Effect of temperature and dissolved oxygen on stress corrosion cracking behavior of P92 ferritic-martensitic steel in supercritical water environment

Accepted Manuscript Effect of temperature and dissolved oxygen on stress corrosion cracking behavior of P92 ferritic-martensitic steel in supercritical water environment Z. Zhang, Z.F. Hu, L.F. Zhang, K. Chen, P.M. Singh PII:

S0022-3115(17)30067-3

DOI:

10.1016/j.jnucmat.2017.10.024

Reference:

NUMA 50556

To appear in:

Journal of Nuclear Materials

Received Date: 11 January 2017 Revised Date:

8 October 2017

Accepted Date: 9 October 2017

Please cite this article as: Z. Zhang, Z.F. Hu, L.F. Zhang, K. Chen, P.M. Singh, Effect of temperature and dissolved oxygen on stress corrosion cracking behavior of P92 ferritic-martensitic steel in supercritical water environment, Journal of Nuclear Materials (2017), doi: 10.1016/ j.jnucmat.2017.10.024. 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.

ACCEPTED MANUSCRIPT

Effect of temperature and dissolved oxygen on stress corrosion cracking behavior of P92 ferritic-martensitic steel in supercritical water environment

Jiangsu Key Laboratory of Advanced Structural Materials and Application Technology, Nanjing 211167, China b School of Materials Science and Engineering, Nanjing Institute of Technology, Nanjing 211167, China c School of Materials Science and Engineering, Tongji University, Shanghai 201804, China d School of Nuclear Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, China e School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta 30332-0245, USA

SC

a

RI PT

Z. Zhanga,b,*, Z. F. Huc, L. F. Zhangd , K. Chend, P. M. Singhe

TE D

M AN U

Abstract: The effect of temperature and dissolved oxygen (DO) on stress corrosion cracking (SCC) of P92 martensitic steel in supercritical water (SCW) was investigated using slow strain rate test (SSRT) and fractography analysis. The SSRT was carried out at temperatures of 400, 500, 600 oC in deaerated supercritical water and at DO contents of 0, 200, 500 ppb at the temperature of 600 oC, respectively. The results of SSRT show that the decrease of ductility at the temperature of 400 oC may be related to the dynamic strain aging (DSA) of P92 steel. The degradation of the mechanical properties in SCW is the joint effect of temperature and SCC. Compared with the effect of temperature, DO in SCW has no significant effect on the SCC susceptibility of P92 steel. The observation of oxide layer shows that large numbers of pores are nucleated in the oxide layer, which is related to vacancy accumulation and hydrogen generated in the oxide layer. Under the combined action of the growth stress and tensile stress, micro cracks, nucleated from the pores in the oxide layer, are easily propagated intergranularly outward to the surface of specimen, and fewer cracks can extend inward along the intrinsic pores to the inner oxide/metal interface, which is the reason for the exfoliation of oxide films.

EP

Key words: ferritic-martensitic (F/M) steel; supercritical water (SCW); temperature; dissolved oxygen; stress corrosion cracking

AC C

*Corresponding author. E-mail addresses: [email protected] (Z. Zhang), [email protected] (Z. F. Hu), [email protected] (L. F. Zhang), [email protected] (K. Chen), [email protected] (P. M. Singh)

1. Introduction

The supercritical water-cooled reactor (SCWR) under the Generation IV program is a promising advanced nuclear systems due to their high thermal efficiency (i.e., about 45 % vs. 33 % efficiency for current Light Water Reactors LWRs), optimization and simplification design. SCWRs are operating at higher pressure and temperatures than LWRs with a direct once-through cycle to increase the fuel efficiency [1-3]. In the SCWRs, the operating temperature is expected to be in the range 500~550 oC, which is above the thermodynamic critical point (374.2 oC, 22.1 MPa), and the water in the system exists in the state of supercritical water [4]. Supercritical water, with significantly different corrosion properties as compared to liquid water, is very aggressive for metallic materials 1

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

being highly corrosive and it may cause stress corrosion cracking during the operation of materials. For the widely application of the advanced nuclear systems, high performance is required for the structural material in terms of its weldability, radiation resistance, creep and fatigue strength, corrosion and stress corrosion cracking susceptibility, etc. [5] The development and characterization of the candidate alloys for specific reactor components at the expected temperatures and pressures is therefore one of the key issues of SCWRs [2]. Among the potential structural materials for SCWR design, austenitic steels, ferritic-martensitic (F/M) steels and Ni-based alloys are considered as candidates for future SCWRs [3-6]. Among these classes of alloys, austenitic steels have a relatively low thermal conductivity, high thermal expansion coefficient and poor weld ability. The Ni-based alloys also have low thermal conductivity and relative high-temperature strength. Most importantly, these two alloys contain substantial amount of Ni, which is susceptible to transmutation under the influence of irradiation [7]. Therefore, due to the lower cost and low swelling under irradiation [2, 8], the low Ni or Ni-free F/M steels present an attractive alternative for in-core and out-of-core application in SCWRs. As one of the advanced F/M steels, P92 is a relatively new improved martensitic heat-resistant steel, with higher creep-rupture strength combined with good oxidation and corrosion resistance at elevated temperatures [9, 10]. It has been used or considered for use as nuclear fission and fusion reactor components as well. However, many SCC failures have been found in components where water is stagnant. These locations may accumulate or remain high content of oxygen. Moreover, there will be temperature gradient during the startup and shutdown of the system. Therefore, it is of great importance to clarify the effects of dissolved oxygen in SCW as well as temperature on SCC of F/M steels. In the previous studies, investigations on the corrosion behavior of 9~12 % Cr F/M steels exposed to SCW have been carried out, and multiple growth mechanisms of the oxide films on the surface of material were proposed [11-13]. Moreover, the SCC susceptibility of F/M steels at room and higher temperature were also investigated [14-18], and results clearly show a decrease in the ultimate tensile stress at a temperature range from 500 to 600 oC in SCW environment. As for the dependence of SCC on dissolved oxygen (DO) content for F/M steel, it indicates that thickness of the surface oxide layer depends on DO content in the water [18, 19]. Actually, the oxide growth rate and associated weight gain of these kinds of steels are dependent on the dissolved oxygen concentration [2, 20]. The formation of a hematite layer formed at the higher dissolved oxygen content SCW may also play a role in the nucleation and growth of stress corrosion cracks. Although considerable efforts have been made, ambiguity still exists due to the complexity of stress corrosion cracking (SCC) mechanism. Therefore, the major purpose of this work is to understand the influence of temperature and dissolved oxygen on SCC susceptibility of the F/M steel from macro and micro scale. Slow strain rate tensile tests (SSRT) was carried out to identify the SCC susceptibility at different conditions. Scanning electron microscopy (SEM) was used to observe the steel surface morphology after fracture and study the nucleation and growth mechanism of stress corrosion cracks.

