Materials and Design 166 (2019) 107599

Contents lists available at ScienceDirect

Materials and Design

j ourna l homepage: www.e lsev ie r .com/ locate /matdes

The iron effect on oxidation and interdiffusion behaviour in MCrAlXcoated Ni-base superalloys

Pimin Zhang a,⁎, Xin-Hai Li b, Johan Moverare a, Ru Lin Peng aa Division of Engineering Materials, Department of Management and Engineering, Linköping University, SE-58183 Linköping, Swedenb Siemens Industrial Turbomachinery AB, SE-61283 Finspång, Sweden

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Fe addition increases β phase fractionand stabilizes β phase in the coating.

• Fe addition promotes a σ phase layerformation and retains Al at 900 °C.

• High Fe addition reduces grain coarsen-ing in the BLZ.

• Less β depletion of the coating due to ahigh Fe addition.

⁎ Corresponding author.E-mail address: pimin.zhang@liu.se (P. Zhang).

https://doi.org/10.1016/j.matdes.2019.1075990264-1275/© 2019 Elsevier Ltd. This is an open access art

a b s t r a c t

a r t i c l e i n f o

Article history:Received 18 September 2018Received in revised form 7 January 2019Accepted 8 January 2019Available online 11 January 2019

Two MCrAlX powders with different Fe content were deposited on an IN792 superalloy using high velocityoxygen-fuel spraying (HVOF). The oxidation and interdiffusion behaviour of the coated specimens were investi-gated at 900, 1000 and 1100 °C for different exposure times. Experimental results showed that high Fe addition inthe MCrAlX coating had no obvious effect on oxidation, however, a great impact on coating-substrate interdiffu-sion and coating degradation was observed. Although no influence on Al diffusivity in IN792, high Fe additionpromotes the formation of a discontinuous σ phase layer at the coating/substrate interface, which retards Aldiffusion in the IN792 substrate at 900 °C. To investigate Fe influence on phase equilibrium, thermodynamiccalculation was also performed. It was shown that high Fe addition in the MCrAlX coating increased theβ phase fraction and stabilized the β phase by extending the β + γ phase region. Besides, high Fe addition inthe MCrAlX coating modifies Cr diffusion, and alters local phase equilibrium at the coating/substrate interface.These two factors suppress inner-β depletion of high Fe containing coating. Grain coarsening of the coatingcan be observed during oxidation, it was retarded by high Fe addition.

icle under the CC BY-NC-ND lice

© 2019 Elsevier Ltd. This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords:CoatingIron additionInterdiffusionThermodynamicsMicrostructure

1. Introduction

Due to the ability to form dense, adherent Al2O3 scales, MCrAlX (M:Ni and/or Co, X: minor elements) coatings have been widely used to

nse (

protect the hot section components in gas turbines, either as overlaysor as bond coats for thermal barrier coatings (TBCs) [1–7]. During oxida-tion at elevated temperature, the life span of MCrAlX coatings is mainlylimited by the aluminium depletion as a result of growth of aluminascale at the coating surface and the interdiffusion between substrateand theMCrAlX coating. The β-NiAl phase inMCrAlX coatings is consid-ered as Al-reservoir, therefore its stability plays an important role inmaintaining good coating performance.

http://creativecommons.org/licenses/by-nc-nd/4.0/).

http://crossmark.crossref.org/dialog/?doi=10.1016/j.matdes.2019.107599&domain=pdfhttp://creativecommons.org/licenses/by-nc-nd/4.0/https://doi.org/10.1016/j.matdes.2019.107599pimin.zhang@liu.sehttps://doi.org/10.1016/j.matdes.2019.107599http://creativecommons.org/licenses/by-nc-nd/4.0/http://www.sciencedirect.com/science/journal/02641275www.elsevier.com/locate/matdes

2 P. Zhang et al. / Materials and Design 166 (2019) 107599

In general, the composition of MCrAlX coatings was designed toprovide better protection against oxidation and hot corrosion duringservice. Early research focused on the optimization of main elementcontent in MCrAlX coatings, i.e. the Ni/Co ratio and the Cr and Alcontents [8,9]. It has been reported by many studies that the addition

coa�nsubstrate

IBDZ BLZ

GPDZ

(c)

C1C1

10 μm

(a)

(b)

50 μm

10 μm

Fig. 1.Cross-section of coating/substrate systemafter coating deposition (a) and insets showingcoating/substrate system after oxidation (c), showing different reaction zones.

of minor reactive element (RE), such as yttrium [10–12], zirconium[13,14], cerium [15,16], hafnium [17–19] and Si [20,21], improves thehigh temperature oxidation resistance of MCrAlX coatings through REeffect [22,23]. The addition of other elements such as Re [24–26], Ta[27–31], Pt [32,33], Ru [34] also improves coating performance against

g

OBDZ

100 μm

C2

10 μm

50 μm

10 μm

cross section of powders; highmagnification of the coating/substrate interface (b); and the

Table 2Chemical composition of as-heat treated MCrAlX coatings [in wt%].

Coating Ni Co Cr Al Y Fe Others

C1bal. 26 15 12 0.3

1.6b0.6

C2 9.6

Table 1Exposure time for oxidation tests at different temperature.

Temperature (°C) Exposure time (h)

900 500, 1500, 3000, 50001000 500, 1500, 5000, 80001100 50, 300, 500, 800

3P. Zhang et al. / Materials and Design 166 (2019) 107599

oxidation. Fe has been treated as a major element in FeCrAlY coatingsfor application on Fe-base superalloys or steel [35–39], and the additionof Fe in Ni-Cr-Al system capable of stabilizing the β phase has been re-ported by Ma et al. [40]. Since the long-term performance of MCrAlXcoatings is directly linked to the β phase stability, Fe as aminor additionin MCrAlX coating for Ni-base superalloy could be promising. Besides,thewide resource of Feworldwidemakes it very cost efficient and envi-ronmental friendly as alloying element in MCrAlX coatings comparedwith others. However, few studies have been focused on the effect ofFe addition as a minor element (b10 wt%) in MCrAlX during high tem-perature oxidation. Hence, a detailed analysis on the Fe-containingMCrAlX coatings with respect to oxidation and interdiffusion betweencoating and substrate is required. In this paper, results obtained fromoxidation test at three different temperature of two MCrAlX coatingswith different Fe content (1.6 and 9.6 wt%) are presented.

2. Experimental

Two tailor-made MCrAlX powders were used: C1 coating with1.6 wt% Fe content and C2 coating with 9.6 wt% Fe content. The powderwas produced in a vacuum atomization reactor in an industrialcompany (powder manufacturer). The powders with size of ~37 μmwere deposited on Φ25.4 × 5 mm IN792 disks (composition in wt%:

0 2000 4000 6000Oxidation Time (h)

0

5

10

15

Alu

min

a sc

ale

thic

knes

s (m

)

C1C2

0 2000Oxidation

0

5

10

15

Alu

min

a sc

ale

thic

knes

s (m

)

C1C2

0 2000 4000 6000Oxidation Time (h)

0

10

20

30

40

OB

DZ

thic

knes

s (m

) C1C2

0 2000 4000Oxidation

0

10

20

30

40

OB

DZ

thic

knes

s (m

) C1C2

(a)

(d) (e)

(b)

900 oC

900 oC

Fig. 2. Alumina scale thickness as a function of oxidation time at 900 °C (a), 1000 °C (b), and 11

12.5 Cr, 9 Co, 4.2 W, 4.2 Ta, 4 Ti, 3.4 Al, 1.9 Mo 0.08C, 0.03 Zr + B, bal-anced by Ni). A JP5000 high velocity oxygen-fuel (HVOF) spray systemwas used to deposit coatings of 200 μm thick with surface roughnessin the range between 6 and 6.5 μm, see Fig. 1. The spray conditionsused for HVOF deposition were: oxygen flow rate 900 l/min with pres-sure of 1.5 MPa, fuel (Kerosene) flow rate 32 l/h, carrier gas (N2) flowrate 8 l/min with pressure of 0.3 MPa, spray distance 380 mm, powderfeed rate 75 g/min, as also described in Ref [41]. Prior to the coating de-position, the substrate surfaces were grit blasted for a better coating/substrate bonding. After the deposition, the as-sprayed samples weresubjected to a heat treatment, including solution annealing at 1120 °Cfor 2 h in vacuum, aging treatment at 845 °C for 24 h in air, and aircooling to room temperature. Such a heat treatment is intended tostrengthen the IN792 substrate [42] and also increases the bondstrength between the coating and substrate [43].

The samples were subjected to oxidation tests in lab air at 900 °C,1000 °C and 1100 °C for different exposure times, as listed in Table 1.After oxidation, the oxidized samples were mounted in epoxy, sec-tioned and polished. A Hitachi SU70 FEG scanning electron microscopy(SEM), equipped with energy dispersive spectroscopy (EDS), wave-length dispersive spectroscopy (WDS) and electron back scatter diffrac-tion (EBSD) system from Oxford instruments, was used to examine thecross-sections of samples. The nominal composition of the two heat-treated coatings, measured on a 20 × 25 μm cross-section area by EDSand WDS, are shown in Table 2.

Accurate element concentration profiles were obtained by EDS areascans. Each data point of the profiles show the chemical composition ofa 100 ∗ 10 μm area parallel to the coating/substrate interface. By usingarea scanning, the composition of each areaswas derived as the averagecomposition of all phases within the area. Besides, EDS point analysiswas employed to measure the composition of different phases nearthe interdiffusion zone with the spatial resolution of 3 μm for Al butbetter than 3 μm for other elements with a higher atomic number. Tobetter describe the microstructure development, the coating and sub-strate were divided into several zones: outer-β-phase depletion zone(OBDZ), β-phase left zone (BLZ), inner-β-phase depletion zone (IBDZ)and γ′-phase depletion zone (GPDZ), as illustrated in Fig. 1c. The thick-ness evolution of different zoneswas quantified by using image analysissoftware. Alumina scale thicknesswasmeasured on the cross-section of

4000 6000 Time (h)

0 200 400 600 800 1000Oxidation Time (h)

0

5

10

15

Alu

min

a sc

ale

thic

knes

s (m

)

C1C2

6000 8000 10000 Time (h)

0 200 400 600 800 1000Oxidation Time (h)

0

10

20

30

40

OB

DZ

thic

knes

s (m

) C1C2

(f)

(c)

1000 oC

1000 oC

1100 oC

1100 oC

00 °C (c) oxidation; and the corresponding OBDZ thickness in (d), (e) and (f), respectively.

