Steam oxidation resistance and performance of newly developed coatings for Haynes® 282® Ni-based alloy

Accepted Manuscript Title: Steam oxidation resistance and performance of newly developed coatings for Haynes® 282® Ni-based alloy Authors: T. Dudziak, L. Boron, A. Gupta, S. Saraf, P. Skierski, S. Seal, N. Sobczak, R. Purgert PII: DOI: Reference:

S0010-938X(17)31510-X https://doi.org/10.1016/j.corsci.2018.04.005 CS 7470

To appear in: Received date: Revised date: Accepted date:

18-8-2017 30-3-2018 2-4-2018

Please cite this article as: Dudziak T, Boron L, Gupta A, Saraf S, Skierski P, Seal S, Sobczak N, Purgert R, Steam oxidation resistance and performance of newly developed coatings for Haynes® 282® Ni-based alloy, Corrosion Science (2010), https://doi.org/10.1016/j.corsci.2018.04.005 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Steam oxidation resistance and performance of newly developed coatings for Haynes® 282® Ni-based alloy

T.

Dudziak1,

L.

Boron1,

A.

Gupta2,

Saraf2,

S.

P.

Skierski1,

S.

Seal2, N.

Sobczak1,

1

Foundry Research Institute, Zakopianska 73, 30-418 Krakow, Poland

2

IP T

R. Purgert3

University of Central Florida, Advanced Materials Processing and Analysis Centre (AMPAC) and

SC R

Materials Science & Engineering, P.O. Box 162455 Orlando, FL 32816 USA 3

U

Energy Industries of Ohio, Park Centre Plaza, Suite 200, 6100 Oak, Tree Boulevard, Independence, OH

A

N

44131

Highlights

 

M

CC E



ED



Steam oxidation study on bare and ceramic coated Haynes® 282® alloy at 800 oC for 2000 hours performed Nano CeO2, Al2O3, YSZ and Al2O3-Graphene oxide coatings with/out bond coat (Ni-5wt.% Al) deposited on Haynes® 282® alloy using air plasma spray High degree of internal oxidation observed in Haynes® 282® alloy Plasma sprayed nano CeO2 coating with no bond coat reduced internal oxidation process extensively Al2O3 based coatings showed poor adhesion and detached from the substrate after 250 hours

PT



Abstract

A

Steam oxidation test was carried out at 800oC for 2000 hours on nine plasma sprayed coatings on gamma-prime (’) strengthened Haynes® 282® Ni-based alloy. The lowest oxidation kinetics and extensive reduction in internal oxidation process was observed in the sample coated with nano CeO2 with no bond coat. This is due to the oxygen deficiency in CeOx that facilitates the inward movement of oxygen arising from the presence of Ce3+ oxidation state and ensures the quick formation of protective, impermeable and adherent layer of Cr2O3. Keywords: A. nickel alloy; ceramic; B. SEM; XRD; C. oxide coatings; oxidation,

1

1. Introduction

M

A

N

U

SC R

IP T

Development of new technology, as well as advancement of existing technologies, requires innovation in mechanical design, functionality and automation. It also requires new materials development and improvement of existing materials to meet the challenges. The new materials developed for energy, aircraft industry, aerospace and propulsion systems must show high-temperature strength and corrosion resistance. One of the most valuable materials for harsh conditions at high temperatures are Ni-based alloys where adequate creep strength and corrosion resistance is required. As the European Union laws related to CO2 emission are becoming more stringent each year, new materials with higher corrosion resistance than existing ones are required to meet the challenges. It is well known that each 1% increase in overall efficiency can result in as much as 3% reduction in CO2 emissions [1]. Therefore, the efforts have been focused on developing ultra-supercritical thermal power plants to enhance the efficiency by increasing the steam temperature and reducing the amount of coal/gas required. The steam temperature in conventional power plants is expected to rise by 50oC to 100oC by 2030 in order to reduce the CO2 emissions further. In coal-fired power plants, the most severe environment is presented by steam oxidation where accelerated degradation of the exposed materials is often met. Oxide formations in boilers and reheaters can result in dangerous conditions such as local overheating, plugging and steam starvation when exfoliated/detached from the base materials. Recently, various steel grades with different amount of Cr concentration have been tested [2, 3]. The 2%Cr, 9%Cr and 12%Cr ferritic steel were preoxidized under 100% dry steam conditions at temperature 550 - 750C. The thickness of the grown oxide was ranged from 15–1150μm, 30–450μm and 25–60μm for 2%Cr, 9%Cr and 12%Cr steel, respectively. 9%Cr and 12%Cr steels showed excellent exfoliation and spalling resistance when subjected to different tensile strain and only through-scale cracking observed. On the other hand, pre-oxidized 2%Cr steel with oxide thickness >380μm scaled off from the specimen without any tensile strain due to the substrate-scale interface failure.

A

CC E

PT

ED

Another group of material with adequate properties to be used in modern coal-fired power stations with high-temperature capability equivalent to 800oC are Haynes® alloys. The Haynes® 282® and Haynes® 230® are the most promising ones. However, these materials undergo severe internal oxidation process when subjected to oxidative conditions as reported recently [4, 5]. Jang et al. have studied the oxidation behavior of Haynes® 230® alloy under air, hot steam and helium environment at 900oC [6]. Haynes® 230® alloy was less susceptible to the change in oxidation environment. Haynes® 230® and 282® alloys are also studied to investigate the metallurgical changes at high temperature for applications in ultra-supercritical steam power plants [7]. The ingress of oxygen via external oxide scale and oxide formation at the grain boundaries cause the internal oxidation of Haynes alloys. The internal oxide consists of Al2O3 and TiO2 phases, which are likely to form from gamma–prime (’) strengthened alloy (Ni3(Ti, Al) phase) [8]. In general, internal oxidation process limits the service life of an alloy. Applying an oxide coat on the steel and other high-temperature structural material has resulted in a significant alterations in oxidation behaviour. Oxide coatings isolate the metal substrate and prevent the direct contact of steam or surrounding oxidizing environment; thus slowing down the rate of oxidation of metal substrates. Polymer [9, 10], oxide, boride [11] and silicide [11] based coatings have been utilized to prevent oxidation of metal substrates [12, 13]. Seal et al. [14] have deposited cerium oxide and rare earth doped nanocrystalline ceria coatings via dip coatings and annealing method on AISI 304, AISI 316 and AISI 321 stainless steel at a temperature as high as 1423K [13]. These coatings have shown excellent corrosion resistance under dry conditions [12, 14, 15, 16]. Al2O3hydroxyapatite composite coatings have been deposited on 316 stainless steel using pulse laser deposition to prevent biocorrosion and improve mechanical properties [12]. In recent years, plasma

2

spray deposition has become a preferred method of coating deposition because of minimal surface preparation, thickness control, and fast process. Nanocerium oxide coated martensitic stainless steel have shown a significant decrease in oxidation kinetics at 1000oC deposited by solution precursor plasma spray route [17]. Therefore, in this research article nine ceramic coatings (S1 – S9), have been deposited on Haynes® 282® substrates to reduce their internal oxidation and extent their service life. The coating – substrate systems have been tested in unique steam facility rig, where deionized water has been circulated in a closed-loop system in order to generate 100% steam conditions.