2. Material and experimental 2.1. Steel sample preparation In this study, the test specimens were machined from a section of a commercially produced P92 pipe in longitudinal direction and tested in as-received condition, i.e. without additional heat treatment prior to the exposure to experimental environment. Nominal chemical composition, specified by the ASTM standard [21], and the measured chemical composition for the steel used in this study is given in Table 1. In order to get fully tempered martensitic structure, the final heat treatment consisted of austenitizing at 1050 oC for 4 hours (followed by air-cooling) and tempering at 760 oC for 6 hours (followed by air-cooling). The initial microstructure of P92 steel acquired by transmission electron microscopy (TEM) is shown in Fig. 1. The prior austensite grain structure is composed of block martensites, which consist of laths decorated with stringers of M23C6 carbides (with particle size 2

ACCEPTED MANUSCRIPT

RI PT

in the range of 40~250 nm) that were Cr enriched. Moreover, there also existed some fine MX type carbides (with particle size of 6~20 nm) rich in V and Nb, which were homogeneously dispersed in the inter-lath regions. In the process of martensitic transformation, the strong interactions between austenite grain boundaries and martensitic laths result in high dislocation densities in the range 1014 m-2 in the matrix [22], while typical values for dislocation densities in pure metals and alloys are 1012 m-2 (undeformed). The basic mechanical properties of P92 steel were acquired from the tensile tests conducted in air at the temperature of 20~600 oC. Three tensile tests were conducted with a strain rate of 5 mm/min for each temperature to get the average value. Table 1 Chemical composition of P92 steel Mn

P

S

Si

Cr

W

Mo

V

ASTM

0.07-

0.3-

≤0.02

≤0.01

≤0.5

8.5-

1.5-

0.3-

A335-2003

0.13

0.6

9.5

2.0

Test

0.1

0.45

8.82

1.57

0.015

0.008

0.3

(b)

N

B

Al

Ni

0.15-

0.04-

0.03

0.001-

≤0.04

≤0.4

0.6

0.25

0.09

0.07

0.006

0.35

0.2

0.078

0.037

0.0027

0.006

0.11

Lath boundaries

M AN U

(a)

Nb

SC

C

Subgrain boundaries

TE D

(wt, %)

Block boundaries

M23C6 MX

Prior austenite grain boundaries

Fig. 1. TEM morphology of tempered martensite structure of P92 steel (a) and schematic diagram of 9Cr

EP

ferritic-martensitic steel [23, 24] (b)

AC C

2.2. Slow strain rate tests and corrosion environment Cylindrical tensile samples (Fig. 2) were machined from as-received P92 steel pipe for SSRT according to the ASTM standard G 129 [25]. The surface of the tensile sample was dry-abraded up to 2000 grit, degreased with ethanol and dried in blowing air. The actual dimensions of these samples after polishing were measured. All tests were started immediately after the polishing and cleaning of a tensile sample. Part of SSRT tests were carried out with a strain rate of 1.0×10-6 s-1 and pressure of 25 MPa in deaerated supercritical water at temperature of 400, 500 and 600 oC to evaluate the effect of temperature on the SCC susceptibility of the steel. The others were conducted at the temperature of 600 oC in aerated supercritical water with dissolved oxygen content of 200, 500 ppb to study the effect of DO on the SCC susceptibility of the steel. All tests were performed in recirculating water loop which can simulate the primary water condition of the pressurized water reactor. The schematic diagram of water loop is given in Fig. 3. It can operate up to temperature of 650 oC and pressure of 31 MPa. The water purity was continuously monitored by inlet and outlet conductivity. DO content was controlled by continuously bubbling of high purity gas or a mixture of high purity gas, such as argon, or a mixture of argon with various percentage of oxygen, depending on the required DO level. The autoclave was heated with three phase ceramic band heaters and temperature was controlled by an artificial intelligent temperature controller. SSRT in Ar at RT and in SCW at 3

ACCEPTED MANUSCRIPT

SC

RI PT

elevated temperature was firstly conducted to compare the mechanical properties of the steel in a non-corrosive environment and a corrosive environment of SCW, respectively. After all SSRT tests, the total elongation and reduction in area of failed specimens were measured to obtain the SCC susceptibility of P92 steel in different environment. The fracture surfaces of specimens and the morphology of the oxide layer were observed under a scanning electron microscope (SEM) to analyze the nucleation and growth of cracks in the oxide layer in SCW.