5 μm

coa�ng

Al2O3

Spinel: (Ni,Co,Cr,Fe)xOy

epoxy

O Al Y

Fe Cr

Co Ni

Fig. 3. Cross section microstructure of thermally grown oxides of C2 coating after 8000 h at 1000 °C oxidation, showing (Ni, Co, Cr, Fe)xOy type spinel.

(a)

(b)

(c)

(d)

β

β

β

β

γ’

γ’

γ’

γ’ c

c

c

c

C2, iso900, 500h

C2, iso900, 5000h

C1, iso900, 5000h

C1, iso900, 500h

30 μm

30 μm

30 μm

30 μm

σ

σ

σ

σ

Fig. 4.Microstructure near coating (upper part)/substrate (lower part) interface (marked by the dash curve). Samples oxidized at 900 °C condition are (a) C1 for 500 h, (b) C1 for 5000 h,(c) C2 for 500 h and (d) C2 for 5000 h. “c” denotes Ti and Ta rich carbides and “σ” denotes topologically-close packed (TCP) sigma phase.

4 P. Zhang et al. / Materials and Design 166 (2019) 107599

5P. Zhang et al. / Materials and Design 166 (2019) 107599

the specimens. Grain size in BLZ of coatings was measured by EBSDmapping with a step size of 0.5 μm using the HKL Channel 5 software.Each mapping area contains at least 2000 grains.

3. Thermodynamic modelling

To fully understand the observed microstructure evolution of thetwo MCrAlX systems with variable Fe content, the experiment workwas supplemented with computational simulation, using Thermo-Calc

(a)

(b)

C1, iso1000, 8000h

β

γ’

γ’

γ’

c

c

βC1, iso1000, 500h

(c)

c

C2, iso1000, 500h

(d)

C2, iso1000, 8000h

γ’

c

Fig. 5.Microstructure near coating (upper part)/substrate (lower part) interface (marked by the(c) C2 for 500 h and (d) C2 for 8000 h. “c” denotes Ti and Ta rich carbides.

software package with the advanced Ni-base thermodynamic databaseTCNI8 [44]. Thermodynamic modelling is utilized to calculate the phaseequilibria and Gibbs free energy of the alloy system.

To investigate the Fe effect on grain coarsening of MCrAlX systems,the grain coarsening rate coefficient of the system was derived by TC-PRISMA module in Thermo-Calc Software. In general, the precipitationmodule is proposed by Chen et al. [45] and calculates the velocity of amoving phase interface in multicomponent systems by identifying theoperating tie-line from the solution of flux-balance equations. The

30 μm

30 μm

30 μm

β

γ’

30 μm

β

dash curve). Samples oxidized at 1000 °C condition are (a) C1 for 500 h, (b) C1 for 8000 h,

Table 3σ phase composition (in wt%) for C1 and C2 coatings at 900 °C.

Coat. Time (h) Al Ti Cr Fe Co Ni Mo Ta W

C1 500 0.9 0.3 41 1 25.6 13.5 6.3 4.5 7.1C2 500 1.4 0.3 35.9 7.4 21.5 16.5 4.9 4.6 7.6

5000 0.4 0.1 42.5 7.2 22.1 10.3 5.9 4.5 7.0

6 P. Zhang et al. / Materials and Design 166 (2019) 107599

model is based on several assumptions: 1) the crystal lattice of theparent and products are assumed to be coherent; 2) the interfacialenergy is solely due to the abrupt compositional change at the interface.Therefore, the calculated coarsening rate coefficient between β/γ can beutilized to evaluate the Fe effect on grain coarsening in MCrAlX coating.

4. Results

Fig. 1a shows the cross-section of the coating/substrate systemafter coating deposition and dendrite-like morphology of the pow-der particles. It's clear that a splat-on-splat structure can be observedwith few pores marking the splat boundaries, and β + γ phasemicrostructure is formed in the coatings (β in dark grey) after thecoating deposition. A 15 μm recrystallization zone can be observedat the coating/substrate interface of both coatings with blue dotlines showing γ grain boundaries (Fig. 1b), due to the plastic deforma-tion induced by grit blasting prior to the coating deposition [46].After the heat treatment, the coating porosity reduced from 1.6 ±0.4% to 0.4 ± 0.2% (C1) and from 1.1 ± 0.2% to 0.3 ± 0.1% (C2). Exceptfor the reduction in the coating porosity, no significant change in thecoating microstructure can be observed after the heat treatment and,therefore, not presented here.

0 2000 4000 6000 8000 10000Oxidation Time (h)

0

20

40

60

IBD

Z th

ickn

ess (

m) C1

C2

(a)

0 2000 4000 6000 8000 10000Oxidation Time (h)

0

20

40

60

80

100

120

GPD

Z th

ickn

ess (

m) C1

C2

(c)

Fig. 6. Inner-β-phase depletion zone (IBDZ) thickness as a function of oxidation time at 100thickness in (c) and (d), respectively.

4.1. Oxidation

Fig. 2 shows the alumina scale thickness measured on the cross-section of coatings after oxidation at 900, 1000, and 1100 °C for differentexposure times, and the corresponding OBDZ thickness. Note that alu-mina scale thickness for oxidation at 1000 °C for 8000 h is missing dueto the spallation during sample preparation.With the increase of oxida-tion temperature and time, the growth of the alumina scale depletes Alfrom the coating, leading to the expansion of the OBDZ. As can be seen,the C1 coating shows similar alumina scale growth behaviour and theOBDZ thickness as the C2 coating at the various conditions. Fig. 3 showsthe cross-sectional microstructure of thermally grown oxide (TGO)of the C2 coating at intact region after 8000 h oxidation at 1000 °C,similar TGO structure can also be observed on C1 coating. Accordingto the EDS mapping, Fe-oxides coexist in the outer layer of TGO as a(Ni, Co, Cr, Fe)xOy type spinel.

4.2. Microstructure evolution near the interdiffusion zone

4.2.1. Microstructure at the coating/substrate interfaceOxidation at high temperature causes interdiffusion between

coating and substrate, and the microstructure evolution at the coating/substrate interface highly depends on both coating composition andoxidation temperature, as presented in Figs. 4 and 5.

Fig. 4 shows the microstructure near the coating/substrate interfaceof the C1 and C2 coatings after 900 °C oxidation for 500 and 5000 h. Thebright white phase was identified as (Ta, Ti)-rich carbides, while thelight grey phase was identified as TCP σ-(Cr, Co) phase by EDS asshown in Table 3. From Fig. 4a for C1, it is clear that microstructure inthe coating side mainly consisted of β and γ phases coexisted withfew small σ grains. In the GPDZ of substrate, a γ/γ′ “palm-tree”-like

0 200 400 600 800 1000Oxidation Time (h)

0

20

40

60

IBD

Z th

ickn

ess (

m) C1

C2

(b)

0 200 400 600 800 1000Oxidation Time (h)

0

20

40

60

80

100

120

GPD

Z th

ickn

ess (

m)

C1C2

(d)

0 °C (a), 1100 °C (b) oxidation, and the corresponding γ′-phase depletion zone (GPDZ)

7P. Zhang et al. / Materials and Design 166 (2019) 107599

microstructure appeared beneath interface due to the cellular recrystal-lization favoured by residual stresses induced by the grit blasting [47].σphase and carbides with different sizes were also observed at 500 h.With prolonged oxidation to 5000 h, the extensive interdiffusion in-duced significantmicrostructural evolution near the interface, includingcoarsening of Kirkendall voids, expansion of theγ′ depletion zone in thesubstrate and dissolution of σ grains in the coating side, but has aminoreffect on the β phase depletion in the coating and γ′ depletion in thesubstrate. However, the microstructural evolution in C2 differed fromC1. In the coating side of C2, a certain amount of small carbides precip-itated and a large amount of σ grains, which coarsened after 5000 h,coexisted with the β and γ phases. The average size of these σ grainsis 5.2 ± 1.5 μm, which is much smaller than the 10–20 μm in Ref [48].The GPDZ of C2 consists of secondary β phase (SB) precipitation, σphase, carbides and γ matrix, which is quite complicated compared tothe “clean” GPDZ of the C1 substrate. As oxidation proceeded to5000 h, the expansion of the GPDZ in the C2 substrate is clear, whilethe coarsening of Kirkendall voids is insignificant.

Fig. 5 shows the microstructure near the coating/substrate interfaceof C1 and C2 coatings at 1000 °C oxidation for 500 and 8000 h. Theinterdiffusion was accelerated due to the increasing element diffusivityat higher temperature, leading to an inner-β-phase depletion in thecoatings. The σ phase, which was no longer stable at 1000 °C, was notobserved in coating or substrate. In the GPDZ of C1 substrate (Fig. 5a),a small amount of secondary β phase coexisted with carbides in the(Cr, Co)-rich γ matrix, which dissolved before 8000 h. The β phasein the C1 coating near the interface depleted fast and was partiallytransformed to γ′ phase. The C2 coating shows less inner-β-depletion

-350 -300 -250 -200 -150 -100 -50 0Distance (Distance ( m)m)

4

5

6

7

8

Al (

wt %

)

50 h300 h500 h800 h

-350 -300 -250 -200 -150 -100 -50 0Distance (Distance ( m)m)

4

5

6

7

8

Al (

wt %

)

500 h3000 h5000 h8000 h

C1-1000 oC

substratecoat

(a) (

(

3.4

3.4

C1-1100 oC

(c)

Fig. 7.Al concentration profile from substrate to the coating/substrate interface of C1 (a) and C2that the baseline of Al concentration is 3.4 wt% (Al content in IN792). An example of Al diffusio

compared with C1, in which the β phase transformed directly toγ phase (Fig. 5b and d). Furthermore, comparing Fig. 5c and Fig. 5d re-veals that the large amount of secondary β phase precipitated at 500 hin the GPDZ of the C2 substrate dissolved before 8000h. Themicrostruc-ture evolution of coatings at 1100 °C are similar to 1000 °C, except thatβphase transformed to γ phase directly in both C1 and C2 coatings.