2. Experimental procedure

IP T

2.1 Materials

SC R

In the study, 10 samples in total have been exposed to high-temperature steam environment including gamma-prime (’) strengthened Haynes® 282® Ni-based alloy. Bare Haynes® 282® alloy served as a reference sample for 9 coated Haynes® 282® coupons with different coatings. Table 1 shows chemical composition of Haynes® 282® alloy according to certificate of analysis received from the producer (Haynes International). For better traceability of the coated Haynes® 282® alloy samples, each sample received unique code listed in Table 2.

U

The rectangle samples with dimensions 10x12mm and a thickness of 4–5mm have

N

been used in the long-term steam oxidation tests. Prior to the experiment, the uncoated

A

sample (reference sample) has been polished using 600 grid SiC paper. Further, the

M

dimensions of reference sample and other samples have been measured using a micrometer and ultrasonically cleaned in isopropanol for 15 minutes at 40oC in order

ED

to degrease the samples and remove other impurities. The samples before and during high-temperature exposure in the steam atmosphere have been accurately weighed

PT

using a digital balance with a resolution of ±0.01mg for masses m<80g. The balance

CC E

was calibrated frequently using its internal calibration function and periodically with test weights.

2.2 Plasma spray deposition of coatings

A

The Haynes® 282® alloy substrates were grit blasted using -100 to +400 mesh alumina grit followed by cleaning with acetone. No pre-treatment was given to the substrate before plasma spray deposition. Coatings of aluminum oxide (Al2O3), aluminum oxide4wt.% reduced graphene oxide (RGO), 8 mole% yttria-stabilized zirconia (YSZ), cerium oxide (CeO2) and Ni-5%Al alloy (Ni185) were deposited using Praxair SG-100 3

plasma gun. Plasma spray parameters used for different coating depositions are summarized in Table 3. Ni-5% Al coating as a bond coat

IP T

The purpose of Ni-5%Al coating is to understand the effect of bond coat on the oxidation resistance of coatings. Therefore, two samples of each coating were prepared, i.e. (i) with bond coat and (ii) without bond coat. The Ni-5%Al coating was deposited using Ni185 powder procured from Praxair Surface Technologies, USA, which is a solid solution of Al in Ni. Its particles were irregular in shape and particle size was in the range of 45-90μm. Cerium oxide coating

SC R

Cerium oxide powder with powder particle size in the range of 50–60nm was procured from Inframat Advanced Materials, USA. The powder was spray dried using Buchi mini dryer B-290 for better flowability. Parameters similar to Ni-5%Al coating were used to deposit cerium oxide coatings. 8 mole% YSZ coating

N

U

The ZrO2 – 8%Y2O3 (YSZ) powder was procured from Praxair Surface Technologies, USA (trade name ZRO-182) and used in as-received condition (agglomerated and sintered). The particle size according to safety data sheet (SDS) was in the range of 125µm/+22µm, means D90 of the particles is 125µm and D10 of the particle is 25µm. Powders come in a normal distribution.

A

Al2O3 coating

Al2O3-4wt% RGO coating

ED

M

The Al2O3 powder was also procured from Praxair Surface Technologies, USA (trade name ALO-101) and used in as-received condition. The particle size according to safety data sheet (SDS) was equivalent to 45µm.

CC E

PT

The Al2O3 powder was procured from Praxair Surface Technologies, USA (trade name ALO-101). Reduced graphene oxide (RGO) was supplied by Garmor Inc., USA. Both powders were used in asreceived condition. ALO-101 and RGO powders were mixed using jar mill at 100rpm. Powder to ball ratio of 6:1 was maintained during jar mixing. Low rpm and powder to ball ratio were selected to minimize the damage to RGO. 2.2 Steam oxidation

A

Schematic of high-temperature steam oxidation rig used in the study is presented in Fig. 1. In this type of steam oxidation test set-up [18], 100% pure steam is generated by pumping water from a reservoir placed underneath the furnace. In the furnace, steam passes over the test samples with a flow rate of 2.833ml/min (1 rotation per minute (rpm) and flows into a condenser before the water returns to the reservoir. Double deionized water was used in the reservoir. The whole system was sealed using stainless steel flanges at both ends, prior to the test. The whole system was purged for 2 hours using oxygen-free nitrogen (OFN). This purge continues through the water reservoir throughout the samples exposure period to minimize the level of oxygen in the system. Prior to the exposure, furnace calibration was required in order to detect hot zone. The calibration process ensured the placement of the samples in the furnace at a test temperature with accuracy ±5 oC. In this study, the test was

4

conducted for 2000 hours and data were collected at the intervals of every 250 hours. The samples were inserted into the hot zone of the furnace using a calibrated stick with the measured distance, in order to place the samples on the Al2O3 plate within the hot zone. 2.3 XRD analysis

 ln(1  Gx)sinθ 2μ

(1)

SC R

X

IP T

The XRD investigations have been conducted by means of D500 Kristalloflex with monochromatic Xray sources Cu (K = 1.54Å) and EMPYREAN Panalytical with Cu X-ray source using Ni filter. The phase analyses on the uncoated Haynes® 282 sample has been performed using two techniques: BraggBrentano (BB) geometry and the geometry of constant angle called grazing incidence using =1o and =3o. In BB geometry, the penetration depth of X-rays can be estimated using the following formula:

where:

Gx - donates the intensity of the primary X-ray, giving important information related to

U

irradiation volume,

N

 - is half of the scattering angle

A

 - linear absorption coefficient.

M

In the grazing angle method, X-rays penetration depth was calculated via the following formula:

where:

ED

 ln(1  Gx)   1 1  ] μ[  sin α sin(2θ  α ) 

(2)

PT

X

CC E

Gx - donates the intensity of the primary X-ray, giving important information related to irradiated volume, this value is equivalent to 0.95 (95%)

A

 - linear absorption coefficient  - grazing incidence angle

The calculated values of Gx for Haynes® 282® where the assumption of Gx = 95% are shown in Table 4.

5

3. Results and Discussion Fig. 2 shows initial states of the surfaces macrostructures captured by using DSLR Canon EOS 70D camera coupled with Canon MP-E 65mm f/2,8 Macro lens. Each of the coatings had a size of around 1 x 1.2cm with a thickness of 0.4cm. The coatings S1 – S9 in as-deposited state showed good adherence to the substrate, uniform macrostructure with no cracks, defects and other deviations. The coatings S1, S2, S4, and S8 has some black spots present from the plasma spray deposition process. The Hitachi TM300 tabletop Scanning Electron Microscope (SEM) in Backscatter Electron mode (BSE) with 15kV accelerated voltage has been used to demonstrate the surface microstructures as shown in Fig. 3.

A

CC E

PT

ED

M

A

N

U

SC R

IP T

The presented microstructures of as-deposited coatings confirm macro findings. The coatings showed a lack of decohesion with the substrate. However, samples S1, S2, S6, S8 and S9 showed slightly different microstructure, where more fine grain particles could be observed. Those particles were not melted during plasma spray deposition. In contrast, S3, S4, S5 and S7 (Al2O3 based coatings) samples indicate fully melted particles on the surface. In addition, it has been found based on EDS analyses that in the S8 sample (Cerium oxide (CeO2)), the black starry-like patches showed enrichment of Cu deriving from SG-100 gun anode and cathode during deposition process, whereas in S1, S2, S4. S6 and S7, the surface was contaminated by Ca enriched spots. The surfaces showed different roughnesses; the most uneven surfaces corresponded to S1, S2, S9 samples while the most even surfaces have been observed in S3, S4, S5 and in the reference sample S10 (not shown here). Finally, EDS chemical composition analyses, performed on the as deposited samples, confirmed the chemical composition stated on delivery notes.