AC C

EP

TE D

M AN U

Fig. 2. Shape and dimension of specimen (units: mm)

Fig. 3. Schematic diagram of SSRT testing system with recirculating water loop

3. Results and Discussion 3.1. Mechanical properties and dynamic strain aging The basic mechanical properties of P92 steel at different temperatures in air are given in Table 2 and Fig. 4. The phenomenon of Portevin-Le Chatelier (PLC), which reflects as the abrupt changes in stress and elongation, can be observed in the range of 400~500 oC. A serrated flow known as PLC effect has already been observed and studied 4

ACCEPTED MANUSCRIPT in these kinds of 9~12 % Cr steels on tensile curves and hysteresis loops [26, 27]. The range of temperature is close to the expected operating temperature of P92 steel in SCWR and greatly modifies the mechanical properties of the steel, which can be critical for the use of the steel in this application. The effect of PLC results from repeated pinning of dislocations, i.e., dynamic strain aging (DSA). This DSA phenomenon accompanied with the abrupt change in mechanical properties will also have an effect on the SCC behavior of P92 martensitic steel in SCW. Table 2 Mechanical properties of P92 steel at different temperatures YS (MPa)

UTS (MPa)

Total elongation (%)

20

547

678

30.2

200

460

580

19.4

300

449

535

18.6

400

504

514

16.6

500

431

472

15.2

600

354

383

18.0

68.6

66.3 66.4 63.8

SC

64.6 78.0

M AN U

800

600

o

200 C

(a)

o

20 C

o

400

o

500 C

200

0

TE D

Stress (MPa)

Reduction of area (%)

RI PT

o

Temperature ( C)

5

10

15

o

600 C

20

Strain (%)

AC C

EP

0

300 C o 400 C

5

25

30

35

ACCEPTED MANUSCRIPT 800

(b)

Stress (MPa)

600

400

200 100

200

300

400

500

100

80

60

700

(c)

Elongation Reduction of area

40

20

0 0

TE D

Elongation and Reduction of Area (%)

600

M AN U

Temperature (ºC)

SC

0

RI PT

Yield stress Ultimate tensile stress

100

200

300

400

500

600

700

EP

Temperature (ºC)

AC C

Fig. 4. The mechanical properties of P92 steel at different temperatures: (a) stress-strain curve, (b) yield and tensile stress, (c) elongation and reduction of area 3.2. Stress-strain behavior of steel in SSRT Tests The results obtained from SSRT of P92 martensitic steel at different temperature and DO content in SCW, as well as other austenitic steels for comparison [6], are listed in Table 3. The corresponding stress-strain curves are shown in Fig. 5. Compared with the stress-strain curve in Ar, the curves tested in deaerated SCW environment have the lower ultimate tensile strength (UTS) and yield strength (YS). It indicates that SCW has great influence on the mechanical behavior of steel. However, the influence of temperature and DO on the mechanical behavior of P92 steel in SCW environment is different, as shown in Fig. 5(a) and Fig. 5(b). In Fig. 5(a), UTS and YS of the steel all drop with the increasing temperature. However, the total elongation to fracture has a minimum value at 400 oC, which is similar to the abrupt change of tensile property of P92 steel observed in Table 2 and Fig. 2. This phenomenon reported and investigated in this kind of steel is attributed to interaction of dislocations with interstitial solute atoms [28, 29]. Katada [30] and Atkinson [31] have shown that DSA will promote the 6

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

environmental assisted cracking (EAC) of materials. In fact, temperature and SCC are two significant factors that affect the strength and ductility of material. In Ar environment, there is no SCC effect and the temperature dominates the properties of the steel. Under a high temperature and certain stress, the material will be subjected to irreversible plastic deformation along with the increasing of time, i.e., creep. The increase of temperature will promote the atom diffusion and plastic damage during the test. While in the SCW environment, the degradation of properties is the joint effect of creep and SCC. Previous studies [32, 33] have shown that creep contribute largest portion to the total crack growth rate of alloy 690 in SCW environment, which is related to the extensive cavity formation ahead of the creak tip. As for P92 martensitic steel, the elevated working temperature and the exposure under stress promote microstructure changes in the subgrains and precipitates due to diffusion process. In order to characterize the effect of temperature on the SCC of material, further fracture morphology will be analyzed using SEM. In Fig. 5(b), The DO has a relatively negligible effect on the YS of steel in SCW, compared with the effect of temperature. The YS is independent of DO content, which is mainly because the effect of oxygen on the specimens was negligible during the elastic deformation period. During the plastic deformation period, when the oxygen has enough time to diffuse and react with samples, consequently increase in the SCC susceptibility of specimens may occurs and further analysis will be discussed in the micro scale to study the effect of DO on the steel in SCW. It can also be seen that, compared with the SSRT results of P92 martensitic steel, the mechanical properties of austenitic steel in SCW at 600 oC are better, especially for the UTS and total elongation of the steel. It shows the better corrosion resistance and high-temperature stability of austenitic steel in SCW at 600 oC. Table 3 SSRT results of P92 martensitic steel and other austenitic steels under different environments Materials

Type

Strain rate (s-1)

Tp (oC)

Environment

UTS (MPa)

YS (MPa)

Total elongation, A (%)

20

Ar

693

597

39.4

Ar

250

231

30.8

400

Deaerated SCW

574

560

16.7

500

Deaerated SCW

394

384

26.5

Deaerated SCW

224

181

22.3

200ppb DO

215

180

21.4

500ppb DO

214

180

20.8

347

172

28.9

325

135

31.2

Martensitic steel

P92

TE D

600 1.0×10

-6

HR3C 316Ti

9.26×10-7

< 10 ppb DO

AC C

Austenitic steel

EP

600

7

ACCEPTED MANUSCRIPT 800

(a) Ar o

400 C 400

RI PT

Stress (MPa)