4.2.2. IBDZ&GPDZWhile no inner-β-phase depletion can be observed at 900 °C up to

5000 h oxidation, oxidation at 1000 and 1100 °C resulted in Fe-content dependent inner-β-phase depletion as Fig. 6 reveals. Forsamples oxidized at 1000 °C, the inner-β-phase depletion shows an“incubation” stage as the depletion initiated at 1500 h and 5000 h forC1 and C2 coatings, respectively. At 1100 °C, inner-β-phase depletionbecomes significant at 50 h in both coatings, indicating little or no incu-bation time. The C1 coating suffers a more rapid inner-β-phase deple-tion, indicating a better resistance against inner-β-phase depletion ofthe C2 coating with high Fe content. Similarly, GPDZ evolution duringoxidation at 1000 and 1100 °C are presented in Fig. 6c and d, respec-tively, which is consistent with themicrostructure observation in Fig. 5.

4.2.3. Al concentration profiles in the substrateFig. 7 shows EDS results of Al concentration profile of C1 and C2

coatings from substrate to the coating/substrate interface during 1000and 1100 °C oxidation at various times (900 °C data shared a similartrend as 1000 °C but will not be presented here). The presence ofsecondary β phase (SB), as also identified in Fig. 5, increases Al concen-tration at the coating/substrate interface of C2 coating. After longer

-350 -300 -250 -200 -150 -100 -50 0Distance (Distance ( m)m)

4

5

6

7

8

Al (

wt %

)

500 h3000 h5000 h8000 h

-350 -300 -250 -200 -150 -100 -50 0Distance (Distance ( m)m)

4

5

6

7

8

Al (

wt %

)

50 h300 h500 h800 h

C2-1000 oC

ing

GPDZ

SB

b)

d)

3.4

3.4

C2-1100 oC

dAl

(b) coatings at 1000 °C oxidation; of C1 (c) and C2 (d) coatings at 1100 °C oxidation. Noten depth, dAl, was marked in (d).

7 7.5 8 8.5 910-17

10-16

10-15

10-14

10-13

C1fit-C1C2fit-C2

0 1000 2000 3000 4000 5000Oxidation Time (h)

0

1000

2000

3000

4000

5000

Squa

re o

f Al d

iffus

ion

dept

h (

m2 )

C1fit-C1C2fit-C2

0 2000 4000 6000 8000Oxidation Time (h)

0

1

2

3

4

5

6

Squa

re o

f Al d

iffus

ion

dept

h (

m2 ) 104

C1fit-C1C2fit-C2

0 200 400 600 800Oxidation Time (h)

0

1

2

3

4

5

6

Squa

re o

f Al d

iffus

ion

dept

h (

m2 ) 104

C1fit-C1C2fit-C2

(a)

(c)

(b)

(d)

900 oC

1100 oC

1000 oC

Fig. 8. Square of Al diffusion depth in IN792 substrate of C1 and C2 coating system as a function of oxidation time at 900 °C (a), 1000 °C (b) and 1100 °C (c); (d) Al diffusivity in IN792 as afunction of inverse temperature.

8 P. Zhang et al. / Materials and Design 166 (2019) 107599

oxidation time, more Cr and Co diffused from coating to substrate(see Section 6.2.2), leading to the transformation ofγ+γ′→γ and for-mation of the GPDZ. Due to the low Al solubility of γ phase, Al wasrejected from GPDZ, forming an Al peak in the substrate next to GPDZ,whose position moves with the expansion of the GPDZ. The Al concen-tration profile after passing the peak follows an error function-typedistribution as a typical diffusion-controlled process. Based on the Al con-centration profile, Al diffusion depth, dAl, which is defined as the distancefrom the left boundary of the GPDZ (the location can be derived fromFig. 6) to the point where Al concentration drops to ~3.4 wt% (Al wt% inIN792) can be measured, as illustrated in Fig. 7d. The γ phase region(GPDZ) is excluded to reveal the true Al diffusivity in IN792. The diffusiv-ity of Al in the IN792 substrate, D, can be estimated by.

D ≈ d2Al=t; ð1Þ

with D following the Arrhenius relationship:

D ¼ D0 exp − QRT� �

ð2Þ

where t is the oxidation time, Q is the activation energy, T is temperatureand R is the gas constant. The obtained square of Al diffusion depth as afunction of oxidation time at various temperatures is plotted in Fig. 8a–c, and the experimental data were fitted to Eq. 2. The true initial point(dAl = 0) cannot be obtained from current experimental observationand is therefore ignored in Fig. 8a–c. It is because the phase transforma-tion in GPDZ greatly affected elemental interdiffusion behaviour at the

beginning of oxidation. The true Al diffusion depth defined here initiatedafter the stabilization of microstructure in GPDZ. It can be seen thatthe Al diffusion depth is similar for C1 and C2 oxidized at 1000(Fig. 8b) and 1100 °C (Fig. 8c), but different at 900 °C, for whichthe Al diffusion depth of C2 is much lower than that of C1 (Fig. 8a).The diffusivity of Al in IN792 substrate at different temperatures, D,can be derived by calculating the slop of corresponding curves inFig. 8a–c. The calculated diffusivity of Al in IN792 substrate at differ-ent temperatures is plotted in Fig. 8d as a function of inverse temper-ature in logarithmic scale, based on Eq. 2. It agrees with the volumediffusion coefficient for aluminium in a nickel base matrix (DAl =2.0 × 10−14 m2/s at 1135 °C) [49]. The activation energy determinedby linear regression analysis of the dash line, is 330 and 340 kJ/molfor the C1 and C2 coating systems, respectively.

4.3. Grain size/microstructure evolution in coating

Examples of EBSD phase distribution maps (Fig. 9) reveal a poly-crystalline and double phase microstructure in the BLZ of thecoatings after exposure at 1000 °C for 500 h and 8000 h, respectively.It is clear that most grains are below 2 μm (Fig. 9a and b) after 500 h,and their distributions are similar in both coatings. After 8000 h ox-idation, the grain size distribution of C1wasmuch scattered between1 and 5 μm with few large γ grains above 6 μm. However, the grainsize distribution of C2 after 8000 h (Fig. 9d) was mostly congregatedbelow 3 μm. Thus, the coarsening of grains is more severe in C1 after8000 h at 1000 °C. Annealing twins of γ grains can be observed inboth coatings.

Grain size ( m)0 2 4 6 8 10 12

rela

tivef

requ

ency

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

betagamma

(a)C1-500h

Grain size ( m)0 2 4 6 8 10 12 14

rela

tivef

requ

ency

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

(b)C2-500h

20 m20 m

gb gb

betagamma

Grain size ( m)0 2 4 6 8 10 12 14

rela

tivef

requ

ency

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

C1-8000h

Grain size ( m)0 2 4 6 8 10 12 14

rela

tivef

requ

ency

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

C2-8000h(c) (d)

20 m

gb gb20 m

betagamma beta

gamma

14

Fig. 9. Histograms for the grain size of C1 (a) and C2 (b) exposed at 1000 °C condition for 500 h, (c) and (d) for 8000 h and the corresponding phase distribution maps at the upper leftcorner. The red and blue areas in the phase distribution map correspond with γ and β phases, respectively. The black and white lines are γ/γ grain boundaries and β/β grain boundarieswith a misorientation larger than 15°, respectively.

9P. Zhang et al. / Materials and Design 166 (2019) 107599

To quantify the coarsening tendency of grains in the BLZ, the averagegrain size as a function of oxidation time for β and γ grains is displayedin Fig. 10 for different oxidation temperatures. The grain size is narrowlydistributed for short oxidation times and becomes more scatteredafter long time exposure, as indicated by the expansion of the errorbar (standard deviation of the grain size). The average size of the grainsincreases with oxidation time at 1000 and 1100 °C, but is constant at900 °C. Although coarsening of grains starts already at the beginningof oxidation, the coarsening of both β and γ grains of C1 coating at1000 °C becomes more significant after 5000 h. However, grains of C2coating are only slightly coarsened after 500 h and remain relativelystable afterwards. At 1100 °C, the coarsening of grains especially the γgrains of C1 coating is considerable.

Fig. 11 shows the β phase fraction in BLZ at different temperaturesas a function of exposure time (Fig. 11a–c), based on the EBSDphase distribution maps, and corresponding BLZ thickness fraction(Fig. 11d–f), defined as the BLZ thickness normalized by the coatingthickness. It can be seen that the β phase fraction of the coatingsdecreases with the increase in oxidation time and temperature, andthe β phase fraction of the C2 coating is generally higher than that ofthe C1 coating. At relatively low oxidation temperature (900 °C), thereduction of the β phase fraction is negligible up to 5000 h oxidationas a result of a low thermal activation energy for either oxidation orcoating-substrate interdiffusion processes, which is consistent withFigs. 2 and 6. For higher oxidation temperatures, the descending trendof the β phase fraction is clear with a higher depletion rate of the C1coating, especially at 1100 °C.

5. Thermodynamic modelling results

As aforementioned, the lower IBDZ thickness of C2 coating (Fig. 6)indicates less β phase depletion in the C2 coating. A possible explana-tion is that the β phase stability is related to the Fe content in thecoating. In this section, thermodynamicmodelling results are presentedto support the theory.

