S. J. Bullard
S. D. Cramer
G. R. Holcomb
U.S. Department of Energy, Albany Research Center, USA
TWI Ltd, Cambridge, United Kingdom
Paper presented at EPRI/DOE Fourth international conference on advances in materials technology for fossil fuel power plants. Hilton Head Island, S Carolina. 26-28 October 2004.
An iron aluminide (Fe3Al) intermetallic coating was deposited onto a F22 (2.25Cr-1Mo) steel substrate using a JP-5000 high velocity oxy-fuel (HVOF) thermal spray system. The as-sprayed coating was examined by electron microscopy and X-ray diffraction and was characterized in terms of oxidation and adhesion.
Fe3Al-coated steel specimens were exposed to a mixed oxidizing/sulfidizing environment at 500, 600, 700, and 800°C for approximately seven days. The gaseous environment consisted of N2-10%CO-5%CO2-2%H2O-0.12%H2S (by volume). All specimens gained mass after exposure to the environment and the mass gains were found to be inversely proportional to temperature increases. Representative specimens exposed at each temperature were cross-sectioned and subjected to examination under a scanning electron microscope (SEM) and X-ray mapping. Results are presented in terms of corrosion weight gain and corrosion product formation. The purpose of the research presented here was to evaluate the effectiveness of an HVOF-sprayed Fe3Al coating in protecting a steel substrate exposed to a fossil energy environment.
Fossil energy environments are among the most aggressive high-temperature environments in industry. These environments range from very high temperatures and complex gas mixtures, such as near coal combustion and coal gasification burners, to lower temperature environments complicated by the presence of ash deposits, such as on waterwalls in coal-fired boilers.
Boiler tubes for fossil environments typically are fabricated from low alloy carbon steels, with chromium and molybdenum as the primary alloy additions. Although chromium is expected to impart corrosion resistance to high-temperature alloys through the formation of a protective oxide layer on the surface, its concentration in boiler tube alloys (typically up to 2.25 wt%) is considered not sufficient to form a protective external scale. While it maybe possible to find or develop improved standalone monolithic alloys that can possess the required mechanical and corrosion resistance properties, it can be more cost effective to use a substrate material with the proper mechanical properties and protect it from aggressive environments by using high-temperature coatings.
Coatings are generally applied as a layer of a protective material onto a low-cost material (the substrate) and can be applied by a variety of physical, chemical, or electrochemical deposition processes. The physical thermal spray technique, such as a high velocity oxyfuel (HVOF) process, has received much attention in the recent years. In the current study a JP5000 HVOF system was used. For the Fe3Al coatings that are of interest to this report there have been studies of the corrosion of the intermetallic alloy applied by thermal spraying [1-4] by weld overlay  , and also of the intermetallic alloy in bulk form [6-8] .
The work presented here was a collaborative effort between the US Department of Energy's Albany Research Center (ARC) and The Welding Institute, Inc. (TWI). All of the work for preparation and pre-exposure analysis of the coated alloys was done by TWI. All of the corrosion tests and post-exposure analyses was done by ARC.
The substrate material selected for this research was F22 steel (nominally 2.25 Cr-1 Mo). This steel was chosen because of its wide application in fossil fuel-fired power plants. The analyzed composition of the F22 alloy, as determined by optical emission spectroscopy (OES), is given in Table 1. F22 is the forged version of the T22 (tube) and P22 (pipe) forms of the alloy.
Table 1: Elemental composition of F22 substrate steel obtained by OES analysis
The coating material used in this study was supplied by Osprey Metals Ltd, UK. It was an inert (Ar) gas atomized powder for HVOF spraying, with a nominal particle size distribution of -45 to +25 µm. The supplied powder sample was analyzed by titration and gravimetric analysis.
