. The column widths are approximately 25 – 65 microns in thickness with 10-micron thick grain boundaries. The preferred orientations of columnar growth consist of the <110> and <211> directions depending on the CVD process variables.

Spot 1 & 2 : Single Phase Platinum Rich Beta Nickel-Aluminum

Spot 3 – 5 : Refractory (Cr,W,Ta,Re) Rich Diffusion Zone

Spot 6 : CMSX-4 Nickel Base Alloy

Composition: Coating composition was measured within the outer 25 microns using quantitative EDXA analysis and Microprobe techniques. The measurements were taken within the outer layer because this layer forms a reservoir of aluminum that acts as a sacrificial layer for protecting CMSX-4 nickel base alloy. EDXA analysis was used to get an overall composition for the top layer that includes all elements. Microprobe analysis was chosen to determine the platinum, aluminum, and nickel compositions at the coating surface and every 5 microns to determine the effects of diffusion during the high temperature CVD process. Results are given in Table 3.

There is clearly something different with the microprobe data from 5 to 25 microns. The microstructure of the top layer is clearly a single-phase layer as can be seen from Figure 3. However, the microprobe compositions between the surface reading and within the first 25 microns of coating differ greatly. The top layer was expected to be a platinum rich beta nickel aluminide phase based upon literature data, as well as the microprobe composition at the surface and EDAX composition 10 microns from the surface. In addition, the Phase Analysis section below confirmed the presence of a beta nickel-aluminum phase at the surface based upon x-ray diffraction. The nickel-aluminum binary phase diagram (Figure 5) gives the composition of aluminum to be between 17 and 39 weight percent for beta nickel-aluminum. This phase is confirmed through the microprobe composition at the surface (36-wt.% Al) and the EDAX composition 10 microns from the surface (20.8 wt.% Al).

In summary, the validity of the microprobe data from 5 to 25 microns is in question, and therefore an understanding of the effects of diffusion at high temperatures could not be determined from this data. Reasons for the questionable microprobe data will be discussed in greater detail in the Summary and Conclusions section.

Phase Analysis: The phase at the surface of the platinum aluminide coating was determined by x-ray diffraction. In addition to the phase structure, the unit cell spacing, a, was determined based upon the x-ray diffraction peaks. Due to the complexity of the data found for the platinum aluminide coating, a baseline study was conducted on Yttria Stabilized Zirconia coating, with the results shown in Appendix 1. In addition a study was conducted on the surface of the platinum aluminide coating, with the results shown in Appendix 2. The surface of the platinum aluminide coating was determined to be a simple cubic (beta) nickel-aluminum phase, which is in agreement with the composition. It was interesting to note that the 100, 200, 210 and 300 peaks did not show up for x-ray diffraction of the coating surface. The reason is the coating grows in a columnar structure in the <110> and <211> directions as can be seen from Figure 4.(Ref. 6) Therefore the 100, 200, 210 and 300 peaks could not be generated at the crystal orientation from the coating surface.

Microstructure: Two oxidation tests at 982°C (1800°F) for 3000-hours and 1149°C (2100°F) for 100-hours in a still air conduction furnace were conducted on the platinum aluminide coated CMSX-4 alloy. In both tests, the coating retained a platinum rich beta nickel-aluminum layer at the coating surface. Figures 6 and 7 are scanning electron micrographs that display the platinum aluminide coating microstructure after 982°C (1800°F) for 3000-hours and 1149°C (2100°F) for 100-hours respectively. Both oxidation tests produced a single phase, two layer coating, where the outer 25 to 30 microns layer consists of a single-phase platinum rich beta NiAl matrix. The second layer is a refractory enriched diffusion zone that is 75 microns after 982°C (1800°F) for 3000-hours (Figure 6), and is 30 microns after 1149°C (2100°F) for 100-hours (Figure 7). The coatings display a high content of dark inclusions at the center of the coating. The inclusions were identified as alumina grit particles that were used in specimen preparation. The coating showed only minimal thickness growth due to diffusion after the 1149°C (2100°F) for 100-hours test, however, the coating thickness grew by 100% after the 982°C (1800°F) for 3000-hours test due to growth of the refractory enriched diffusion zone. The coating thickness growth is strongly influenced by the test temperature and duration. A larger growth in thickness can be achieved after a longer test duration even at a lower exposure temperature.

Spot 1 : Single Phase Platinum Rich Beta Nickel-Aluminum

Spot 2 – 5 : Refractory (Cr,W,Mo,Re) Rich Diffusion Zone

Spot 6 : CMSX-4 Nickel Base Alloy

Spot 1 & 2 : Single Phase Platinum Rich Beta Nickel-Aluminum

Spot 3 – 6 : Refractory (Cr,W,Ta,Re) Rich Diffusion Zone

Spot 7 : CMSX-4 Nickel Base Alloy

Composition: Coating composition was measured 10 microns from the coating surface using quantitative EDXA analysis. The measurements were taken in the same location in the as-coated and oxidized samples. Results are given in Table 4. It was evident that both Al and Pt concentrations decreased after the oxidation exposures. The aluminum content of the coating decreases due to the continuous spallation, i.e. alumina mechanically removed from the surface, and reformation of the alumina scale. The platinum content decreased due to additional interdiffusion with the nickel base alloy. The nickel and aluminum concentrations can be extrapolated to 100% to determine the binary phase present as shown in Figure 5. Both oxidation samples form a beta nickel-aluminum phase (17 to 39 wt.% Al), with the 982°C (1800^oF) / 3000 hours and 1194°C (2100^oF) / 100 hours sample having 22.2 and 20.5 weight percent aluminum respectively.