5.1. Fe effect on phase equilibrium

Due to the small amount of non-Fe minor elements additions(the sum of minor elements is b1.5 wt%), the C1 and C2 coatings canbe treated as a Ni-Co-Cr-Al-xFe system. Fig. 12 shows isothermal sectionfor Ni19Co13.5Cr-xAl-yFe at 900, 1000 and 1100 °C. The contents of Coand Cr are set based on the EDS measurement in the BLZ of samplesoxidized for 8000 h at 1000 °C (Table 4). The inward diffusion of Fefrom coating to substrate leads to depletion of Fe and the Fe contentin the C2 coating plateaued at 5–6 wt% after long-term oxidation,therefore the upper limit of 7 wt% Fe range in Fig. 12 covers the full Fecomposition range instead of 9.6 wt% from the original composition.From Fig. 12, γ′ phase is destabilizedwith increasing temperature, lead-ing to the disappearance of the γ + γ′ and γ + γ′ + β regions, on thecontrary, a decreasing temperature shows the opposite effect. Similarly,the increasing Fe content in the system destabilizes γ′ phase. Iso-content Al line marks the phase boundary, at a constant Al content, asa function of Fe content in the system. The upward drifting of γ + β/γphase boundary (purple dash line) from isocontent of Al line suggests

0

aver

age

grai

n si

ze (

1

2

Time (h)200 400 600 800 1000

m)

3

4

5

6C1C2

Time (h)0 200 400 600 800 1000

aver

age

grai

n si

ze (

m)

1

2

3

4

5

6C1C2

Time (h)0 1000 2000 3000 4000 5000 6000

aver

age

grai

n si

ze (

m)

0

0.5

1

1.5

2

2.5

3

C1C2

Time (h)0 1000 2000 3000 4000 5000 6000

aver

age

grai

n si

ze (

m)

0

0.5

1

1.5

2

2.5

3

C1C2

Time (h)0 2000 4000 6000 8000 10000

aver

age

grai

n si

ze (

m)

0

1

2

3

4

5C1C2

Time (h)0 2000 4000 6000 8000 10000

aver

age

grai

n si

ze (

m)

0

1

2

3

4

5C1C2

900 °C,

1000 °C,

1100 °C,

900 °C,

1000 °C,

1100 °C,

(c)

(e) (f)

(a)

(d)

(b)

Fig. 10. Average grain size of β and γ phase as a function of oxidation time within the C1 and C2 coatings at 900, 1000 and 1100 °C oxidation.

10 P. Zhang et al. / Materials and Design 166 (2019) 107599

that Fe extends the γ + β dual phase region for a fixed Al content, asindicated by the black arrow. Themechanism bywhich the Fe additionsaffect the phase transformation of the C1 and C2 coatings can beexplained by phase diagram and the microstructural degradationsequence of the C1 and C2 coatings are predicted by the blue and red ar-rows, respectively. At 900 and 1000 °C, the degradation of the C1 systemfollows the phase transformation from γ + γ′ → γ + γ′ + β → γ + β.However, the increase of Fe content in the C2 system destabilizes γ′phase, which leads to γ + β → γ. At 1100 °C, only γ + β → γ transfor-mation can be observed for both coatings since the γ′ phase is no longerstable.

The role of Fe addition induced shifting of phase equilibrium can beexplained from the view of Gibbs free energy. Fig. 13a shows the Gibbsfree energy for β + γ phases in Ni19Co13.5Cr10Al-xFe system(balanced by Ni) as a function of Fe content at 900 °C calculated withthe Thermo-Calc software. Unlike the Co-free Ni-Cr-Al-xFe system [40],the addition of Co inNi-Co-Cr-Al-xFe system stabilizesβ+γ phase equi-libriumand suppresses the formation ofγ′ andα-Cr phase, leavingβ+γphase at the equilibrium condition [50]. It is observed that the absolutevalue of Gibbs free energy of β phase increases with the increasing ofFe content in contrast to that of γ phase. Similar calculated resultswere observed at 1000 and 1100 °C, therefore not presented here.

0 2000 4000 6000 8000 10000Oxidation Time (h)

0.3

0.4

0.5

0.6

pha

se v

olum

e fr

actio

n

C1C2

0 2000 4000 6000Oxidation Time (h)

0.3

0.4

0.5

0.6

pha

se v

olum

e fr

actio

n

C1C2

(a) (b) (c)

0 2000 4000 6000Oxidation Time (h)

0.5

0.6

0.7

0.8

0.9

1

BLZ

thic

knes

s fra

ctio

n (%

)

C1C2

0 2000 4000 6000 8000 10000Oxidation Time (h)

0.5

0.6

0.7

0.8

0.9

1

BLZ

thic

knes

s fra

ctio

n (%

)

C1C2

0 200 400 600 800 1000Oxidation Time (h)

0.5

0.6

0.7

0.8

0.9

1

BLZ

thic

knes

s fra

ctio

n (%

)

C1C2

(d) (e) (f)

0 200 400 600 800 1000Oxidation Time (h)

0.3

0.4

0.5

0.6

pha

se v

olum

e fr

actio

n

C1C2

Fig. 11. β phase fraction of C1 and C2 coatings as a function of time at (a) 900 °C, (b) 1000 °C and (c) 1100 °C oxidation, and the corresponding β-phase left zone (BLZ) thickness fraction(BLZ thickness normalized by the coating thickness) in (d), (e) and (f), respectively.

11P. Zhang et al. / Materials and Design 166 (2019) 107599

The total Gibbs energy of system and Gibbs energy of an individualphase can be expressed as:

Gtotal ¼Xp

f pGp cið Þ ð3Þ

Gp cið Þ ¼Xi

ciGpi þ RT

Xi

ci lnci þXi

XjN i

cic j �Xmn¼0

nLpi; j ci−c j� �n

þXi

XjN i

XkN j

cic jckLpi; j;k ð4Þ

γ

γ+γ’

Fig. 12. Isothermal section for Ni19Co13.5Cr-xAl-yFe at 1000 °C, dash lines denote new phase bindicated. Degradation sequence of C1 and C2 coatings are marked with blue arrow and red ar

where p is the individual phase, fp is the mole fraction of phase p; i, j, kdenote the different components in phase p; R is the gas constant, T isthe absolute temperature; ci is the concentration of component i inphase i; Li, jp and Li, j, kp are binary and ternary interaction parameters,respectively. Following the same discussion in Ref [40], the equilibriumof system is achieved by theminimization of the total Gibbs free energyof the system, i.e. ΔGtotal ≤ 0. Hence, Eq. 3 for a β+ γ dual phase systemcan be simplified as fβ ≥ 1/(1 − |Gβ/Gγ|). When the absolute value ofGibbs free energy of β phase increased while that of γ phase decreased

β+γ

iso-content Al

β+γ+γ’

C1

C2

oundaries at 1100 °C (purple arrow) or 900 °C (green arrow), isocontent line of Al is alsorow, respectively.

Table 4Average chemical composition measured in the BLZ of coatings [in wt%].

Coat. Ni Al Ti Cr Fe Co Mo Ta W Condition

C1 bal. 10.7 ± 0.5 1.0 ± 0.3 13.5 ± 0.4 0.8 ± 0.2 19.0 ± 0.7 1.0 ± 0.3 2.2 ± 0.4 1.0 ± 0.1 1000 °C–8000 hC2 bal. 10.2 ± 0.3 1.3 ± 0.4 12.7 ± 0.7 5.1 ± 0.5 17.9 ± 1.1 1.2 ± 0.1 2.2 ± 0.3 1.1 ± 0.3C1 bal. 10.5 ± 0.7 1.4 ± 0.2 12.8 ± 0.5 0.8 ± 0.3 19.0 ± 0.8 1.0 ± 0.2 1.0 ± 0.2 0.5 ± 0.2 1100 °C–800 hC2 bal. 9.7 ± 0.5 1.5 ± 0.3 12.9 ± 1.0 5.0 ± 0.6 18.0 ± 0.6 1.2 ± 0.2 1.0 ± 0.2 0.9 ± 0.5

Mass percent of Fe in the system0 2 4 6 8 10G

ibbs

Fre

e En

ergy

of d

iffer

etn

phas

es (J

) 104

-8

-7

-6

-5

-4

-3

-2

-1

phase phase

Fig. 13. Gibbs free energy for each phase as a function of the Fe concentration at 900 °C.

12 P. Zhang et al. / Materials and Design 166 (2019) 107599

due to the increasing of Fe content, the mole fraction of β phaseincreased. Since the Gibbs energy of an individual phase is stronglydependent on the concentrations of different components (ci) in thatphase, the increasing β phase fraction by the addition of Fe in theMCrAlX system can be explained.

5.2. Coarsening rate coefficient as a function of Fe content

Fig. 14a shows coarsening rate coefficient ofβ phase inγmatrix andγphase in βmatrix in NiCoCrAl-xFe coating systems as function of temper-ature. The alloy systems are defined in Table 4 to simulate the β and γphase coarsening. From Fig. 14a, the coarsening rate coefficient of bothsystems increases with temperature, and coarsening rate coefficient ofC1 system is higher than that of C2, and γ phase coarsening in β matrixis higher than β phase in γ matrix. The coarsening rate coefficient of βphase in γ matrix and γ phase in β matrix in Ni19Co13Cr10Al-xFe wascalculated as a function of Fe mass percent at 1000 °C in Fig. 14b. It isclear that the coarsening rate coefficient of β/γ phase, especially γphase in βmatrix, decreases with the increasing Fe content, resulting ina retardation of the β and γ grain coarsening. However, such effect, i.e.coarsening of grains ceased at ~2 μm (Fig. 10), is only observed at 900°C oxidation and C2 coating at 1000 °C oxidation. In general, coarseningof particles is also controlled by the short-range element diffusion, lessgrain coarsening at 900 °C oxidation (Fig. 10) is due to a low element dif-fusivity as evidenced by the low coating-substrate interdiffusion at 900 °Cand less grain coarsening of C2 coating at 1000 °C oxidation (Fig. 10) isprobably due to addition of Fe (Fig. 14b); on the contrary, coarsening ofβ grains in C1 coating after 5000 h oxidation at 1000 °C and in both sys-temat 1100 °C oxidation and can be readily observed after 50 h (Fig. 10e).