600

o

500 C 200

o

600 C 0

10

30

40

M AN U

300

20

Strain (%)

SC

0

Temperature: 600ºC

(b)

Ar

Stress (MPa)

200

0 µg/L

200 µg/L

100

0

TE D

500 µg/L

0

10

20

30

EP

Strain (%)

AC C

Fig. 5. Stress-strain curves of P92 steel in SCW: (a) temperature, (b) DO The susceptibility to SCC can be evaluated using the ratio between percentage elongation to failure or UTS in aggressive medium and inert environment, such as argon, air or inert solution. To investigate the influence of temperature and DO on SCC susceptibility in SCW environment, UTS loss rate Is and elongation loss rate IA were calculated respectively, and the definition of the rate are

I S = (1 −

σE ) × 100% σ0

(1)

I A = (1 −

AE ) × 100% A0

(2)

where σE and AE are the UTS and elongation of steel in SCW environment, respectively, and σ0 and A0 are UTS and elongation of steel measured in Ar. The results of Is and IA are shown in Fig. 6. It can be seen that Is and IA increase 8

ACCEPTED MANUSCRIPT with increase of DO content in SCW, however, the increment is relatively small. It is reported that the SCC susceptibility I equals or smaller than 20 % is considered as an index of immunity to SCC [34], it is possible to conclude that SCC occurs in SCW at 600 oC but there is no significant difference between results in SCW with or without DO. 33

44.0

32

31

43.0

30

42.5

IA (%)

Is (%)

43.5

RI PT

44.5

29

42.0

41.5

SC

Is IA

28

27

0

100

200

300

500

M AN U

DO (ppb)

400

Fig. 6. The influence of DO on SCC susceptibility calculated by UTS and elongation loss rate

AC C

EP

TE D

3.3. Fracture morphology after SSRT To investigate the effect of temperature and DO on SCC behavior of P92 steel in the micro scale, scanning electron microscopy (SEM) analysis was performed on fractured specimens to identify the fracture morphology. The SEM photographs of the fracture surface morphology obtained after testing at different temperature are shown in Fig. 7. It can be seen that the fracture morphology of P92 steel at 400 oC exhibits totally different characteristics compared to that at 500 and 600 oC. The fracture morphology of specimen at 400 oC is a kind of ductile tensile fracture, however, the reduction of area is obviously decreased compared with that at 500 and 600 oC. Moreover, there is almost no obvious stress corrosion cracks (SCCs) on the surface of specimen near the fracture. The reduction of ductility of the steel at 400 oC is related to the occurrence of DSA in this range of temperature. When the material is subjected to a plastic deformation at a certain strain rate and temperature range, the occurrence of DSA can result in strain concentration or plastic strain localization at the tip of crack. This kind of local concentration will promote the breakdown of oxide film in the front of crack tip. As a consequence, the disruption of the protective film will accelerate the nucleation and growth of crack in SCW environment. In comparison, the fractures at 500 and 600 oC are typical stress corrosion fracture and there are large numbers of stress corrosion cracks observed on the surface of specimen near the fracture. The reduction in area is increased with increase of temperature, which manifests a high degree of plastic deformation. The presence of stress corrosion cracks can also decrease the effective load bearing area and result in the decline of UTS. In Fig. 8, the fracture morphology of specimens investigated in SCW with different DO content shows similar feature, and large numbers of stress corrosion cracks can be observed on the surface of specimen near the fracture. To characterize the relationship of SCC with temperature and DO, the stress corrosion cracks on the side surface of the SSRT specimens were observed, as shown in Fig. 7 and Fig. 8. The number of the stress corrosion cracks, along with the maximum crack length and width was counted and measured in a random area of 0.65 mm × 0.65 mm on the side surface of specimen. The average value was obtained from three measurements and the statistical results are shown in Fig. 9. It can be seen that the initiation and growth of stress corrosion cracks is strongly affected by the temperature of 9

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

SCW. Stress corrosion cracks are detected at 400 oC, however, numerous small stress corrosion cracks are visible at higher magnification on the side surface, as is shown by the inset picture in Fig. 7(b). The number of stress corrosion cracks on the surface of specimen is decreased, while the maximum length and width of stress corrosion cracks are increased with the increasing temperature. It might result from the interaction and coalescence of small stress corrosion cracks. Once the stress corrosion cracks develop on the gauge surface, the cracks will grow faster due to the decrease in steel strength. Therefore, the SCC of P92 steel can be accelerated dramatically with the increase in temperature under the constant tensile strain rate. It is well known that a high temperature accelerates all the degradation process in steels. The movement of dislocations and the migration of grain boundaries can be thermally activated by increasing of temperature. As for P92 martensitic steel, it has been proved that its structure is composed of martensite laths decorated with stringers of M23C6 carbides at the boundaries and high dislocation densities in the range 1014 m-2 inside the subgrains. During plastic deformation at high temperature, the thermally activated moving dislocations in the subgrain are easily piled up at the low-angle lath and prior austensite grain boundaries. According to Mughrabi [35], the local accumulation of pile-ups caused by this irreversibility is the origin of crack initiation. Moreover, the small M23C6 carbides at the boundaries may also interact with the moving dislocations and lead to dislocation stress concentration at the boundaries. Therefore, these boundaries may be the sites susceptible to initiation of stress corrosion crack in SCW environment. According to the results in Fig. 5(b), DO plays a minor role in increasing SCC susceptibility of the steel in SCW, which is confirmed by the change of crack number in Fig. 9. It can be seen that the crack density is approximately the same at 600 °C with 0, 200 and 500 ppb of DO. However, the length of crack is increased with increasing of DO. Previous studies [2, 19] have reported that the thickness and growth rate of oxide layer of martensitic steels increase dramatically with increasing of DO content in SCW. Thick oxide layer can exert higher stress at crack tips, which might be the reason why the crack grows faster with the increasing DO content. As for the slight decrease in the crack width in higher DO content, enough long corrosion time might result in the corrosion of crack in higher DO and the newly formed crack will be covered by the newly formed oxide layer. The morphologies of oxide layer of P92 steel will be further observed and analyzed to understand the effect of temperature and DO on the nucleation and propagation of cracks in the oxide layer.