Table 4. Platinum Aluminide Coating Composition after Oxidation Tests

Phase Analysis: The phase at the surface of the platinum aluminide coating after 982°C (1800^oF) / 3000 hours and 1194°C (2100^oF) / 100 hours oxidation testing was determined by x-ray diffraction, with results shown in Figure 8. The x-ray diffraction peaks after oxidation testing showed very low count rates (<300 CPS) due to interdiffusion between the coating and alloy. Only 2 – 3 peaks could be distinguished, however these peaks matched the first two peaks (110 & 111) found in the as-coated sample. Therefore, the surface of the platinum aluminide coating was determined to be a simple cubic (beta) nickel-aluminum phase, which is in agreement with the composition. The x-ray diffraction results are displayed in Figures 8.

1. Platinum Aluminum Coating Surface after 982°C (1800^oF) / 3000 hours

The combination of scanning electron microscope with EDAX capability, x-ray diffraction and microprobe techniques was used to determine the phases and compositions of platinum aluminide coatings in the top aluminum rich layer before and after oxidation testing. While the results were limited in some cases there was still sufficient data to conclude that the top layer of coating is a platinum rich beta nickel-aluminum phase.

Microprobe analysis was chosen to determine the platinum, aluminum, and nickel compositions at the coating surface and every 5 microns to determine the effects of diffusion during the high temperature CVD process. Unfortunately, the microprobe analysis produced questionable results. The questionable results were obtained due to many factors. The microprobe reading at the surface of the coated sample showed believable results. This surface reading was tested using a 10-micron beam spot size that gave sufficient counts of 1,000 to 10,000. The beam spot size was reduced to 2 microns to evaluate the specimen in cross-section, and this reduced the counts to the range of 0 to 3000. By increasing the testing time, sufficient counts may have been achieved to give better results. In addition, the smaller beam size tended to drift, and could be measuring a different portion of the sample. Finally, if platinum rich beta nickel-aluminum standards could have been made, they might have given more accurate results than platinum, nickel, and aluminum elemental standards.

In conclusion, a good understanding of the advantages and shortcomings of the various techniques was achieved. The Scanning Electron Microscope with EDAX capability is an excellent tool to determine coating microstructure, and give a reasonable qualitative compositional analysis. Microprobe is used to achieve a very accurate qualitative compositional analysis. X-Ray Diffraction is a very powerful technique to not only determine the phases present, but the crystal structures and unit cell spacing as well.

1. Petit, F.S. and Meier, G.H., 1984, "Oxidation and Hot Corrosion of Superalloys," Proceedings, Fifth International Symposium on Superalloys, The metallurgical Society of AIME, Seven Springs, Pennsylvania, pp. 651-687.
2. Goward, G.W., Boone, D.H., 1971, "Mechanisms of Formation of Diffusion Aluminide Coatings on Nickel Base Superalloys," ASM Transactions, Vol. 60, pp. 228-241.
3. Smith, J.S. and Boone, D.H., 1990, "Platinum Modified Aluminides – Present Status," Paper No. 90-GT-319 presented at the Gas Turbine and Aeroengine Congress and Exposition, Brussels, Belgium.
4. Erickson, G.L., 1995, "A New, Third-Generation, Single-Crystal, Casting Superalloy, Journal of Metals, vol. 47, No. 4, pp. 36-39.
5. Alloy Phase Diagrams, Vol. 3, ASM International, 1992, Sec. 2, Pg. 49.
6. Lee, W, April 1998, "Formation of Grain Boundary in Diffusion Beta Nickel Aluminide Coatings by CVD," Presented at the AGTSR’s Metallic Coatings Specialty Workshop, Stevens Institute of Technology, Hoboken, New Jersey.
7. Arrhenius, G., Lecture Notes from Solids in Nature - MATS 290, University of California, San Diego, Winter Quarter 2000.

a = 5.14 angstroms

The dual peaks of the Face Centered Cubic (FCC) crystal represent deformation of the FCC unit cell. The unit cell is most likely tetragonal due to close spacing between the dual peaks and the planes where they occur. The dual peaks occur on the 200, 311, and 400 planes and therefore it is hypothesized that the 100 plane is deformed.

a = 2.91 angstroms

The 100, 200, 210 and 300 peaks (Figure 12) did not show up for x-ray diffraction of the platinum aluminide specimen surface. The coating grows in a columnar structure in the <110> or <211> direction as can be seen from Figure 4.(Ref. 6) Therefore the 100, 200, 210 and 300 peaks could not be generated at the crystal orientation from the coating surface.