6. Discussion

6.1. Fe effect on oxidation

Comparedwith those elementswithhigh affinity to oxygen, such asAland reactive element (Y, Hf), the Gibbs formation energy of Fe-oxides andthe corresponding equilibriumoxygenpartial pressure of Fe-oxide forma-tion are close to Ni and Co oxides (NiO and CoO) in MCrAlX system, aslisted in Table 5. At the beginning of the oxidation, O reacts with Ni, Co,Cr, Fe and Al on the coating surface to form an external layer of spinelconsisting of a mixture of oxides [51]. Therefore, the formation of Fe-oxide-containing spinel is expected [52]. With reduced penetration ofoxygen through the external oxide layer, a lower partial pressure ofoxygen level is expected at oxide/coating interface, where only Al2O3growth is thermodynamically feasible [53]. Once Al2O3 grows into acontinuous and dense inner layer, the formation of bi-layer of TGO canbe observed. The alumina scale and outer-β-phase depletion behaviourof C1 and C2 coatings are similar at different conditions (Fig. 2), indicatinglittle influence of increasing Fe content on the alumina growth behaviour.Besides, over-doping of REs [54] in MCrAlY could cause negative effectsuch as fast Al2O3 scale growth. No such negative effect can be foundwith the increasing Fe content in the C2 coating.

6.2. Fe effect on coating-substrate interdiffusion

The interdiffusion between coating and substrate occurs duringoxidation, i.e., Al, Cr and Co diffuse inwards from coating to substrate;

refractory element (Ta, W, Mo) and Ti diffuse outwards from substrateto coating. Inward diffusion of Cr and Co from the coating to thesubstrate destabilized γ′ [46,55–57], leading to the formation ofthe GPDZ. However, due to the rejection of γ′-forming elements(Ti and Ta) from γ in the GPDZ, the formation of (Ta, Ti)-rich carbides(denoted by ‘c’) is more enhanced in the GPDZ than in the bulk sub-strate. The phase transformation in GPDZ can be generally described as:

γ þ γ0 →Al;Cr;Co

γ þ c ð5Þ

Due to the complexity of phase equilibrium in theGPDZ, the Fe effecton the interdiffusionwill be discussed separately for the different oxida-tion temperatures with respect to the inducedmicrostructural changes.

6.2.1. 900 °CThepresence of theσphase is only observed at 900 °C. Small amount

of σ phase can be observed at 500 h in the IBDZ of the C1 coating(Fig. 4a), due to the supply of W and Mo from the substrate enablingthe formation of σ phase. At longer oxidation times, the expansion oftheγ phase zone near the interface dissolvedσdue to thehigh solubilityof the γ phase for W and Mo. Hence, the phase transformation of C1coating can be described by Eq. 6. For C2 coating, large amounts of σphase can be observed. The phase transformation in the GPDZ of C2coating at 900 °C can be modified as:

γ þ γ0 →Al;Cr;Co;Fe

γ þ β þ σ þ c ð6Þ

With the inward diffusion of Fe from coating to substrate, it is clearthat the high Fe content in the GPDZ of the C2 coating greatly affectsthe phase equilibrium: the reprecipitation of secondary β phase wasenhanced and stabilized even at 5000 h (Fig. 4d). The reprecipitationof the β phase in γ matrix enhanced the precipitation of the σ phase

0 2 4 6Mass percent of Fe in the system

0

0.5

1

1.5

2

Coa

rsen

ing

rate

coe

ffic

ient

(m3 /

s) 10-24

in in

900 950 1000 1050 1100Temperature (°C)

0

0.2

0.4

0.6

0.8

1

Coa

rsen

ing

rate

coe

ffic

ient

(m3 /

s) 10-23

in -C1 in -C1 in -C2 in -C2

(a) (b)

Fig. 14. Coarsening rate coefficient of β phase in γmatrix and γ phase in βmatrix in NiCoCrAl-xFe coating systems as function of temperature (a) and mass percent of Fe at 1000 °C (b),calculated by Thermo-Calc software.

13P. Zhang et al. / Materials and Design 166 (2019) 107599

due to rejection of Cr and Co from β, and MC-carbides. The σ grainsformed in the GPDZ of the C2 coating coarsened with the increasing ox-idation time, as evidenced by the coarsening of σ grains observed inboth coating and substrate near the interface after 5000 h oxidation(Fig. 4d). It can be explained by the fact that the coarsening of the βphase further rejected W and Mo, which forced these σ precipitates tocongregate and grow into large precipitates with continuous supply ofCr from the coating. These σ precipitates distributed discontinuouslyin a row near the interface and were rather stable keeping a similarcomposition after 5000 h oxidation, as indicated in Table 3. Some βgrains remained stable after 5000 h oxidation in C2 coating, coexistedwith large σ grains in the IBDZ; on the contrary, a thin discontinuouslayer of β-depletion zone near the coating/substrate interface in C1coating can be observed in Fig. 4b. It has been reported that the σphase in Re-Cr system shows a low interdiffusion coefficient severalorders magnitude lower than γ phase [58]. The potential of Cr-richprecipitation to suppress interdiffusion has been proved in Ref [59,60]. As aforementioned, Fe shows no significant influence on Al diffusiv-ity in IN792 substrate (Fig. 8d). The logarithmic Al diffusivity shows agood linear fitting with inverse temperature according to Eq. 3. Similaractivation energies indicate that the Al diffusion in both the C1 and C2system follows a same thermally activated behaviour, which is indepen-dent of Fe content. However, it is clear that the Al diffusion depth in C2systemwas reduced at 900 °C at 500 h. A similar Al diffusivity but differ-ent diffusion depth in IN792 indicates that certain amounts of Al wasconsumed at the GPDZ by the formation of secondary β phase whenthe Al flux passed through the GPDZ at the beginning of diffusion.Then the rest of Al passing through the GPDZ diffused further into theIN792 substrate, the true Al diffusion depth is therefore “delayed”,giving a less depth for C2 (Fig. 9a). This is consistent with the formationof β + γ + σ region in the GPDZ of C2 system in Fig. 4. Therefore, itis reasonable to assume that the formation of the σ phase diffusionbarrier retards the Al diffusion in the C2 system at 900 °C. Besides, less

Table 5Gibbs free energy of oxide formation at 1000 °C and equilibrium oxygen partial pressurefor different oxides [62].

ΔG (kJ/mol) pO2 (bar)

NiO −254 4.3E-11CoO −278 1.1E-12FexOy −360~−326 9.4E-16–2.8E-14Cr2O3 −536 1.7E-22Al2O3 −844 1.6E-35Y2O3 −1023 7.6E-43

Kirkendall voids can be observed in the C2 coating interface (Fig. 4)after 5000 h oxidation, which again suggests a retardation of coating-substrate interdiffusion behaviour. Kirkendall voids are responsible forreduction in fatigue strength and coating peeling off due to reductionof its adhesion, less degradation ofmechanical properties to the C2 coat-ing system is expected. Furthermore, the reduced Al loss in the coatingcan prolong the coating's oxidation life, as also reported by Jiang et al.[48]. Coating-substrate interdiffusion causes Al-depletion in the coatingreducing coating life span during oxidation, the coating service life canbe prolonged by hindering the Al diffusion with such a diffusion barrierlayer at 900 °C. Thus, the σ phase layer in C2 coating served as an in-situdiffusion barrier, as also reported by Jiang et al. [48].

6.2.2. 1000 °C and 1100 °CWith the increasing of oxidation temperature at 1000 and 1100 °C, σ

phase is not stable for any of the coatings as indicated in Figs. 5 and 6.Certain amount of secondary β phase formed in the GPDZ in both coat-ings, is transformed to γ phase eventually, as indicated by Fig. 7. There-fore, the phase transformation in the GPDZ of both systems at 1000 and1100 °C can be described as.

γ þ γ0 →Al;Cr;Co;Fe

γ þ β þ c →Al;Cr;Co;Feγ þ c ð7Þ

Compared with C1, a thicker GPDZ on the C2 coating due to thepresence of the Fe concentration gradient can be observed. Fig. 15shows element concentration profile near the coating/substrate inter-face after 800 h at 1100 °C oxidation for C1 and C2 coatings. Near thecoating/substrate interface (marked with black dash line), the GPDZ(left) and the IBDZ (right) boundaries are marked with dash lines(blue for C1, red for C2).

The apparent uphill diffusion of Cr can be observed near the coating/substrate interface. A similar Cr activity can be derived in both of thecoating and substrate (roughly between 2.42–2.55 ∗ 10−3 at 1100 °C)before interdiffusoin occurred. After interdiffusion occurred, Al diffusioninto substrate decreases Cr activity in IBDZ, resulted in Cr diffusion fromBLZ to the coating/substrate interface, as also reported by Elsaß et al.[61].Moreover, amuch lower solubility of Cr inγ+γ′ phase (substrate)orβ+γ phase (coating) thanγ phase (GPDZ and IBDZ) also contributesto the uphill diffusion behaviour. Compared between two coating/substrate systems, the large Fe gradient between the C2 coating/substrateshows different impacts on elemental diffusion behaviour: Co diffusion isnot much affected by the addition of Fe indicated by the overlapping ofconcentration profile in GPDZ of C1 and C2 coatings; on the contrary, Crconcentration peak position shifts leftwards, and the gradient of peak is

4

6

8

10

12

Al (

wt %

) C1C2

10

15

20

Co

(wt %

)

C1C2

GPDZ IBDZ

-250 -200 -150 -100 -50 0 50 100Distance ( m)

0

1

2

3

4

5

Fe (w

t %)

C1C2

-250 -200 -150 -100 -50 0 50 100Distance ( m)

10

12

14

16

Cr (

wt %

)

C1C2

Fig. 15. Element concentration profile near the coating/substrate interface after 800 h at 1100 °C oxidation, the coating/substrate interface is marked with black dash line, GPDZ and IBDZboundaries are marked with dash lines (blue for C1, red for C2).