Disc specimens of the steel to a 25 mm dia. x 5.8 mm thickness were cut and machined prior to the surface preparation for HVOF spraying. This was followed by degreasing and grit blasting using angular alumina grit type F60 (nominalsize 0.25-0.3 mm). The substrate surface was preheated to a temperature between 70-80°C using a hot air blower prior to coating deposition. A batch of 10 disc specimens were attached to each other using a washer as spacer in between and rotated on a lathe. The spray gun was aligned perpendicularly and directed towards the axis of the rotating discs. The spray gun was scanned across the rotating discs in a series of horizontal passes and the circumference was coated. After this operation, flat faces were coated using a separate program. For this, the disc specimens were fixed on a rotating turntable, while the spray gun was scanned across the flat surface of the rotating discs in aseries of vertical passes. All faces of the disc specimens were coated to a nominal thickness of 500µm. The JP5000 system from Praxair Surface Technologies was used for the deposition of the Fe3Al coating and the parameters used for spraying are given in Table 2.
Table 2: Parameters used to prepare iron aluminide coatings
|Oxygen supply pressure
|Fuel (Kerosene) supply pressure
|Carrier gas (Ar) flow rate
|Oxygen flow rate
|Fuel flow rate
|Powder feed rate
|Vertical scan rate
|Number of passes
|Maximum surface temperature
Characterization of as-prepared coating
Standard metallographic techniques were used to prepare powder and coating cross sections ready for SEM examination. The oxygen content in the initial powder and in the sprayed coating was measured using the inert gas fusion technique with Leco TC-136 equipment. The X-ray diffraction (XRD) technique comprised of collecting the spectrum, using CuK α radiation, from samples of the powder and the coating detached from the substrate. The powder and coating samples were scanned over a range of 2 θ from 10° to 80° with a counting time of 1 second per 0.02° 2 θ step. The equipment used was a Philips X-ray diffractometer. Coating adhesion was measured using the tensile pull method in accordance with the test procedure ASTM D4541 'Standard Test Method for Pull-off Strength of Coatings Using Portable Adhesion Tester'. The adhsive used for bonding the coating surface and the test dolly was a high strength epoxy 'FM1000' supplied by CytecIndustries.
High-temperature corrosion testing
The high-temperature corrosion behavior of the coated specimens was investigated by exposing them to a mixed oxidizing/sulfidizing environment in a tube furnace. The gaseous environment was identical to one originally used by Marderet al. [1-3] This was done with the intention of being able to compare the corrosion results of the weld overlay Fe3Al coatings reported in reference 5 to the HVOF-sprayed Fe3Al coatings reported here in similar environments. The gas composition was N2-10%CO-5%CO2 -2%H2O-0.12%H2S (by volume). This gas mixture reportedly  has values of Log PO 2 = -19 and Log PS 2 = -8.
Test specimens were exposed at gas temperatures ranging from 510°C to 793°C. The specimens were placed in the tube furnace, heated to the required temperature under flowing N2 (414 mL/min), and then exposed for 160 hours to the mixed oxidizing/sulfidizing gaseous environment. The total gas flow rate for the test environment was 500 mL/min. The experiment was stopped by replacing the mixed gas with flowing N2 and then allowing the furnace to cool down to room temperature.
Corrosion product morphology
Corrosion samples intended for examination on the SEM were first coated with epoxy to keep the sample intact, cut near the center of the specimen to produce a cross section, mounted metallographically in epoxy, and then polished.
The SEM used to analyze the samples was a JEOL 7000F with field emission electron gun (FEG) and with built-in energy dispersive spectroscopy (EDS) and wavelength dispersive spectroscopy (WDS) facilities.
Results and discussion
The titration and gravimetry analysis confirmed that the powder contained 14.3 wt% Al and 85.3 wt% Fe. Since the stoichiometric composition of Fe3Al is 13.87 wt % Al, this powder had a slight excess of Al. Figure 1 shows that the XRD pattern of the iron aluminide powder matches the peaks for the Fe3Al and AlFe standards.
A secondary electron (SE) image of several powder particles is shown in Figure 2. This image shows that the particles are mostly spherical in shape typical of those produced by inert gas atomization and with diameters that agree well with the size distribution data in Table 3.