10

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

Fig. 7. SEM morphologies of the fracture surface obtained after testing at different temperatures in SCW environment: (a) (b) 400 oC, (c) (d) 500 oC, (e) (f) 600 oC

11

RI PT

ACCEPTED MANUSCRIPT

3000

1000

2

Crack number (n/mm )

Crack width 800

2000

1500

0

TE D

500

Crack number

400 / 0

500 / 0

600 / 0

600

400

Crack length

80

Maximum crack length (µ m)

2500

1000

100

Crack number Maximum crack length Maximum crack width

60

40

20

Maximum crack width (µ m)

M AN U

SC

Fig. 8. SEM morphologies of the fracture surface obtained after testing in SCW environment with different DO content: (a) (b) 200 ppb, (c) (d) 500 ppb

200 0 600 / 200

600 / 500

o

EP

Temperature and DO content ( C/ppb)

AC C

Fig. 9. The variation of SCC size on the side surface of specimen 3.4. Morphology and structure of oxide layer To understand the structure and morphology of oxide layer, and the nucleation and growth of cracks in the oxide layer of the P92 steel in SCW environment, axial cross-sectional observation of tested SSRT specimens was carried out, as shown in Fig. 10. It can be seen that almost all the stress corrosion cracks occur near the necking region due to large deformation in this region. The difference in deformation behavior between the oxide layer and base metal will result in the breakdown of oxide layer under tensile stress. In general, the deformation behavior of oxide layer under tensile strain can be described by three contributions [36]: elastic contribution, plastic contribution mainly by diffusional creep in the oxide, and contribution from lateral-oxide growth effects. Among them, the plastic contribution is rather small sometimes negligible due to the fact that the necessary five independent dislocation slip systems do not exist in the temperature range [37]. Once the oxide layer breaks, a fresh metallic surface is exposed, and the corrosion proceeds by the outward diffusion of iron ions and the inward 12

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

diffusion of oxygen. In the oxide layer, some pores are firstly nucleated in the front area of the crack tip, as can be seen in the magnified images in Fig. 10(a). Under the action of internal and external stress, the micro cracks can nucleate from these pores and propagate in the oxide layer by connecting the adjacent micro cracks. Fewer cracks, which initiated from the pores in the outer oxide layer, can propagate inward along the intrinsic pores in the inner oxide layer to the inner-metal interface, as can be seen in Fig. 10(b). These through-cracks and interface-cracks will undermine the stability of the oxide layer, and eventually lead to the breakdown of the oxide layer.

Fig. 10. Cross-sectional SEM images of fractures after SSRT at 600 oC/ 0 ppb in SCW: (a) Pores nucleate in the front area of crack tip, (b) Cracks propagate in the oxide layer 13

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Similar surface morphology and the oxide layer structure, along with the chemical composition are shown in Fig. 11. It has been reported that ferritic-martensitic steels exhibit a three-layered oxide structure with a Fe-rich outer layer containing only large columnar Fe3O4, a Cr-rich inner oxide layer containing a mixture of small equiaxed Fe3O4 and FeCr2O4 and a diffusion layer containing a mixture of Cr-rich oxide precipitates located along the lath and grain boundaries [38]. The outer oxide layer is distinct as shown in Fig. 11 and the outer-inner interface is thought to be the original metal-water interface. This is in agreement with the oxidation mechanism of an outer oxide formed by outward diffusion of the iron cations (Fe2+), and inner oxide formed by inward diffusion of oxygen anions (O2-) [39]. The EDX scan line in Fig. 11 contains the iron Kα, the chromium Kα and the oxygen Kα. The EDX plots clearly show the three oxide layers. The outer oxide layer contains only iron and oxygen, which has an approximate thickness of 10 µm, is mainly composed of Fe3O4. While the inner oxide layer is enriched in chromium compared to the base metal and the diffusion layer, which has an approximate thickness of 12 µm, is composed of Fe-Cr spinel oxide. The chromium content of the diffusion layer is intermediate to that of the metal and the inner oxide layer, and the thickness of the diffusion layer is 1~2 µm. Additionally, a chromium enrichment peak (circled) is noticed at the interface of inner and diffusion layer, which indicates the formation of Cr2O3 film at the inner oxide/metal interface [40, 13]. It has been reported that this continuous layer of Cr2O3 can stop or dramatically slow down the diffusion of oxygen beyond the layer and therefore serves as a barrier for further oxidation [40]. It can also be seen that there is zone of porosity located right at the inner oxide-diffusion layer interface, which is also adjacent to the location of the Cr2O3 phase. The slower diffusion of both iron and oxygen in Cr2O3 compared to their diffusion in spinel [41], leading to the inability of the iron diffusing from the diffusion layer to compensate for the influx of iron vacancies, can result in the formation of pores at this interface [13, 42]. There are also relative tiny pores in the inner layer, as shown in Fig. 10 and Fig. 11. Bischoff et al [43] have reported that these small pores located in the inner layer are linked with the chromium-rich regions. Since the pores are believed to form through the coalescence of iron vacancies that migrate inward from the highly oxidized outer oxide layer, which suggests that iron migrates preferentially from these chromium-rich regions. The micro cracks nucleated in the outer layer can propagate inward along the small pores into the base metal, as shown in Fig. 10(b). Compared with the relatively dense and protective inner layer, the outer layer is porous and non-protective, therefore, the micro cracks are preferentially nucleated in the outer oxide layer and make it more non-protective in SCW. The following analysis is then focused on the outer layer to study the nucleation and propagation of cracks.