14 P. Zhang et al. / Materials and Design 166 (2019) 107599

reduced resulting in a broader Cr peak in C2 coating. Therefore, a largerGPDZ but smaller IBDZ of C2 coating can be explained. Based on Fig. 15,the Cr concentration profiles intersected at the same Cr content atGPDZ left boundary (~14%) and the IBDZ right boundary (15–16%) for

C

C2

Fig. 16. Isothermal section for NiCoCrAlFe at 1100 °C, β + γ/γ phase bounda

both coatings, suggesting a dominate factor of Cr content for the localphase equilibrium. To better understand Cr role on phase equilibrium,an isothermal section for NiCoCrAlFe system is presented in Fig. 16. Theβ + γ/γ phase boundaries for C1 and C2 coating are marked in blue

1

+

ries are marked for C1 system (blue) and C2 system (red), respectively.

15P. Zhang et al. / Materials and Design 166 (2019) 107599

and red colour, respectively. From Fig. 16, the β + γ/γ boundaryintersects at certain Cr levelwhenAl content is above ~6%,which demon-strates the dominate factor of Cr content; the same theory applies to γ+γ′/γ phase equilibrium at the left GPDZ boundary.

6.3. Fe effect on coating degradation

Both coating surface oxidation reaction and coating-substrateinterdiffusion contribute to the Al depletion of the MCrAlX coating,which result in a coating microstructural degradation. Due to a similarsurface oxidation reaction, the influence of different Fe content on coat-ing degradationmainly depends on the Al interdiffusion, phase stabilityof the coating and grain coarsening in the BLZ. By the formation of an σphase diffusion barrier near the coating/substrate interface at 900 °C,the Al diffusion into the IN792 substrate of the C2 coating system isretarded. Accelerated coating microstructural degradation can be ob-served for oxidation at 1000 and 1100 °C. Based on thermodynamicmodelling results, Fe stabilizes the β phase and extends the β + γphase region to a low Al range. Fe addition also modified Cr diffusionbehaviour which changes the local phase equilibrium. It is clear thatthe addition of Fe shows no influence on Al diffusivity. Consideringthe potential to form in-situ σ phase as interdiffusion barrier, it is inter-esting to investigate the temperature range tomaintain the formation oftheσ phase in C2/IN792diffusion couple; and the possibility to formdif-fusion barrier on other superalloys. This could be the subject of futurework. These factors help to suppress inner-β-phase depletion due tothe coating-substrate interdiffusion, especially for C2 coating withhigh Fe content showing an “incubation” stage at 900 and 1000 °Coxidation. When fast coating degradation occurs at 1100 °C oxidation,the β phase volume fraction of C2 coating plateaus at 45% while thatof C1 coating drops catastrophically. And the β phase volume fractionin the coating decreases as the increase of oxidation time, C2 coatinggenerally shows higher β phase volume fraction in the BLZ at variousconditions (Fig. 11). Another effect of increasing Fe content in thecoating is to reduce the tendency of grain coarsening based on theexperimental observation (Figs. 11 and 12), which is also confirmedby modelling as shown in Fig. 14.

7. Conclusion

In this study, the role of Fe on twoMCrAlX coatings (1.6 and 9.6 wt%of Fe addition for C1 andC2 coatings respectively) on IN792 substrates isstudied with respect to on oxidation and interdiffusion behaviour at900, 1000 and 1100 °C. The major outcomes of this study can be listedas follow:

1. Fe in theMCrAlX coatings takes part in oxidation reaction at the coat-ing surface by the formation of spinel type (Ni,Co,Cr,Fe)xOy oxides atouter layer of TGO due to a similar oxide Gibbs formation energy asCo and Ni. However, the increase of the Fe content doesn't affectthe Al2O3 growth and the overall oxidation resistance of the coating;

2. Combined experimental results and thermodynamic calculationsshow that increasing Fe content in the MCrAlX coating increasesthe β phase fraction and stabilizes the β phase by extending theβ + γ phase region to the lower Al region in the phase diagram;

3. High Fe content shows great influence on coating-substrate inter-diffusion depending on oxidation temperature. However, it showsminor influence on Al diffusivity in IN792, except that the forma-tion of σ phase layer retained certain amount of Al in the GPDZ at900 °C oxidation. A significant effect is that it modifies Cr diffu-sion, which alters the local phase equilibrium near the coating/substrate interface, and therefore reduces the inner-β-depletionof the C2 coating;

4. The grain coarsening in the BLZ of the C2 coating, especially duringoxidation at 1000 °C, is retarded by the high Fe addition, which isalso confirmed by modelling;

5. The results in this paper combine experiment and thermodynamiccalculations and present a full investigation of the effect of differentFe level addition on oxidation and interdiffusion of MCrAlX coatedIN792. It improves the understanding of the effect of Fe addition onMCrAlX coating performance, and also validates the reliability touse thermodynamic calculation to guide MCrAlX coating composi-tion design.

CRediT authorship contribution statement

Pimin Zhang: Conceptualization, Data curation, Formal analysis, Investi-gation, Methodology, Software, Validation, Visualization, Writing - origi-nal draft, Writing - review & editing. Xin-Hai Li: Conceptualization,Writing - review & editing. Johan Moverare:Writing - review & editing.Ru Lin Peng: Conceptualization, Writing - review & editing.

Acknowledgement

Financial support from Siemens Industrial Turbomachinery AB(Finspång, Sweden) and Swedish Energy Agency through KME consor-tiumare greatly acknowledged. In addition, the authors are also gratefulfor the Swedish Government Strategic Research Area in MaterialsScience on Functional Materials at Linköping University (Faculty GrantSFO-Mat-LiU 2009-00971) for financial support.

Data availability

All data included in this study are available upon request by contactwith the corresponding author.

References

[1] N. Eliaz, G. Shemesh, R.M. Latanision, Hot corrosion in gas turbine components, Eng.Fail. Anal. 9 (2002) 31–43, https://doi.org/10.1016/S1350-6307(00)00035-2.

[2] J.R. Nicholls, N.J. Simms, W.Y. Chan, H.E. Evans, Smart overlay coatings — conceptand practice, Surf. Coat. Technol. 149 (2002) 236–244, https://doi.org/10.1016/S0257-8972(01)01499-2.

[3] M.J.M.J. Pomeroy, Coatings for gas turbine materials and long term stability issues,Mater. Des. 26 (2005) 223–231, https://doi.org/10.1016/j.matdes.2004.02.005.

[4] P. Zhang, K. Yuan, R.L. Peng, X.H. Li, S. Johansson, Long-term oxidation of MCrAlYcoatings at 1000 °C and an Al-activity based coating life criterion, Surf. Coat. Technol.332 (2017) 12–21, https://doi.org/10.1016/j.surfcoat.2017.09.086.

[5] Y. Chen, X. Zhao, P. Xiao, Effect ofmicrostructure on early oxidation ofMCrAlY coatings,Acta Mater. 159 (2018) 150–162, https://doi.org/10.1016/j.actamat.2018.08.018.

[6] B.Y. Zhang, G.J. Yang, C.X. Li, C.J. Li, Non-parabolic isothermal oxidation kinetics oflow pressure plasma sprayed MCrAlY bond coat, Appl. Surf. Sci. 406 (2017)99–109, https://doi.org/10.1016/j.apsusc.2017.02.123.

[7] J. Wang, M. Chen, Y. Cheng, L. Yang, Z. Bao, L. Liu, et al., Hot corrosion of arc ion platingNiCrAlY and sputtered nanocrystalline coatings on a nickel-based single-crystal super-alloy, Corros. Sci. 123 (2017) 27–39, https://doi.org/10.1016/J.CORSCI.2017.04.004.

[8] E. Lang, E. Bullock, The effect of coatings on the HTmechanical properties of Ni-basesuperalloys, Final Rep. Cost. (1982) 50.

[9] K. Natesan, Corrosion-erosion behavior of materials in a coal-gasification environ-ment, Corrosion 32 (1976) 364–370.

[10] J. Toscano, R. Vaβen, A. Gil, M. Subanovic, D. Naumenko, L. Singheiser, et al., Param-eters affecting TGO growth and adherence onMCrAlY-bond coats for TBC, Surf. Coat.Technol. 201 (2006) 3906–3910, https://doi.org/10.1016/j.surfcoat.2006.07.247.

[11] A. Gil, V. Shemet, R. Vassen, M. Subanovic, J. Toscano, D. Naumenko, et al., Effect ofsurface condition on the oxidation behaviour of MCrAlY coatings, Surf. Coat.Technol. 201 (2006) 3824–3828, https://doi.org/10.1016/j.surfcoat.2006.07.252.

[12] J. Wang, M. Chen, L. Yang, L. Liu, S. Zhu, F. Wang, et al., The effect of yttrium additionon oxidation of a sputtered nanocrystalline coating with moderate amount of tanta-lum in composition, Appl. Surf. Sci. 366 (2016) 245–253, https://doi.org/10.1016/j.apsusc.2016.01.088.

[13] B.a. Pint, Experimental observations in support of the dynamic-segregation theoryto explain the reactive-element effect, Oxid. Met. 45 (1996) 1–37, https://doi.org/10.1007/BF01046818.

[14] D. Naumenko, V. Shemet, L. Singheiser, W.J. Quadakkers, Failure mechanisms ofthermal barrier coatings on MCrAlY-type bondcoats associated with the formationof the thermally grown oxide, J. Mater. Sci. 44 (2009) 1687–1703, https://doi.org/10.1007/s10853-009-3284-3.

[15] T. Amano, T. Taguchi, Spalling of the surface oxide formed on Ni-(20,40,60,80)Cr al-loys with small additions of Ce and Si, J. Alloys Compd. 193 (1993) 20–22, https://doi.org/10.1016/0925-8388(93)90297-Z.