Table 3: Powder particle size distribution (data supplied by Osprey Metal Powders)
|Powder type||Wt % of powder in size range|
Table 4: Measured Oxygen content of the initial powder and final sprayed coating
|Identification||Oxygen wt%||Mean oxygen wt%|
| || || |
A backscattered electron (BSE) image of a cross section of a single Fe3Al particle in Figure 3 shows the particle to be dense, homogeneous, and uniform in atomic number contrast. This suggests that the powder particles are single-phased. The measured level of oxygen content in the powder particles was very low at about 0.01 wt% as shown in Table 4.
The morphology of the as-sprayed Fe3
Al coating surface produced using the JP5000 HVOF system is shown in Figure 4
. The surface appearance is rough and typical for HVOF sprayed coatings produced with a low combustion gas temperature, and suggests that the coating is formed primarily due to deformation in which smaller sized particles appear fully deformed and larger particles are only partially deformed. A BSE image of a coating cross section is presented in Figure 5
. The coating microstructure is typical of JP5000 sprayed coatings comprising partially deformed layers of powder particles 'lamellae' parallel to the test piece surface.
The SEM image of the cross section shows a very low coating porosity (typically <1% when measured using the image analysis software), and some splat boundary and fine gray contrast features well distributed in the coating. The cross section also shows a low level of particle oxidation due to HVOF spraying with a measured oxygen level of 0.25wt%. These gray contrast features are believed to be fine oxide stringers. The cross section also displays a well-bonded coating/substrate interface. This was confirmed by tensile adhesion measurements which gave a mean coating/substrate bond strength value of 24Mpa ( Table 5) with the failure occurring predominantly at the coating/substrate interface (coating adhesion) with some failure occurring between particle splats (coating cohesion). The coating thickness of 475-525µm was typical and was close to the target thickness of 500µm.
Table 5: Adhesion of the sprayed coatings
|Coating||Tensile failure stress, MPa||Mean||Mode of failure|
|a||b||c||A, %||B, %|
a, b, and c = different measurements
A - coating/substrate adhesion; B - coating cohesion
% values are estimates from visual examination
X-ray diffraction patterns ( Figure 6) of the as-deposited Fe3Al coating are similar to that of the starting powder ( Figure 1), except for three minor peaks which are no longer clear above the background at 2 θ values of 31, 55, and 73°. This was probably due to the HVOF spraying process.
High-temperature corrosion behavior
The results of exposing the Fe3Al coated steel substrates to the mixed oxidizing/sulfidizing environment from 510 to 790°C are summarized in Figure 7. The data show that all specimens gained weight and that the weight gain was inversely proportional to temperature increase. Data reported previously  for a weld-overlay coating of Fe-14.4Al-2Cr would extrapolate to more than 5 mg/cm2 weight gain after the 160 hours used in this report. Even though the referenced alloy  contained from 2 to 5 % Cr and the exposure time was shorter, the weight gains are of the same order of magnitude (in fact almost the same) at 500°C. This suggests that the HVOF coatings have corrosion resistance comparable to that of the weld-overlay Fe3Al alloy. Bulk Fe 3 Al (Fe-28 at % Al) alloys exposed to an Ar-H2S-H2-H2O environment (Log PO 2 = -20 and Log PS 2 = -6) gained approximately 0.2 mg/cm 2 after a 290 hour exposure at 800°C  .
BSE images of the thermal spray Fe3Al coatings after exposure to a mixed oxidizing/sulfidizing environment at 510 and 790°C are shown in Figure 8 with element x-ray maps for sulfur and oxygen. The BSE images of the coating cross sections after environmental exposures at both temperatures show a clean coating/substrate interface (without substrate corrosion) at both temperatures. Although corrosion attack of the coating exposed at 510°C is clearly visible with an abundance of gray-contrast phases and darker contrast stringers around splat boundaries, such features were not present in the as-sprayed coating prior to environmental exposures, Figure 5. The atomic number contrast in the BSE images suggests the presence of elements with lower atomic number than iron or aluminum. The oxygen maps in Figure 8 suggest that the gray contrast regions represent oxides possibly due to high-temperature oxidation whereas the darker contrast stringers seen in the BSE images are cracks between particle splats. These cracks are extended to about 300 µm outward from the substrate. Analysis of higher magnification oxygen, iron, and aluminum x-ray maps (not shown) suggests that these areas represent aluminum oxides. The coating exposed at 790°C (Figure 8) had very few such features and displayed a microstructure very close to the one prior to exposure (Figure 5). The BSE image of the cross section of the coating exposed at 790°C also had (Figure 8) a continuous gray contrast layer on the as-sprayed surface of the coating (not clearly see in Figure 8).