14

SC

RI PT

ACCEPTED MANUSCRIPT

M AN U

Fig. 11. SEM morphologies and EDX data of the oxide layer after SSRT at 600 oC/ 0 ppb in SCW environment

AC C

EP

TE D

The morphology of outer oxide layer of P92 steel after SSRT in SCW environment are shown in Fig. 12. As can be seen in Fig. 12(a), the outer oxide layer formed on the surface of P92 steel after SSRT in 400 oC SCW is dense and composed of large equiaxed grains. While at the temperature of 500 oC, many dispersively distributed pores are detected on the surface of outer oxide layer. The size and number of the holes on the surface of outer oxide layer grows with increasing temperature, which is consistent with the previous work [38, 44]. Some researchers believed that the growth of hole may be related to the dissolution of the oxide on the outer surface of oxide layer in SCW. Previous studies showed that the iron oxides can dissolve and deposit when the alloys are exposed to the high temperature water environments [45, 46]. When the oxides dissolve, the metal ions diffuse into SCW with increasing exposure time, leaving some vacancies in the oxide layer. As a consequence, the cavities or micro pores are formed due to vacancy accumulation [47]. With the outward diffusion of the iron cations (Fe2+) and reaction with H2O, the Fe3O4 in the outer layer grows. The thickness of outer oxide layer measured in Fig. 12 in different SCW environment is shown in Fig. 13, it can be seen that the thickness of the outer oxide layer is increased with increasing of temperature and DO. According to the morphologies in Fig. 11, the thickness of the inner oxide layer is almost the same as that of the outer oxide layer. With the inward transport of oxygen anions (O2-), both the outer Fe3O4 oxide layer and Fe-Cr inner oxide layer became thicker. Apart from the accumulation of vacancies in the oxide layer, some pores may also nucleate in the oxide layer by trapping or ejection of hydrogen [48]. The related process and reaction equations will be discussed in the next part. Zhu et al [49] investigated the role of DO in the oxidation of martensitic steel and showed that both the molecular oxygen and molecular water contributed to the growth of oxide scale. They used 18O2 isotope for oxidation study and showed that the concentration of the oxidant -18O2 was approximately six orders of magnitude smaller than that of the oxidant -H216O at the SCW/oxide surface, which indicated the main oxygen source for the oxidation process under high temperature SCW conditions is water. It can also be used to explain the big difference of stress-strain curves between the test in Ar and SCW environment, and the relatively small effect of DO on the SCC behavior of material in Fig. 5. The outer oxide layer shows a rather high porosity. The number of pores may provide a path for inward short circuit transport of oxygen and water molecules, resulting in the growth of the inner layer oxide [50, 51].

15

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

16

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

EP

Fig. 12. SEM morphologies of oxide layer after SSRT in different SCW environment: (a) 400 oC/0 ppb, (b) 500 oC/0 ppb, (c) 600 oC/0 ppb, (d) 600 oC/200 ppb, (e) 600 oC/500 ppb 25

Outer layer thickness

Outer layer thickness (µm)

AC C

20

15

10

5

0 400 / 0

500 / 0

600 / 0

600 / 200

600 / 500

o

Temperature and DO content ( C/ppb)

Fig. 13. The variation of oxide thickness in different SCW environment 17

ACCEPTED MANUSCRIPT

3Fe + 4H 2O → Fe3O4 + 4H 2

EP

2Cr + 3H 2 O → Cr2O3 + 3H 2

TE D

M AN U

SC

RI PT

3.5. Initiation and growth of cracks in oxide layer Scanning electron micrographs in Fig. 14 show morphologies of secondary cracks observed on the fracture after SSRT tests in SCW at 600 oC. These morphologies are obtained at the edge of fracture to analyze the initiation and propagation of cracks in the oxide layer. It can be seen that with increasing of DO content from 0 to 500 ppb in SCW, more secondary cracks are observed and they are propagated intergranularly in the outer oxide layer. It indicates the oxide layer become more un-protective in SCW environment with the higher DO. Fig. 15 is the SEM morphologies of the radial cross section of specimens. It shows that pores nucleated in the oxide layer after SSRT tests in SCW environment, as shown in Fig. 15(a). SCW with high temperature and DO can promote the formation of large numbers of submicron pores in the outer magnetite layer, especially at the interface of outer-inner oxide layer. As shown in Fig. 15(b), the porous outer layer is initially exfoliated from the original metal-water interface due to the growth of interface-cracks at the interface of the outer-inner oxide layer. The secondary cracks observed are formed when the through-cracks in the oxide layer extend outward along the magnetite grain boundaries to the surface of fracture. Moreover, the micro cracks nucleated in the outer oxide layer can also extend inward along the intrinsic pores in the inner oxide layer to the inner oxide/metal interface, as can be seen in Fig. 10(b). Eventually, the entire oxide layer can be exfoliated from the metal with the growth of cracks along the inner oxide-metal interface, as shown in Fig. 15(b). Based on the related literature [40-48] and the morphologies observed in the present work, the nucleation and propagation process of cracks in the oxide layer could be reasonably described by the following steps in Fig. 16. At the initial stage of exposure in SCW, Fe diffuses outwards along grain boundaries and short circuit paths, in the meantime, the columnar magnetite (Fe3O4) forms and H2 is released through the reaction (3). At the same time, Cr can react with the inward migration of oxygen and the discrete Cr2O3 islands are formed at the original metal/H2O interface (reaction (4)). And then Cr2O3 produced further reacts with Fe through the reaction (5) and forms Fe-Cr spinel (FeCr2O4) and more H2 is released. At the magnetite/H2O interface, an extra discontinuous hematite (Fe2O3) is formed through the reaction (6) at the grain boundary of magnetite [52, 53] and the network of porous outer oxide layer is formed due to the dissolution of the oxide. (3) (4) (5)