[16] K. Ogawa, K. Ito, T. Shoji, D.W.W. Seo, H. Tezuka, H. Kato, Effects of Ce and Si addi-tions to CoNiCrAlY bond coatmaterials on oxidation behavior and crack propagation

https://doi.org/10.1016/S1350-6307(00)00035-2https://doi.org/10.1016/S0257-8972(01)01499-2https://doi.org/10.1016/S0257-8972(01)01499-2https://doi.org/10.1016/j.matdes.2004.02.005https://doi.org/10.1016/j.surfcoat.2017.09.086https://doi.org/10.1016/j.actamat.2018.08.018https://doi.org/10.1016/j.apsusc.2017.02.123https://doi.org/10.1016/J.CORSCI.2017.04.004http://refhub.elsevier.com/S0264-1275(19)30019-X/rf0040http://refhub.elsevier.com/S0264-1275(19)30019-X/rf0040http://refhub.elsevier.com/S0264-1275(19)30019-X/rf0045http://refhub.elsevier.com/S0264-1275(19)30019-X/rf0045https://doi.org/10.1016/j.surfcoat.2006.07.247https://doi.org/10.1016/j.surfcoat.2006.07.252https://doi.org/10.1016/j.apsusc.2016.01.088https://doi.org/10.1016/j.apsusc.2016.01.088https://doi.org/10.1007/BF01046818https://doi.org/10.1007/BF01046818https://doi.org/10.1007/s10853-009-3284-3https://doi.org/10.1007/s10853-009-3284-3https://doi.org/10.1016/0925-8388(93)90297-Zhttps://doi.org/10.1016/0925-8388(93)90297-Z

16 P. Zhang et al. / Materials and Design 166 (2019) 107599

of thermal barrier coatings, J. Therm. Spray Technol. 15 (2006) 640–651, https://doi.org/10.1361/105996306X147081.

[17] R. Mevrel, Cyclic oxidation of high-temperature alloys, Mater. Sci. Technol. 3 (1987)531–535, https://doi.org/10.1080/02670836.1987.11782264.

[18] B.A. Pint, K.L. More, I.G. Wright, Effect of quaternary additions on the oxidation be-havior of Hf-doped NiAl, Oxid. Met. 59 (2003) 257–283, https://doi.org/10.1023/A:1023087926788.

[19] T. Gheno, B.C. Zhou, A. Ross, X. Liu, G. Lindwall, Z.K. Liu, et al., A thermodynamic ap-proach to guide reactive element doping: Hf additions to NiCrAl, Oxid. Met. 87(2017) 297–310, https://doi.org/10.1007/s11085-016-9706-0.

[20] V.K. Tolpygo, D.R. Clarke, Surface rumpling of a (Ni, Pt)Al bond coat induced by cy-clic oxidation, Acta Mater. 48 (2000) 3283–3293, https://doi.org/10.1016/S1359-6454(00)00156-7.

[21] P. Dai, Q. Wu, Y. Ma, S. Li, S. Gong, The effect of silicon on the oxidation behavior ofNiAlHf coating system, Appl. Surf. Sci. 271 (2013) 311–316, https://doi.org/10.1016/j.apsusc.2013.01.189.

[22] B.A. Pint, On the formation of interfacial and internal voids in α-Al2O3 scales, Oxid.Met. 48 (1997) 303–328, https://doi.org/10.1007/BF01670505.

[23] D. Naumenko, B.A. Pint, W.J. Quadakkers, Current thoughts on reactive element ef-fects in alumina-forming systems: in memory of John Stringer, Oxid. Met. 86(2016)https://doi.org/10.1007/s11085-016-9625-0.

[24] N. Czech, F. Schmitz, W. Stamm, Improvement of MCrAlY coatings by addition ofrhenium, Surf. Coat. Technol. 68–69 (1994) 17–21, https://doi.org/10.1016/0257-8972(94)90131-7.

[25] W. Beele, N. Czech, W.J. Quadakkers, W. Stamm, Long-term oxidation tests on a re-containing MCrAlY coating, Surf. Coat. Technol. 94–95 (1997) 41–45, https://doi.org/10.1016/S0257-8972(97)00473-8.

[26] B. Han, Y. Ma, H. Peng, L. Zheng, H. Guo, Effect of Mo, Ta, and Re on high-temperature oxidation behavior of minor Hf doped β-NiAl alloy, Corros. Sci. 102(2016) 222–232, https://doi.org/10.1016/j.corsci.2015.10.011.

[27] T.A. Taylor, D.F. Bettridge, Development of alloyed and dispersion-strengthenedMCrAlY coatings, Surf. Coat. Technol. 86–87 (1996) 9–14, https://doi.org/10.1016/S0257-8972(96)02961-1.

[28] P. Fox, G.J. Tatlock, Effect of tantalum additions on oxidation of overlay coated superal-loys, Mater. Sci. Technol. 5 (1989) 816–827, https://doi.org/10.1179/mst.1989.5.8.816.

[29] L. Yang, M. Chen, J. Wang, Z. Bao, S. Zhu, F. Wang, Diffusion of Ta and its influence onoxidation behavior of nanocrystalline coatings with different Ta, Y and Al contents,Corros. Sci. 126 (2017) 344–355, https://doi.org/10.1016/j.corsci.2017.07.017.

[30] R. Chen, X. Gong, Y. Wang, G. Qin, N. Zhang, Y. Su, et al., Microstructure and oxida-tion behaviour of plasma-sprayed NiCoCrAlY coatings with and without Ta onTi44Al6Nb1Cr alloys, Corros. Sci. 136 (2018) 244–254, https://doi.org/10.1016/j.corsci.2018.03.008.

[31] A.G. Mora-García, H. Ruiz-Luna, M. Mosbacher, R. Popp, U. Schulz, U. Glatzel, et al.,Microstructural analysis of Ta-containing NiCoCrAlY bond coats deposited byHVOF on different Ni-based superalloys, Surf. Coat. Technol. 354 (2018) 214–225,https://doi.org/10.1016/J.SURFCOAT.2018.09.025.

[32] W.J. Quadakkers, V. Shemet, D. Sebold, R. Anton, E. Wessel, L. Singheiser, Oxidationcharacteristics of a platinizedMCrAlY bond coat for TBC systems during cyclic oxida-tion at 1000 °C, Surf. Coat. Technol. 199 (2005) 77–82, https://doi.org/10.1016/j.surfcoat.2004.11.038.

[33] B. Gleeson, W. Wang, S. Hayashi, D.J. Sordelet, Effects of platinum on the interdiffu-sion and oxidation behavior of Ni-Al-based alloys, Mater. Sci. Forum 461–464(2004) 213–222, https://doi.org/10.4028/www.scientific.net/MSF.461-464.213.

[34] P. Zhang, R. Lin Peng, X.-H. Li, S. Johansson, Investigation of element effect on high-temperature oxidation of HVOF NiCoCrAlX coatings, Coatings 8 (2018) 129, https://doi.org/10.3390/coatings8040129.

[35] R. Prescott, M.J. Graham, The formation of aluminum oxide scales on high-temperature alloys, Oxid. Met. 38 (1992) 233–254, https://doi.org/10.1007/BF00666913.

[36] A.Weisenburger, A. Heinzel, G. Müller, H. Muscher, A. Rousanov, T91 cladding tubeswith and without modified FeCrAlY coatings exposed in LBE at different flow, stressand temperature conditions, J. Nucl. Mater. 376 (2008) 274–281, https://doi.org/10.1016/j.jnucmat.2008.02.026.

[37] J. Zhang, A review of steel corrosion by liquid lead and lead-bismuth, Corros. Sci. 51(2009) 1207–1227, https://doi.org/10.1016/j.corsci.2009.03.013.

[38] M.N. Gussev, K.G. Field, Y. Yamamoto, Design, properties, and weldability of ad-vanced oxidation-resistant FeCrAl alloys, Mater. Des. 129 (2017) 227–238, https://doi.org/10.1016/j.matdes.2017.05.009.

[39] J. Eklund, B. Jönsson, A. Persdotter, J. Liske, J.E. Svensson, T. Jonsson, The influence of sil-icon on the corrosion properties of FeCrAl model alloys in oxidizing environments at600 °C, Corros. Sci. 144 (2018) 266–276, https://doi.org/10.1016/j.corsci.2018.09.004.

[40] K. Ma, F. Tang, J.M. Schoenung, Investigation into the effects of Fe additions on theequilibrium phase compositions, phase fractions and phase stabilities in the Ni-Cr-Al system, Acta Mater. 58 (2010) 1518–1529, https://doi.org/10.1016/j.actamat.2009.10.059.

[41] K. Yuan, R. Lin Peng, X.H. Li, S. Johansson, Hot corrosion behavior of HVOF-sprayedCoNiCrAlYSi coatings in a sulphate environment, Vacuum 122 ( (2015) 47–53,https://doi.org/10.1016/j.vacuum.2015.09.015.

[42] J. Yang, Q. Zheng, H. Zhang, X. Sun, H. Guan, Z. Hu, Effects of heat treatments on themicrostructure of IN792 alloy, Mater. Sci. Eng. A 527 (2010) 1016–1021, https://doi.org/10.1016/j.msea.2009.10.026.

[43] S. Saeidi, K.T. Voisey, D.G. McCartney, The effect of heat treatment on the oxidationbehavior of HVOF and VPS CoNiCrAlY coatings, J. Therm. Spray Technol. 18 (2009)209–216, https://doi.org/10.1007/s11666-009-9311-8.

[44] Thermo-Calc Software TCS, Ni-Alloys Mobility Database Version 4, 2015.[45] Q. Chen, J. Jeppsson, J. Ågren, Analytical treatment of diffusion during precipitate

growth in multicomponent systems, Acta Mater. 56 (2008) 1890–1896, https://doi.org/10.1016/j.actamat.2007.12.037.