6

3.1 Oxidation Kinetics Fig. 4 shows kinetic data for the exposed samples coated with ceramic based coating applied on the surface of Haynes® 282® Ni-based alloy.

SC R

IP T

The kinetic results presented in Fig. 4 clearly demonstrate that all the tested materials indicated similar behaviour in terms of mass change. The highest mass gain has been recorded for S9 sample (1.38mg/cm2) and the lowest one for S8 sample (1.13mg/cm2). Surprisingly, the uncoated sample presented almost the lowest mass gain under steam oxidation conditions showing the final reading of 1.17mg/cm2, hence the total difference between the most resistant sample and that of the lowest resistance showed the value of 0.25mg/cm2. In the samples S1, S6 and S8 some steady state oxidation process or reducement of mass gain compared to other exposed materials has been found between 500–1250 hours. The observed behaviour could be a result of protective semi layer formation at the coating – substrate interface or at the coating bond coat interface, however, after 1250 hours of exposure mass gain of the exposed samples have again been clearly observed.

ED

M

A

N

U

In addition, for some samples, the test conditions were too harsh; therefore, the samples S3, S4, S5 and S7 have been removed from further tests after 250 hours of exposure due to detachment of the plasma sprayed coating. Detachment occurred when high levels of internal stresses appeared due to incompatibility deriving from the coefficient of liner thermal expansion (CTE) between the coating and the substrate. Additionally, plasma spray deposited coatings contain compressive residual strain due to the high velocity (200-300 m/s) impact of semi-molten particles with a substrate and high cooling rate (108K/s). The magnitude of residual strain increases with increasing thickness [19, 20]. Detached coatings had higher thickness compared to coatings that did not detached from Haynes® 282® Ni-based alloy (thickness shown later). The other coatings have been tested till the final exposure of 2000 hours. The calculated values for n factor indicate that the mass gain values cannot be fitted into the parabolic rate law. The calculated n factor using slope function where log(mass gain) vs. log(time) in seconds showed values other than 0.5 as should be observed for the trends following parabolic rate law. The coatings showed some deviations from the parabolic rate constant; the exponent n values indicate: S1 n = 0.24, S2 = 0.25, S6 n = 0.32, S8 n = 0.41, S9 n = 0.26 and finally S10 n = 0.40 as shown in Fig. 5.

A

CC E

PT

The findings indicate that the coated Ni-based alloy showed different mechanism governed high temperature oxidation degradation at 800oC for 2000 hours. Nevertheless, the S2, S6 and S9 coatings showed exponent factor with value of n = 0.26-0.32 indicating cubic time dependence law. According to Quadakkers et al. [21], the process in most cases can be explained by assuming that oxygen grain boundary diffusion is the dominating scale growth process in combination with an increase in oxide grain size with growth and perhaps with time. The following calculations showed as well that only two systems, cerium oxide (CeO 2) with no bond coat (S8) and Haynes® 282® alloy (S10), indicate nearly parabolic rate constant mechanism of oxide formation with n factor = 0.40. It is interesting to note that no bond coat in sample S8 showed high influence on the corrosion behaviour and oxide growth kinetics. The sample S8 with no bond coat showed near parabolic time dependence rate of oxide growth, whereas the sample S2 with the bond coat showed cubic time dependence law according to presented calculations. The kp values for cerium oxide with no the bond coat (CeO2) (S8) and Haynes® 282® alloy (S10) samples have been determined using data from Fig. 5, the results are shown in Fig. 6. The calculated values are listed in Table 5. The values clearly indicate that the more negative value of kp is associated with the uncoated Haynes® 282® alloy, however the difference between calculated values is not big. Hence it can be assumed that both,

7

cerium oxide with no bond coat (CeO2) (S8) and Haynes® 282® alloy (S10) samples, indicate the same mechanism of oxide scale formation in terms of kinetic law. 3.2 XRD analysis

IP T

Due to coating thickness deposited on the Haynes® 282®, the XRD analysis has been performed only on the uncoated Ni-based alloy. Furthermore, in order to obtain a full spectrum of the oxide scale formed during steam oxidation, the process of XRD analysis has been conducted in BB mode and by grazing incidence using =1o and =3o. Fig. 7 shows XRD pattern for the uncoated alloy after steam oxidation test at 800oC for 2000 hours. It has been found that under steam conditions, the alloy underwent the development of Cr2O3, TiO2 and MnCr2O4 phases mainly, with some (Ni) suggesting the detection of metal matrix phase. 3.2 Macro investigations of the exposed samples

CC E

PT

ED

M

A

N

U

SC R

The post exposure observations revealed that the coatings (S1, S2, S6, S8, S9) under steam oxidation atmosphere showed high resistance to crack formation and to steam oxidation process. Nevertheless, the Al2O3 based coatings S3, S4, S5 and S7 detached from the substrate indicating poor adhesion. Hence, the samples S3, S4, S5 and S7 were withdrawn after 250 hours from the further test at 800 oC. The other samples showed a lack of steam condition influence and the coatings showed strong bond to the metallic substrate with no sign of cracks and spallation (Fig. 8). It is believed that spallation of thick Al2O3 based ceramic coatings is related to the difference of coefficient of thermal expansion (CTE). In fact, CTE coefficients found for various ceramic phases indicate that high difference between Al2O3 and Haynes® 282® alloy exists (summarized in Table 6) [22,23,24,25]. It is interesting to note that the deposition of a bond coat between the coating and the substrate showed a lack of influence. This finding indicates that strains developed during cooling period from the high-temperature regime have been too high to be accommodated by the bond coat. Moreover, it is believed that another scenario besides strain relaxation could appear providing the detachment of the deposited Al2O3 ceramic coating. The Thermal Barrier Coating (TBC) effect of the metallic bond coat is to enhance the adhesion of the ceramic coating (TBC) to the substrate and provide additional corrosion resistance to the metal substrate. However, under conditions where high partial pressures are met, as under steam oxidation in pure water steam, extensive oxidation of the deposited bond coat may appear resulting the formation of Thermally Grown Oxide (TGO) [26], and the formation of the TGO with the critical thickness of 6–7m [27, 28]. Development of TGO with high oxidation rate may induce additional forces governing detachment mechanism of the ceramic coatings instead of protection of the base alloy against further oxidation via the formation of Al2O3 from TGO. The development of Al2O3 is essential in order to build up diffusion barrier against oxidation at high temperature. 3.3 Micro investigations of the exposed samples

A

The surface microstructures of the exposed coatings are shown in Fig. 9. The microstructures after steam oxidation at 800oC for 2000 hours show lack of cracks and other defects on the surface. It has been noted that the samples S1 and S6 (YSZ) have been covered by spherical particles rich in Si–O elements according to EDS investigations, suggesting the formation of SiO2 under steam oxidizing conditions. The other samples were free from SiO2 formation. The effect of SiO2 formation has been recently discused by Menna et al. [29] suggesting that Si is an impurity in YSZ coating. On the other hand, Wilhelm and Howarth [30] suggested that addition of Si to YSZ and the formation of SiO2 decreases sintering temperature required to obtain a relative densities higher than 95% in order to enhance mechanical properties or to regulate grain size and grain growth.