2Fe3O4 + H 2 O → 2Fe2O3 + H 2

(6)

AC C

Cr2O3 + Fe + H 2O → FeCr2O4 + H 2

In the outer oxide layer, some pores are formed near the outer-inner oxide interface, as shown in Fig. 15, which may be caused by trapping or ejection of hydrogen released by the reactions. On the one hand, the hydrogen released by the reactions in the oxide layer may diffuse outward through the oxide layer. Fujii and Meussner [54, 55] suggested that the outward diffusing hydrogen may form water molecular where it encounters a sufficient oxygen potential, which may possibly triggering dissociation of the layer structure into the enlarging pores. The DO in SCW can affect oxidation kinetics by reacting with metallic ions and increase the oxygen potential distribution through the oxide scale [49]. Therefore, the outward iron migration and hydrogen diffusion are increased, meanwhile, the inward oxygen penetration is simultaneously increased with higher DO content. This is the main reason that the thickness of oxide layer is increased and the enlarging process of pores is accelerated. Moreover, the hematite (Fe2O3) is easily formed on the surface of oxide layer in such high oxygen partial pressure environment at high temperature [56, 57]. When the oxide formed dissolves with exposure time, the oxygen outside can easily 18

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

diffuse through the porous network and react with the outward diffusing hydrogen. All of these factors promote the initiation and growth of pores in oxide layer of high-oxygen exposed specimens. On the other hand, the hydrogen released by the reactions in the oxide layer may also diffuse inward to the metal. As known, hydrogen embrittlement (HE) is a failure mode caused by the presence of a relatively small amount of hydrogen. It may trigger catastrophic failures at small applied loads and cause degradation of ductility and toughness of material [58, 59]. In general, microstructural heterogeneities such as dislocations, grain boundaries, precipitates and vacancies can act as traps of hydrogen. According to the hydrogen-enhanced localized plasticity (HELP) mechanism, hydrogen must be transported by moving dislocations in the material. During SSRT, the generated hydrogen is redistributed with the hydrogen-enhanced mobility of dislocations to the boundaries and the dislocation-precipitate interfaces so that the concentration of hydrogen at these sites increases. As a consequence, the boundary cohesive strength and the interaction energy of dislocation-obstacles interaction is decreased [60], and micro cracks occur by micro voids coalescence at these sites. Once the crack is nucleated, according to adsorption-induced dislocation emission (AIDE) mechanism, hydrogen can trigger the release of dislocations from the advancing crack tip, causing crack growth and intense deformation in the crack vicinity [59].

19

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

Fig. 14. SEM morphologies of secondary SCCs after testing in different DO content SCW environment: (a) 0 ppb, (b) 200 ppb, (c) 500 ppb

Fig. 15. SEM morphologies of pores nucleated in the oxide layer after SSRT at 600 oC/ 2000 ppb (a) Pores nucleated in the outer oxide layer, (b) Exfoliation of the outer and inner oxide layer 20

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

During SSRT in SCW environment, there are three sources of stress in the oxide layer: one is the growth stress or residual stress which results from the local volume expansion during growth of oxide layer. The other one is thermal stress, which is caused by cooling/heating or thermo-cycling. It may also participate cracking due to the differences in thermal expansion coefficients between the base metal and the oxide. But in this case the thermal stress is negligible due to the stable temperature control during the SSRT. The last one is the applied tensile stress. Based on the observations and analysis above, the schematic of the cracks formed in the oxide layer is shown in Fig. 16. As shown in Fig. 16(a), the nucleation of pores is the result of vacancy accumulation and hydrogen released in the oxide layer. Under the combined action of oxide layer growth stress and applied tensile stress, micro cracks are nucleated from the pores and propagated in the oxide layer, as shown in Fig. 16(b). The cracks extend intergranularly outward to the surface of specimen, which leads to the formation of secondary cracks on the surface of the fracture, as shown in Figs. 14(b) and 14(c). Considering that the outer oxide layer is porous and non-protective while the inner layer is dense and protective, the outer layer is firstly exfoliated from the outer-inner oxide interface and makes the inner oxide layer exposed to the SCW environment. Moreover, few cracks nucleated in the outer oxide layer can propagate inward along the tiny pores in the inner oxide layer. With the growth of cracks at the inner oxide/metal interface, as a consequence, the entire oxide layer is exfoliated from the base metal, as shown in Fig. 16(c).