[46] M. Elsaß, M. Frommherz, A. Scholz, M. Oechsner, Interdiffusion in MCrAlY coatednickel-base superalloys, Surf. Coat. Technol. 307 ( (2016) 565–573, https://doi.org/10.1016/j.surfcoat.2016.09.049.

[47] D. Texier, D. Monceau, Z. Hervier, E. Andrieu, Effect of interdiffusion on mechanicaland thermal expansion properties at high temperature of a MCrAlY coated Ni-basedsuperalloy, Surf. Coat. Technol. 307 ( (2016) 81–90, https://doi.org/10.1016/J.SURFCOAT.2016.08.059.

[48] S.M. Jiang, C.Z. Xu, H.Q. Li, J. Ma, J. Gong, C. Sun, High temperature corrosion behav-iour of a gradient NiCoCrAlYSi coating I: microstructure evolution, Corros. Sci. 52(2010) 1746–1752, https://doi.org/10.1016/j.corsci.2010.01.025.

[49] H. Danflou, M. Macia, T.H. Sanders, T. Khan, Mechanisms of formation of serratedgrain boundaries in nickel base superalloys, Superalloys 1996 1996, pp. 119–127,https://doi.org/10.1023/A:1018694716756.

[50] D.R.G. Achar, R. Munoz-Arroyo, L. Singheiser, W.J. Quadakkers, Modelling of phaseequilibria in MCrAlY coating systems, Surf. Coat. Technol. 187 (2004) 272–283,https://doi.org/10.1016/j.surfcoat.2004.02.018.

[51] F. Cao, B. Tryon, C.J. Torbet, T.M. Pollock, Microstructural evolution and failure char-acteristics of a NiCoCrAlY bond coat in “hot spot” cyclic oxidation, Acta Mater. 57(2009) 3885–3894, https://doi.org/10.1016/j.actamat.2009.04.039.

[52] R. Fetzer, A. Weisenburger, A. Jianu, G. Müller, Oxide scale formation of modifiedFeCrAl coatings exposed to liquid lead, Corros. Sci. 55 (2012) 213–218, https://doi.org/10.1016/j.corsci.2011.10.019.

[53] Y.Z. Liu, X.B. Hu, S.J. Zheng, Y.L. Zhu, H. Wei, X.L. Ma, Microstructural evolution of theinterface between NiCrAlY coating and superalloy during isothermal oxidation,Mater. Des. 80 (2015) 63–69, https://doi.org/10.1016/j.matdes.2015.05.014.

[54] A. Strawbridge, P.Y. Hou, The role of reactive elements in oxide scale adhesion,Mater. High Temp. 12 (1994) 177–181, https://doi.org/10.1080/09603409.1994.11689484.

[55] J.D. WhITTENBERGER, in: C.T. Sims, N.S. Stoloff, W.C. Hagel (Eds.), A Review of:“Superalloys II”, Wiley-Interscience Publication John Wiley & Sons, New York, NY,1987https://doi.org/10.1080/10426919208947432 , (615 pages, hardcover WileyEd, 1992).

[56] T. Galiullin, A. Chyrkin, R. Pillai, R. Vassen, W.J. Quadakkers, Effect of alloying ele-ments in Ni-base substrate material on interdiffusion processes in MCrAlY-coatedsystems, Surf. Coat. Technol. 350 (2018) 359–368, https://doi.org/10.1016/J.SURFCOAT.2018.07.020.

[57] J. Zhong, J. Liu, X. Zhou, S. Li, M. Yu, Z. Xu, Thermal cyclic oxidation and interdiffusionof NiCoCrAlYHf coating on a Ni-based single crystal superalloy, J. Alloys Compd. 657(2016) 616–625, https://doi.org/10.1016/j.jallcom.2015.10.148.

[58] T. Narita, F. Lang, K.Z. Thosin, T. Yoshioka, T. Izumi, H. Yakuwa, et al., Oxidation be-havior of nickel-base single-crystal superalloy with rhenium-base diffusion barriercoating system at 1,423 K in air, Oxid. Met. 68 (2007) 343–363, https://doi.org/10.1007/s11085-007-9079-5.

[59] E. Godlewska, K. Godlewski, Chromaluminizing of nickel and its alloys, Oxid. Met. 22(1984) 117–131, https://doi.org/10.1007/BF00656900.

[60] K. Zhang, Q.M. Wang, C. Sun, F.H. Wang, Preparation and oxidation behaviorof NiCrAlYSi coating on a cobalt-base superalloy K40S, Corros. Sci. 50 (2008)1707–1715, https://doi.org/10.1016/j.corsci.2008.01.033.

[61] M. Elsaß, M. Frommherz, A. Scholz, M. Oechsner, Interdiffusion in MCrAlY coatednickel-base superalloys, Surf. Coat. Technol. 307 ( (2016) 565–573, https://doi.org/10.1016/j.surfcoat.2016.09.049.

[62] I. Barin, O. Knacke, O. Kubaschewski, Thermochemical Properties of InorganicSubstances, Springer-Verlag, 1977https://doi.org/10.1007/978-3-662-02293-1.

https://doi.org/10.1361/105996306X147081https://doi.org/10.1361/105996306X147081https://doi.org/10.1080/02670836.1987.11782264https://doi.org/10.1023/A:1023087926788https://doi.org/10.1023/A:1023087926788https://doi.org/10.1007/s11085-016-9706-0https://doi.org/10.1016/S1359-6454(00)00156-7https://doi.org/10.1016/S1359-6454(00)00156-7https://doi.org/10.1016/j.apsusc.2013.01.189https://doi.org/10.1016/j.apsusc.2013.01.189https://doi.org/10.1007/BF01670505https://doi.org/10.1007/s11085-016-9625-0https://doi.org/10.1016/0257-8972(94)90131-7https://doi.org/10.1016/0257-8972(94)90131-7https://doi.org/10.1016/S0257-8972(97)00473-8https://doi.org/10.1016/S0257-8972(97)00473-8https://doi.org/10.1016/j.corsci.2015.10.011https://doi.org/10.1016/S0257-8972(96)02961-1https://doi.org/10.1016/S0257-8972(96)02961-1https://doi.org/10.1179/mst.1989.5.8.816https://doi.org/10.1016/j.corsci.2017.07.017https://doi.org/10.1016/j.corsci.2018.03.008https://doi.org/10.1016/j.corsci.2018.03.008https://doi.org/10.1016/J.SURFCOAT.2018.09.025https://doi.org/10.1016/j.surfcoat.2004.11.038https://doi.org/10.1016/j.surfcoat.2004.11.038https://doi.org/10.4028/www.scientific.net/MSF.461-464.213https://doi.org/10.3390/coatings8040129https://doi.org/10.3390/coatings8040129https://doi.org/10.1007/BF00666913https://doi.org/10.1007/BF00666913https://doi.org/10.1016/j.jnucmat.2008.02.026https://doi.org/10.1016/j.jnucmat.2008.02.026https://doi.org/10.1016/j.corsci.2009.03.013https://doi.org/10.1016/j.matdes.2017.05.009https://doi.org/10.1016/j.matdes.2017.05.009https://doi.org/10.1016/j.corsci.2018.09.004https://doi.org/10.1016/j.actamat.2009.10.059https://doi.org/10.1016/j.actamat.2009.10.059https://doi.org/10.1016/j.vacuum.2015.09.015https://doi.org/10.1016/j.msea.2009.10.026https://doi.org/10.1016/j.msea.2009.10.026https://doi.org/10.1007/s11666-009-9311-8http://refhub.elsevier.com/S0264-1275(19)30019-X/rf0220https://doi.org/10.1016/j.actamat.2007.12.037https://doi.org/10.1016/j.actamat.2007.12.037https://doi.org/10.1016/j.surfcoat.2016.09.049https://doi.org/10.1016/j.surfcoat.2016.09.049https://doi.org/10.1016/J.SURFCOAT.2016.08.059https://doi.org/10.1016/J.SURFCOAT.2016.08.059https://doi.org/10.1016/j.corsci.2010.01.025https://doi.org/10.1023/A:1018694716756https://doi.org/10.1016/j.surfcoat.2004.02.018https://doi.org/10.1016/j.actamat.2009.04.039https://doi.org/10.1016/j.corsci.2011.10.019https://doi.org/10.1016/j.corsci.2011.10.019https://doi.org/10.1016/j.matdes.2015.05.014https://doi.org/10.1080/09603409.1994.11689484https://doi.org/10.1080/09603409.1994.11689484https://doi.org/10.1080/10426919208947432https://doi.org/10.1016/J.SURFCOAT.2018.07.020https://doi.org/10.1016/J.SURFCOAT.2018.07.020https://doi.org/10.1016/j.jallcom.2015.10.148https://doi.org/10.1007/s11085-007-9079-5https://doi.org/10.1007/s11085-007-9079-5https://doi.org/10.1007/BF00656900https://doi.org/10.1016/j.corsci.2008.01.033https://doi.org/10.1016/j.surfcoat.2016.09.049https://doi.org/10.1016/j.surfcoat.2016.09.049https://doi.org/10.1007/978-3-662-02293-1

The iron effect on oxidation and interdiffusion behaviour in MCrAlX coated Ni-base superalloys1. Introduction2. Experimental3. Thermodynamic modelling4. Results4.1. Oxidation4.2. Microstructure evolution near the interdiffusion zone4.2.1. Microstructure at the coating/substrate interface4.2.2. IBDZ&GPDZ4.2.3. Al concentration profiles in the substrate

4.3. Grain size/microstructure evolution in coating

5. Thermodynamic modelling results5.1. Fe effect on phase equilibrium5.2. Coarsening rate coefficient as a function of Fe content

6. Discussion6.1. Fe effect on oxidation6.2. Fe effect on coating-substrate interdiffusion6.2.1. 900°C6.2.2. 1000°C and 1100°C

6.3. Fe effect on coating degradation

7. Conclusionsection21AcknowledgementData availabilityReferences