8

3.4 Micro investigations of cross-sectioned samples

PT

ED

M

A

N

U

SC R

IP T

Cross–sectional images of the samples exposed for 2000 hours steam oxidation test are shown in Fig. 10. The observed cross-sections captured after an exposure for 2000 hours at 800oC indicate different thickness of the plasma deposited coatings and oxide scale formed on the surface of the Haynes® 282® alloy. Since in this study, only one set of samples have been investigated, the thickness of the deposited ceramic layers have been unknown prior to exposure to steam environment. Nevertheless, looking at the SEM images it has been observed that every sample has different thickness. The samples S1 and S2 have showed thickness of 550-650m, S6 samples has been covered by 450–500 m thick ceramic layer. Finally S8 and S9 samples have been protected against steam oxidation condition via 30 and 50 m thick coating and thermally grown oxide layers, respectively. It has been noted that under steam oxidation conditions, the Haynes® 282® alloy undergoes high rate of internal oxidation process [31, 32, 33]. It appears that the penetration of oxygen beneath the TGO or underneath the oxide scale (samples without the bond coat) varied. The measurements have been performed based on scale bar attached to each BSE image. The measurements have shown that the in the S1 sample, internal penetration reached only 20–50m, in S2 the internal oxidation zone reached at least 250m, the S6, S8 samples have showed internal penetration via oxygen up to 100m, finally, the samples S9 and S10 severe internal oxidation of 30 and 25m, respectively. The variety of internal oxidation penetration depends mainly from the coating porosity and permeability for oxygen molecules. Since the deposited coatings showed high porosity level, it can be stated that oxygen from dissociation of H2O molecules diffused inwardly through the porous coatings. In order to prove difference in oxygen permeability with the oxide structure EDS X-Ray mappings were performed instead of line scan EDS concentration profiles. The EDS X-Ray mappings are presented in Figs. 11 – 13. The mappings were performed on three different samples, S2, S8 and S10 only. The following samples have been chosen because: CeO2 with no bond coat (S8) sample showed the lowest weight gain according to kinetic data shown in Fig. 4, CeO2 with the bond coat (S2) has been shown different kinetic law, whereas S10 is the uncoated Haynes® 282® sample. It is also evident that the sample with no bond coat (S8) developed dense oxide scale scattered within the CeO2 matrix where Cr diffused outwardly from the Ni based metal matrix consuming (reducing pO2 pressure) O from ambient atmosphere (steam); hence the process was controlled by inward O diffusion and Cr outward diffusion to CeO2 structure. In the case of sample S2 with the bond coat, the O from an ambient atmosphere was consumed only for the oxide scale formation at the CeO2–metal matrix interface, the oxide scale developed according to cubic law where oxygen grain boundary diffusion is the dominating scale growth process according to work presented by Quadakkers et al.[21].

A

CC E

In the case of the S2 sample, EDS X-Ray mapping in Fig. 11 using SEM in BSE mode presents the formation of several layers over the substrate. The TiO2 and Cr2O3 layers must be the dense and adhered native oxide layers on the substrate, suggesting similar mechanism as in the uncoated alloy and access of oxygen from ambient moist atmosphere with no restrictions. The formation of nickel oxide between enriched cerium oxide in the very top layer is very interesting. It shows that the formation of enriched Cr2O3 layer at the coating – substrate interface is not dense enough, therefore mass transport of high number of Ni ions to build up enriched NiO layer throughout that layer (Cr2O3) is not a problem. The formation of this enriched zone of NiO within the coating deriving from equal distribution of Cr within the CeO2 based coating, and inability of the formation of dense Cr2O3 oxide scale at the interface. In the S8 sample (Fig. 12), the formation of dense Cr2O3 oxide layer has been observed in the coating – substrate interface. The porous structure of CeO2 showed enough mass transport of oxygen to

9

SC R

IP T

develop dense, adherent Cr2O3 oxide scale with some enrichment of TiO2. Cerium oxide (CeOx) is a redox active material; therefore, both Ce3+ and Ce4+ oxidation states are present at the surface of the coatings. Plasma sprayed CeOx coatings and nanoparticles consist of 30-40% Ce3+ and the rest Ce4+ [34, 35]. Several mechanism have been reported for the high temperature oxidation protection of CeOx in literature [36]. Oxygen deficiency in CeOx facilitates the inward movement of oxygen due to the presence of Ce3+ oxidation state and helps in quick formation of protective, impervious and adhered layer of Ce2O3. [14]. Additionally, segregation of Ce3+ and Ce4+ at grain boundaries obstructs the migration of cations and prevents the further oxide scale formation at the alloy surface. Inward diffusion of oxygen and limited diffusion of Cr3+ and other cations prevent the generation of voids and therefore improve the adhesion of oxide scale with the alloy substrate and develop a strong interface at the oxide scale-alloy interface [36, 37, 38]. As a result of improvement in oxide scale adherence, the alloy system can be subjected for a greater number of thermal cycles without oxide scale layer rupture. In our present study, the samples were taken out of the stem oxidation furnace every 250 hours to measure mass change. Thus, alloy-coating systems were subjected to a total of 8 thermal cycles. No scale rupture or delamination have been observed further confirming the good adhesion of scale due to the CeOx coatings. Fig. 13 shows EDS X-Ray mappings on the uncoated Haynes® 282® alloy in steam oxidations. The performed mappings revealed that the alloy underwent oxidation by the formation of thick, dense Cr2O3 oxide scale with locally distributed enrichment of TiO2. Underneath the oxide scale, internal precipitation of Al and Ti oxide formed suggesting high inward oxygen flux throughout the Cr2O3 oxide layer. The oxide stability depends on activity, Gibbs free energy of formation ( G T ) concentration, and partial pressure of oxidizing gas and finally, diffusivity of an element. At 800oC when deionised water is transported from reservoir to the hot furnace, firstly the following reaction spontaneously occurs:

(3)

M

1 H 2 O(l)  H 2 (g)  O 2 (g) 2

A

N

U

o

ΔG H2O  230.000  8.14T  ln(T )  9.25T

ED

(4)

CC E

PT

Decomposition of water releases oxygen, which surrounds the external alloy surface, provide highest activity for the oxidation of elements. It was found that among several elements from the bulk alloy, the formation of three oxidation products have been observed including Cr2O3, TiO2 and finally MnCr2O4 which is further supported by XRD analysis (Fig. 7). These phases have been developed via reactions 5 – 8:

(5)

3 Cr(s)  O2 (g)  Cr2O3 (s) 2

(6)

A

1 Ti(s)  O 2 (g)  TiO2 (s) 2

1 Mn(s) O 2 (g)  MnO(s) 2

(7)

MnO(s) Cr2 O3 (s)  MnCr2 O 4 (s)

(8)