A comparison of the oxygen and sulfur x-ray maps in Figure 8 can be used to suggest an explanation for the corrosion results presented in Figure 7 which showed that weight gain due to corrosion decreased with increasing temperature while the environment was kept constant. The x-ray maps in Figure 8 suggest that the corrosion attack changes from sulfidation along splat boundaries of the outer coating (approximately 150 µm) to oxidation of the inner coating surface (approximately 300 µm). This trend to more oxidation and less sulfidation at higher temperatures is consistent with the oxygen and sulfur x-ray maps for specimens exposed at intermediate temperatures of 590 and 710°C (not shown). At further higher temperatures around790°C, a continuous layer of oxide film is formed (see oxygen map in Figure 8) that is believed to seal the surface of the HVOF coating thus preventing further penetration and attack by sulfidation. An examination of aluminum x-ray maps for the samples at 510 to 790°C suggests that the oxide formed on the outer surface is an aluminum oxide.
A low porosity (<1%), homogeneous and uniformly thick layer of Fe3Al intermetallic coating with little oxidation (~0.25 wt%) and good coating adhesion (~24 MPa) was deposited onto F22 steel using an inert gas atomised Fe3Al powder with a JP5000 HVOF system.
A F22 steel substrate can be well protected in a mixed oxidizing/sulfidizing environment (typical of a fossil energy environment) at a range of temperatures (510-790°C) using the JP5000 sprayed Fe3Al coating of 500 µm nominal thickness. Weight gains resulting from the reaction of the coating with the gaseous environment decrease with increasing temperature from 510 to 790°C.
EDS x-ray maps suggest that sulfur penetrates the outer surface splat boundaries at lower temperatures but is prevented from doing so at higher temperatures due to the formation of a continuous and protective oxide film on the outer surface, believed to be an aluminium oxide.
- K R Luer, J N DuPont, and A R Marder, 'High-Temperature Sulfidation of Fe3Al Thermal Spray Coatings at 600°C', Corrosion, 56(2), 2000, pp. 189-198.
- T C Totemeier, R N Wright and W D Swank, 'Microstructure and Stresses in HVOF Sprayed Iron Aluminide Coatings', J of Thermal Spray Technology, 11 (3), September 2002, pp.400-408.
- A Magnee, E Offergeld, M Leroy and A Lefort, 'Fe-Al Intermetallic Coating Applications to Thermal Energy Conversion Advanced Systems', Proceedings of the 15th International Thermal Spray Conference, Nice, France, May 1998, pp.1091-1096.
- S T Bluni and A R Marder, 'Effects of Thermal Spray Coating Composition and Microstructures on the Coating Response and Substrate Protection at High Temperatures', Corrosion, 52 (3), March 1996, pp.213-218.
- J R Regina, J N DuPont, and A R Marder, 'Corrosion Behavior of FeAlCr Weld Overlay Coatings in High-Temperature Oxidizing/Sulfidizing Environments,' Proceedings of the EPRI International Conference on Materials and Corrosion Experience for Fossil Fuel Power Plants, Hilton Head, SC (USA), Electric Power Research Institute, November 2003.
- J R Regina, J N DuPont, and A R Marder, 'Corrosion Behavior of Fe-Al-Cr Alloys in Sulfur- and Oxygen-Rich Environments in the Presence of Pyrite,' Corrosion, 60(5), 2004, pp. 501-509.
- B A Pint, K L More, P F Tortorelli, W D Porter, and I G Wright, 'Optimizing the Imperfect Oxidation Performance of Iron Aluminides,' Materials Science Forum, 369-372, 2003, pp. 411-418.
- J H DeVan, 'Oxidation Behavior of Fe3Al and Derivative Alloys,' Oxidation of High Temperature Intermetallics, ed. T Grobstein and J Doychak, The Minerals, Metals & Materials Society, 1988, pp. 107-115.