10

For the following reactions, Gibbs free energy of formation ( G T ) at the temperature of interest [39] o

can be calculated; the results are shown in Table 7. A chromia scale is forming due to the more negative value of Gibbs free energy of formation ( G T ) at 800oC, then a MnCr2O4 oxide develops with the formation of MnO as an intermediate step. The driving force for these reactions is the lowering of Gibbs free energy listed in Table 7. According to crystal field theory calculations, Mn2+ has a higher mobility than Cr3+ in Cr2O3 [40], which supports rapid transport of Mn2+ ions to the chromia scale and the formation of MnO which reacts with Cr2O3 to form MnCr2O4 according to reaction 8. Jang et. al. [6] have also reported the formation of a Cr2O3 (inner) and MnCr2O4 (outer) layers at 900C in air, hot steam and helium environment which improves the corrosion resistance [6]. The formation of different oxide layers has been observed when Ni based super alloys were subjected to high temperature oxidation at 850-1000C for 20000 hours in air [41]. Introduction of Mn or other elements forming spinel phases can lower the activity of Cr2O3, for example by substitution of Ni or Mn for Cr, or by the formation of MnCr 2O4 spinel. The spinel is essentially a matrix of Cr2O3 and MnO according to reaction 8. Furthermore, Mn reduces the activity of Cr in the oxide scale, either from solid solution replacement of chromium with manganese when low levels of manganese is present or from the formation of manganese-chromium spinel MnCr2O4 when high levels of manganese are added. The beneficial effect of reduction in chromium activity leads to a predicted reduction in chromium evaporation by as much as a factor of 35 at 800°C and 55 at 700 °C [42]. Due to the MnCr2O4 spinel formation Cr evaporation is highly reduced in Ni-Cr rich alloys. A similar process of Cr evaporation from Cr rich scale is well known in Fe-Cr rich materials where steam environment invokes evaporation of Cr via the formation of CrO2(OH)2 from the scale consisting Cr2O3 [43]. It should be indicated that chromia evaporation can proceed only when high partial pressure of oxygen can be met [44]. Chromia evaporation via the formation of CrO2(OH)2 phases is not considered in this study due to low partial pressure of oxygen derived from H2O. According to Opila [45], in steam atmospheres with a low oxygen partial pressure, there is minimal formation of the volatile Cr oxyhydroxide with no influence on the corrosion behavior due to a low oxygen partial pressure in steam. The following discussion of steam oxidation performance of the coated Haynes® 282® needs to cover as well by the formation of TiO2 phase found by X-ray diffraction analyses. The formation of TiO2 can be rationalized by thermodynamic principles as well. When Ni based alloys such as Haynes® 282® containing 2.1wt% Ti are exposed at high temperature, it is expected that TiO2 phase is possible to

PT

ED

M

A

N

U

SC R

IP T

o

CC E

form. The formation of TiO2 is dictated by the most negative Gibbs free energy formation ( G T ) from the phase of interest (Cr2O3, TiO2, MnCr2O4) according to Table 7. Bearing in mind that oxygen partial pressure is high enough for the formation of the phases (Cr2O3, TiO2, MnCr2O4), calculated activities for o

Ti and Cr only (MnCr2O4 due to high Gibbs free energy formation ( G T ) has not been taken into account) clearly show that Ti possesses higher activity than that for Cr at 800oC. The activities of the elements have been calculated using the following formulas:

A

G OTiO2,T   RT ln K TiO2   RT

o

aTiO 2 aTi  pO 2

G OCr2O32,T   RT ln K Cr2O3   RT

aCr2 O 3 (aCr ) 2  ( pO 2 ) 3 / 2

(9)

(10)

Fig. 14 shows the minimum activity graph where Log a vs. Log pO2 is present, the blue thick stripe indicates Log pO2 in water steam at 800oC according to work carried out by Dooley [43]. Based on Fig.

11

14, it has been found that Ti element has showed higher activity than that presented by Cr. Therefore, when Ni-based alloy with Ti and Cr is exposed to steam oxidation environment the following mechanism is occurring. Due to the large magnitude negative value of Gibbs free energy formation (

G To ) and higher activity, the formation of TiO2 is likely to form in the initial period of steam oxidation. However, after some time, when activity and concentration of Ti are reduced, highly concentrated Cr with slightly lower Gibbs free energy formation ( G T ) start to develop stable Cr2O3 phase. Further, the process obeys the formation of MnCr2O4 due to the higher mobility of Mn2+ than Cr2+ in Cr2O3 phase as mentioned previously. Furthermore, the process is active for every alloy exposed in this study, due to high porosity of the deposited coating and ingress of oxygen from the dissociated water molecule was, in fact, unlimited. Therefore, partial pressure of oxygen at the coating – substrate interface was the same as in the external coating surface exposed to steam atmosphere.

IP T

o

GTO = - 749.6KJmol

N

U

SC R

Finally, the process conducted under steam oxidation showed that the exposed coated alloy underwent different degrees of internal oxidation phenomena. The lowest degree was found to be in S1 coated Haynes® 282®, where YSZ with bond coat system effectively protected the alloy against internal oxidation process. The highest level of internal oxidation process was observed in S2, S6, S8 samples respectively, the S9 and finally S10 samples showed lower internal oxidation phenomena. The results suggest that the samples without bond coat or with porous coating structure showed high degree of internal oxidation degradation. The process is driven by the ability of Ti and Al from gamma prime Ni3(Al,Ti) phase to oxidation reaction due to high negativity of Gibbs free energy formation of TiO2 and Al2O3:

and - 892.2KJmol -1 that is required to form the oxide. In Haynes® 282® concentration of Al compared to Ti concentration is much lower resulting that the formation of TiO2 is much more prevalent than Al2O3; hence, the alloy suffers mainly due to the formation of internally TiO2 phase.

M

A

-1

ED

4. Conclusions

PT

The aim of this work was to study high temperature resistance and performance of the plasma spray coatings deposited on the Haynes® 282® alloy surface under steam oxidation conditions at 800oC for 2000 hours. Based on the results, the following conclusions can be made:

CC E

1) Four coatings: S3 (Aluminium oxide – 4 wt% reduced graphene oxide (RGO) – with no bond coat), S4 (Aluminium oxide (Al2O3) – with the bond coat), S5 (Aluminium oxide – 4 wt% reduced graphene oxide (RGO) – with the bond coat) and S7 (Aluminium oxide (Al2O3) – no bond coat), detached from the alloy surface indicating poor adhesion and stress development upon cooling to room temperature. 2) The uncoated Haynes® 282® sample (S10) showed similar corrosion resistance ability as the CeO2 no bond coat sample (S8), showing nearly parabolic mechanism of the oxide growth,

A

3) The impact of bond coat in S2 and S8 samples has been observed. 4) The highest mass gain was obtained for S9 (bond coat - no top coat (control)), the lowest mass gain has been found for S8 (Cerium oxide (CeO2) with no bond coat). 5) The formation of extensive internal oxidation regions in S1, S2 samples have been found. 6) The formation of Cr2O3, MnCr2O4 and TiO2 phases has been observed on the surfaces of the uncoated Haynes® 282® sample

12

Acknowledgment The project team would like to acknowledge that the undertaken project and its conclusions would not have been possible without the support, assistance and encouragement of the following individuals and organizations for which the author is deeply appreciative: University of Central Florida

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Energy Industries of Ohio a Non-Profit Corporation

13

References [1] J. Henry, G. Zhou, T. Ward, Lessons from the past: materials-related issues in an ultra-supercritical boiler at Eddystone plant, Mater High Temp. 24 (2007) 249-258. [2] N. Otsuka, Fracture behaviour of steam-grown oxide scales formed on 2–12%Cr steels, Mater High Temp. 22(1-2) (2005) 131 – 138.

IP T

[3] S.R. Paterson, R. Moser, T. Rettig, Electric Power Research Institute/VGB Technische Vereinigung der Grosskraftwerksbetreiber in Atomic Energy Commission USA -Reports-; 8-1; Interaction of ironbased materials with water and steam by EPRI; (1993) [4] T. Dudziak, V. Deodeshmukh, L. Backert, N. Sobczak, M. Witkowska, W. Ratuszek, K. Chrusciel, A. Zielinski, J. Sobczak, G. Bruzda, Phase investigations under steam oxidation process at 800 oC for 1000 h of advanced steels and Ni-based alloys, Oxid. Met. 87(1/2) (2016) 139–158

SC R

[5] T. Dudziak, L. Boron, V. Deodeshmukh, J. Sobczak, N. Sobczak, M. Witkowska, W. Ratuszek, K. Chruściel, Steam oxidation behaviour of advanced steels and Ni based alloys at 800 oC, Journal of Materials Engineering and Performance, 2017, DOI: 10.1007/s11665-017-2535-8

U

[6] J. Changheui, K. Daejong, K. Donghoon, S. Injin, R. Woo-Seog, Y. Young-Sung, Oxidation behaviors of wrought nickel-based superalloys in various high temperature environments, Trans. Nonferrous Met. Soc. China., 21(7) (2011) 1524-1531.

A

N

[7] D.L. Klarstrom, L.M. Pike, V.R. Ishwar, Nickel-Base Alloy Solutions for Ultrasupercritical Steam Power Plants, Procedia Engineering, 55 (2013) 221 - 225.

M

[8] F.A. Pérez-González, N.F. Garza-Montes-de Oca, R. Colás, High temperature oxidation of the Haynes 282© nickel-based superalloy, Oxid. Met. 82(3) (2014) 145-161. [9] D.A. Wrobleski, Corrosion resistant coating, US 3024176 A

ED

[10] M. Tiitu, Corrosion resistant coating, US 5658649 A [11] Newell C. Cook, Corrosion resistant coating, US 3024176 A

PT

[12] S. Bajpai, A. Gupta, S.K. Pradhan, T. Mandal, K. Balani, Crack propagation resistance of α-Al2O3 reinforced pulsed laser-deposited hydroxyapatite coating on 316 stainless steel, JOM 66(10) (2014) 2095 – 2107.

CC E

[13] S. Patil, S.C. Kuiry, S. Seal, R. Vanfleet, Synthesis of nanocrystalline ceria particles for high temperature oxidation resistant coating, J. Nanopart. Res. 4 (2002) 433–438.

A

[14] R. Thanneeru, S. Patil, S. Deshpande, S. Seal, Effect of trivalent rare earth dopants in nanocrystalline ceria coatings for high-temperature oxidation resistance, Acta Mater. 55 (2007) 3457– 3466. [15] S. Seal, S.K. Bose, S.K. Roy, Improvement in the oxidation behavior of austenitic stainless steels by superficially applied, Cerium oxide coatings, Oxid. Met. 41(I/2) (1994) 139 – 178. [16] S.K. Roy, S. Seal, S.K. Bose, M. Caillet, Effect of superficially applied cerium oxide coating on the non-isothermal oxidation of AIS1321 grade stainless steel, J. Mater. Sci. Lett. 12 (1993) 249-251.

14

[17] V. Viswanathan, R. Filmalter, S. Patil, S. Deshpande, S. Seal, High-temperature oxidation behavior of solution precursor plasma sprayed nanoceria coating on martensitic steels J. Am. Ceram. Soc. 90(3) (2007) 870–877. [18] M. Lukaszewicz, N.J. Simms T. Dudziak, J.R. Nicholls, Effect of Steam Flow Rate and Sample Orientation on Steam Oxidation of Ferritic and Austenitic Steels at 650 and 700° C, Oxid. Met. 79(5/6) (2013) 473 – 483.

IP T

[19] S. Ariharan, A. Gupta, A. Keshri; A. Agarwal; K. Balani, Size Effect of Yttria Stabilized Zirconia Addition on Fracture Toughness and Thermal Conductivity of Plasma Sprayed Aluminum Oxide Composite Coatings, Nanosci Nanotech Let 4(3), (2012) 323-332.

SC R

[20] D. Ward, A. Gupta, S. Saraf, C. Zhang, T.S. Sakthivel, S. Barkam, A. Agarwal, Functional NiAlgraphene oxide composite as a model coating for aerospace component repair, Carbon (105) (2016) 529-543.

[21] W. J. Quadakkers, D. Naumenko and E. Wessel, Growth Rates of Alumina Scales on Fe–Cr–Al Alloys, Oxid. Met. 61(1/2) (2004) 17-37. [22] Haynes® 282® Alloy – Haynes International Booklet

U

[23] S. Stecura, W.J. Campbell, Thermal expansion and phase inversion of rare-earth oxides, U. S. Department of the Interior, Bureau of Mines (1961)

A

N

[24] K. Wefers, Ch. Misra, Oxides and Hydrooxides of Aluminium, Alcoa Laboratories, Alcoa Technical Paper No 19, Aluminum Company of America (1987)

M

[25] J. Park, Bioceramics: Properties, Characterizations, and Applications Springer US (2008) [26] M. Di Ferdinando, A. Fossati, A. Lavacchi, U. Bardi, F. Borgioli, C. Borri, C. Giolli, A. Scrivani, Isothermal oxidation resistance comparison between air plasma sprayed, vacuum

ED

[27] V.K. Tolpygo, D.R. Clarke, K.S. Murphy, Oxidation-induced failure of EB-PVD thermal barrier coatings, Surf. Coat. Technol. 146–147 (2001) 124 131.

PT

[28] V.K. Tolpygo, D.R. Clarke, in: N.B. Dahotre, J.M. Hampikian, J.E. Morral (Eds.), Elevated Temperature Coatings: Science and Technology IV, TMS, Warrendale, PA, 93 (2001).

CC E

[29] M. Menna, Abo-Zeid, M. S. El-Deab, , A. AbdelKareem, Omayma A.M. El-Kady, A. M. Daher, Impurities Effect on the Charge Mobility of Yttria-Stabilized Zirconia, Int. J. Electrochem. Sci., 11 (2016) 3137 – 3146. [30] R.V. Wilhelm and D. S. Howarth, Iron oxide-doped ytria-stabilized zirconia ceramic — iron solubility and electrical conductivity, Ceram. Bull. 58 (1979) 228 – 232.

A

[31] H.J. Christ, L. Berchtold, H.G. Sockel, Oxidation of Ni-Base Alloys in Atmospheres with Widely Varying Oxygen Partial Pressures, Oxid. Met., 26(1/2) (1986) 45-76. [32] P.M. Scott, An Overview Of Internal Oxidation As A Possible Explanation Of Intergranular Stress Corrosion Cracking Of Alloy 600 In PWRS, 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems—Water Reactors, The Minerals, Metals & Materials Society (TMS), (1999)

15

[33] K.A. Christofidou, N.G. Jones, M.C. Hardy, H. J. Stone, The Oxidation Behaviour of Alloys Based on the Ni–Co–Al–Ti–Cr System, Oxid. Met. 85(3) (2016) 443-458. [34] V. Viswanathan, R. Filmalter, S. Patil, S. Deshpande, S. Seal, High-Temperature Oxidation Behaviour of Solution Precursor Plasma Sprayed Nanoceria Coating on Martensitic Steels, J. Am. Ceram. Soc., 90(3) (2007) 870–877. [35] A. Gupta, S. Das, C.J. Neal, S. Seal, Controlling the surface chemistry of cerium oxide nanoparticles for biological applications, J. Mater. Chem. B4 (2015) 3195 - 3202.

IP T

[36] S. Seal, S. K. Roy, S. K. Bose, S.C. Kuiry, Ceria based high temperature coatings for oxidation prevention JOM-e 52, 2000, 1-8 [37] S. K. Roy, S. Seal, S. K. Bose, M. Caillet, Effect of superficially applied cerium oxide coating on the non-isothermal oxidation of AIS1321 grade stainless steel, J. Mater. Sci. Lett. 12(4) (1993) 249-251.

SC R

[38] S. Seal, S. K. Bose and S. K. Roy, Improvement in the oxidation behaviour of austenitic stainless steels by superficially applied, cerium oxide coatings, Oxid. Met., 41(I/2) (1994) 139 - 178.

[39] C. Alcock, Thermochemical Processes Principles and Models, Butterworth-Heinemann, A Elsievier PLC group (2001)

N

U

[40] M.G. C. Cox, B. McEnaney, and V.D. Scott, Chemical diffusion model for partitioning of transition elements in oxide scales on alloys, Philosophical Magazine, 26(4) (1972) 839–851.

A

[41] F. Palmert, Oxidation and degradation of nickel-base alloys at high temperatures, Master of Science Thesis Stockholm, Sweden 2009

M

[42] G.R. Holcomb, D.E. Alman, The Effect of Manganese Additions on the Reactive Evaporation of Chromium in Ni-Cr Alloys, Department of Energy DOE/ARC-05-002

ED

[43] R.B. Dooley, Program on Technology Innovation: Oxide Growth and Exfoliation on Alloys Exposed to Steam, Final Report, Electric Power Research Institute (EPRI) 2007

PT

[44] D. J. Young, B. A. Pint, Chromium Volatilization Rates from Cr2O3 Scales into Flowing Gases Containing Water Vapour, Oxid. Met. 66(3), (2006) 137-153.

A

CC E

[45] E.J. Opila, Volatility of Common Protective oxides in High Temperature Water Vapour: Current Understanding and Unanswered Questions, Mater. Sci. Forum, 464 (2004) 765 - 774.Figure Caption

16

Figr-1List of Figures Fig. 1. Schematic diagram of steam oxidation rig used in the present study Fig. 2. Macro pictures of the as-deposited coatings on Haynes® 282® alloy substrates: A) S1, B) S2, C) S3, D) S4, E)S5, F) S6, G) S7, H) S8 and I) S9 Fig. 3. Microstructures of the as-deposited coatings on Haynes® 282® Ni-based alloy substrates: A) S1, B) S2, C) S3, D) S4, E)S5, F) S6, G) S7, H) S8 and I) S9

IP T

Fig. 4. Kinetic data for the uncoated and coated Haynes® 282® alloy exposed in steam atmosphere for 2000 hours at 800 oC. Fig. 5. A log-log plot of mass gain vs. time to determine the exponent n factor for the exposed samples. Fig. 6. The kp values for samples S8 and S10 exposed for 2000 hours at 800 oC.

SC R

Fig. 7. XRD spectrum of the Ni-based uncoated Haynes® 282® alloy exposed to steam oxidation atmosphere at 800 oC for 2000 hours.

Fig. 8. Macrostructures of the coated Haynes® 282® alloy exposed to steam oxidation atmosphere at 800 oC for 2000 hours: A) S1, B) S2, C) S6, D) S8, E) S9, and F) S10.

N

U

Fig. 9. SEM micrographs of the coated Haynes® 282® alloy exposed to steam oxidation atmosphere at 800 oC for 2000 hours in BSE mode: A) S1, B) S2, C) S6, D) S8, E) S9, and F) S10.

M

A

Fig. 10. Cross-section microstructures of the coated Haynes® 282® Ni-based alloy exposed in steam oxidation atmosphere at 800 oC for 2000 hours carried out by SEM in BSE mode: A) S1, B) S2, C) S6, D) S8, E) S9, and F) S10. Figures D1 – F1 are higher magnification of figures D-F respectively. Fig. 11. EDS elemental mappings of sample S2 (CeO2 with the bond coat) after 2000 hours of exposure in steam atmosphere at 800 oC

ED

Fig. 12. EDS elemental mappings of sample S8 (CeO2 with no bond coat) after 2000 hours of exposure in steam atmosphere at 800 oC

PT

Fig. 13. EDS elemental mappings of sample S10 after 2000 hours of exposure in steam atmosphere at 800 oC

A

CC E

Fig. 14. The minimum activities of Ti and Cr at 800 oC in steam atmosphere

17

flowmeter 11

1

5 3

9

1

8

IP T

4 13

SC R

7

1) Heat resistant furnace

2

3) Nitrogen inlet

12

A

4) Cold de ionized water inlet

N

U

2) Cylinder with nitrogen for water bubbling system

5) Al2O3 holder with the samples 10

M

6) Reaction chamber made of stainless steel 7) Peristaltic pump for water circulation within the system

PT

ED

8) Hot water cooler/condenser

A

CC E

Fig. 1. Schematic diagram of steam oxidation rig used in the present study

18

E

F

G

U

SC R

D

C

IP T

B

A

I

ED

M

A

N

H

PT

Fig. 2. Macro pictures of the as-deposited coatings on Haynes® 282® alloy

A

CC E

substrates: A) S1, B) S2, C) S3, D) S4, E) S5, F) S6, G) S7, H) S8 and I) S9

19

I 100 µm

F

PT

A

CC E

100 µm

D

100 µm

G

ED

E

C

M

100 µm

i

N U SC R

B

A

A

100 µm

100 µm

H

100 µm

100 µm

Fig. 3. Microstructures of the as-deposited coatings on Haynes® 282® Ni-based alloy substrates: A) S1, B) S2, C) S3, D) S4, E) S5, F) S6, G) S7, H) S8 and I) S9

100 µm

20

IP T SC R U N A M

Fig. 4. Kinetic data for the uncoated and coated Haynes® 282® alloy exposed in

A

CC E

PT

ED

steam atmosphere for 2000 hours at 800 oC

21

IP T SC R U N A M

A

CC E

PT

exposed samples

ED

Fig. 5. A log-log plot of mass gain vs. time to determine the exponent n factor for the

22

IP T SC R U N A M

A

CC E

PT

ED

Fig. 6. The kp values for samples S8 and S10 exposed for 2000 hours at 800 oC

23

IP T SC R U N A

M

Fig. 7. XRD spectrum of the Ni-based uncoated Haynes® 282® alloy exposed to

A

CC E

PT

ED

steam oxidation atmosphere at 800 oC for 2000 hours.

24

D

IP T

C

SC R

B

M

A

N

U

A

F

A

CC E

PT

ED

E

Fig. 8. Macrostructures of the coated Haynes® 282® alloy exposed to steam oxidation atmosphere at 800 oC for 2000 hours: A) S1, B) S2, C) S6, D) S8, E) S9, and F) S10.

25

A

B

YSZ

Cerium oxide

SiO2 particles 100 µm

C

IP T

100 µm

D

SC R

YSZ Cerium oxide

SiO2

U

particles

100 µm

N

100 µm

F

ED

Bond coat

M

A

E

MnCr2O4, Cr2O3, TiO2 oxide scale

CC E

PT

100 µm

100 µm

A

Fig. 9. SEM micrographs of the coated Haynes® 282® alloy exposed to steam oxidation atmosphere at 800 oC for 2000 hours in BSE mode: A) S1, B) S2, C) S6, D) S8, E) S9, and F) S10.

26

I N U SC R

A

C

A

B

500 µm

F

500 µm

500 µm

E1

D1

CC E

PT

ED

E

M

500 µm

D

500 µm

100 µm

500 µm

A

500 µm

27

I N U SC R

F1

Fig. 10. Cross-section microstructures of the coated Haynes® 282® Ni-based alloy exposed in steam oxidation atmosphere at 800 oC for 2000 hours carried out by SEM in BSE mode: A) S1, B) S2, C) S6, D)

A

S8, E) S9, and F) S10. Figures D1 – F1 are higher magnification of figures D-F respectively

A

CC E

PT

ED

M

500 µm

28

IP T SC R U N A M ED PT CC E A

Fig. 11. EDS elemental mappings of sample S2 (CeO2 with the bond coat) after 2000 hours of exposure in steam atmosphere at 800 oC

29

BSE

O

Cr

SC R

Ce

50 m

IP T

50 m

50 m

N

U

50 m

Al

ED

M

A

Ni

PT

50 m

Mn

A

CC E

Ti

50 m

50 m

50 m

Fig. 12. EDS elemental mappings of sample S8 (CeO2 with no bond coat) after 2000 hours of exposure in steam atmosphere at 800 oC

30

BSE

O

20 m

20 m

IP T

Ni

SC R

Cr

20 m

Ti

ED

M

A

N

Al

U

20 m

A

CC E

Mn

20 m

PT

20 m

20 m

Fig. 13. EDS elemental mappings of sample S10 after 2000 hours of exposure in steam atmosphere at 800 oC

31

-40

-35

-30

-25

-20

-15

-10

-5

0

30

30

20

20

10

10

0

0

IP T

Log a

-45

-10

-10

-20

-20

SC R

-30 -40 -40

-35

-30

-25 -20 Log pO2

aTi

-10

-5

0

N

aCr

-15

-40

U

-45

-30

A

CC E

PT

ED

M

A

Fig. 14. The minimum activities of Ti and Cr at 800 oC in steam atmosphere

32

List of Tables Table 1 Chemical composition of Ni-based alloy (wt.%) used for high-temperature steam oxidation study Table 2 Unique nomenclature of the tested materials under steam conditions Table 3 Plasma spray deposition parameters for the coatings deposited on Haynes® 282® alloy used in the study

IP T

Table 4 Depth penetration for the selected samples and the selected phases using two different geometries Table 5 The kp values for S8 and S10 samples exposed to the steam environment for 2000 hours at 800 oC

SC R

Table 6 Linear coefficient of thermal expansion (CTE) of the ceramics used in this work

A

CC E

PT

ED

M

A

N

U

Table 7 Gibbs free energy values of the oxides of interest evaluated at 800 oC for the exposed uncoated Haynes® 282® alloy in a steam atmosphere

33

Table 1 Chemical composition of Ni-based alloy (wt.%) used for high-temperature steam oxidation study

Haynes®

Fe

Cr

Co

Mo

Si

Mn

Cu

La

Ti

Al

C

W

B

Bal.

1.5

20

10

8.5

0.15

0.3

-

-

2.1

1.5

0.06

-

0.05

A

CC E

PT

ED

M

A

N

U

SC R

IP T

282®

Ni

34

Table 2 Unique nomenclature of the tested materials under steam conditions Bond coat

Top coat

S1

Yes

Yttria stabilised zirconia (YSZ)

S2

Yes

Cerium oxide (CeO2)

S3

No

Aluminium oxide – 4 wt% reduced graphene oxide (RGO)

S4

Yes

Aluminium oxide (Al2O3)

S5

Yes

Aluminium oxide – 4 wt% reduced graphene oxide (RGO)

S6

No

Yttria stabilised zirconia (YSZ)

S7

No

Aluminium oxide (Al2O3)

S8

No

Cerium oxide (CeO2)

S9

Yes

No top coat (control)

S10

No

Reference sample (Haynes® 282® alloy)

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Sample code

35

Table 3 Plasma spray deposition parameters for the coatings deposited on Haynes® 282® alloy used in the study Ni185

ZRO-182

ALO-101

Al2O3-GO

CeO2

Current (A)

600

600

850

850

600

Power (kW)

19.5

18.8

19.4

19.4

19.5

Primary gas (SCFH)

85 (Ar)

68 (Ar)

82 (Ar)

82 (Ar)

85 (Ar)

Secondary gas (SCFH)

10 (H2)

104 (He)

40 (He)

40 (He)

Carrier gas (SCFH)

8.5 (Ar)

49 (Ar)

13 (Ar)

13 (Ar)

Powder feed rate (rpm) 2

3

2.5

Standoff distance (mm) 115

100

100

90

90

SC R

10 (H2)

8.5 (Ar)

2.5

2

100

115

90

90

A

CC E

PT

ED

M

A

N

U

90

Spray angle ()

IP T

Parameter

36

Table 4 Depth penetration for the selected samples and the selected phases using two different geometries Sample

BB

Grazing

geometry

incidence =1o 0.47 m

4.64 - 11.76 m

incidence

=3o 1.35 m

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Haynes® 282®

Grazing

37

Table 5 The kp values for S8 and S10 samples exposed to the steam environment for 2000 hours at 800 oC Sample ID Kp [(mg/cm2)2 s-1] 1.81E-07

Haynes® 282® alloy (S10)

2.03E-07

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Cerium oxide no bond coat (CeO2) S8

38

Table 6 Linear coefficient of thermal expansion (CTE) of the ceramics used in this work Mean coefficients of thermal expansion (CTE) [10-6*K-

Material

1

]

Ref.

13.95

[20]

CeO2

12.0

[21]

Al2O3

8.5

[22]

ZrO2

7 (monoclinic); 12 (tetragonal)

IP T

Haynes® 282®

A

CC E

PT

ED

M

A

N

U

SC R

[23]

39

Table 7 Gibbs free energy values of the oxides of interest evaluated at 800 oC for the exposed uncoated Haynes® 282® alloy in a steam atmosphere

TiO2

Cr2O3

MnO

MnCr2O4

G To [KJ mol -1]

- 749.6

- 566.9

- 618.0

- 40.0

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Phase

40