21

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 16. Schematic of the cracks formed in the oxide layer of P92 steel in SCW environment: (a) Pores nucleated in the oxide layer, (b) Micro cracks propagated along the pores, (c) Exfoliation of the oxide layer

22

ACCEPTED MANUSCRIPT 4. Conclusions

M AN U

SC

RI PT

Based on the SSRT results and SEM analysis of the fractures samples, the main conclusions are drawn as following: (1) The decrease of ductility at intermediate temperature of 400~500 oC is attributed to the occurrence of dynamic strain aging in P92 steel, which may affects the SCC behavior of the steel in SCW environment. The degradation of the mechanical properties in SCW environment is the joint effect of temperature and SCC. (2) Compared with the effect of temperature, DO in SCW has no significant effect on the SCC susceptibility of P92 steel, which may be related to the fact that the main oxygen source for the oxidation process in SCW is water instead of molecular oxygen. High temperature and high oxygen potential distribution through the oxide scale can accelerate the diffusion of ions and formation of large numbers of pores in the oxide layer. (3) The nucleation of pores in the oxide layer is the result of vacancy accumulation and generated hydrogen. With the inward diffusing of hydrogen, the hydrogen-enhanced mobility of dislocations pile up at prior austensite and martensite lath boundaries make the crack easily nucleated at these boundaries. (4) Under the combined action of the growth stress and tensile stress, the micro cracks are nucleated from the enlarging pores in the oxide layer. The cracks in the oxide layer are easily propagated intergranularly outward to the surface of specimen, and fewer cracks can also extend inward along the intrinsic pores to the inner oxide/metal interface, which is the reason for the exfoliation of oxide layer in SCW environment.

AC C

References

EP

TE D

Acknowledgement This work was jointly supported by Shanghai Key Laboratory for R&D and Application of Metallic Functional Materials in Tongji University, Corrosion Laboratory for Nuclear Power Materials in Shanghai Jiao Tong University, Corrosion and Materials Chemistry Research Laboratory in Georgia Institute of Technology and Jiangsu Key Laboratory of Advanced Structural Materials and Application Technology in Nanjing Institute of Technology. The authors are also grateful for the financial support from the National Natural Science Foundation of China (with grant No. 50871076) and International Exchange Program for Graduate Students, Tongji University (with grant No. 2016020004).

[1] T. Nishitani, M. Enoeda, A. Akiba, et al., Recent progress in solid breeder blanket development at JAEA, Fusion. Sci. Technol. 52 (2007) 971–978. [2] G.S. Was, P. Ampornrat, G. Gupta, et al., Corrosion and stress corrosion cracking in supercritical water, J. Nucl. Mater. 371 (2007) 176–201.

[3] N. Muthukumar, J.H. Lee, A. Kimura, SCC behavior of austenitic and martensitic steels in supercritical pressurized water, J. Nucl. Mater. 417 (2011) 1221–1224. [4] R. Novotny, P. Hahaner, J. Siegl, et al., Stress corrosion cracking susceptibility of austenitic stainless steels in supercritical water conditions, J. Nucl. Mater. 409 (2011) 117–123. [5] Y. Watanabe, H. Abe, Y. Daigo, et al., Environmentally assisted cracking of sensitized stainless steel in supercritical water : effects of physical property of water, GENES4/ANP2003, 2003, Paper 1183. [6] L. Tan, X. Ren, K. Sridharan, et al., Corrosion behavior of Ni-based alloys for advanced high temperature water-cooled nuclear plants, Corros. Sci. 50 (2008) 3056–3062. 23

ACCEPTED MANUSCRIPT [7] T.R. Allen, Workshop on Higher Temperature Materials for Advanced Nuclear Energy Systems, DOE Office of Nuclear Energy, Science and Technology, La Jolla, CA, 2002, March. [8] R.L. Klueh, A.T. Nelson, Ferritic/martensitic steels for next-generation reactors, J. Nucl. Mater. 371 (2007) 37– 52. [9] P.J. Ennis, A.Czyrska-Filemonowicz, Recent advances in creep-resistant steels for power plant applications. Sadhana. 28 (2003) 709–730. [10] A. Vyrostkova, A. Kroupa, J. Janovec, et al., Carbide reactions and phase equilibria in low alloy Cr–Mo–V

RI PT

steels tempered at 773–993 K. Part I: Experimental measurements.Acta Mater. 46 (1998) 31–38.

[11] L. Martinelli, F. Balbaud-Célérier, A. Terlain, et al., Oxidation mechanism of a Fe–9Cr–1Mo steel by liquid Pb–Bi eutectic alloy (Part I), Corros. Sci. 50 (2008) 2523–2536.

[12] L. Tan, X. Ren, T.R. Allen, Corrosion behavior of 9–12% Cr ferritic–martensitic steels in supercritical water, Corros. Sci. 52 (2010) 1520–1528.

SC

[13] J. Bischoff, A.T. Motta, Oxidation behavior of ferritic–martensitic and ODS steels in supercritical water, J. Nucl. Mater. 424 (2012) 261–276.

[14] A. Ramar, P. Spätig, R. Schaublin, Analysis of high temperature deformation mechanism in ODS EUROFER

M AN U

97 alloy, J. Nucl. Mater. 382 (2008) 210–216.

[15] S.S. Hwang, B.H. Lee, J.G. Kim, et al., SCC and corrosion evaluations of the F/M steels for a supercritical water reactor, J. Nucl. Mater. 372 (2008) 177–181.

[16] H.S. Cho, H. Ohkubo, N. Iwata, et al., Improvement of compatibility of advanced ferritic steels with super critical pressurized water toward a higher thermally efficient water-cooled blanket system, Fusion Eng. Des. 81 (2006) 1071–1076.

[17] J.H. Lee, R. Kasada, A. Kimura, et al., Influence of alloy composition and temperature on corrosion behavior of ODS ferritic steels, J. Nucl. Mater. 417 (2011) 1225–1228.

TE D

[18] H. Je, A. Kimura, Stress corrosion cracking susceptibility of oxide dispersion strengthened ferritic steel in supercritical pressurized water dissolved with different hydrogen and oxygen contents, Corros. Sci. 78 (2014) 193–199.

[19] N.Q. Zhang, X. Hong, B.R. Li, et al., Influence of the dissolved oxygen content on corrosion of the ferritic-

EP

martensitic steel P92 in supercritical water, Corros. Sci. 56 (2012) 123–128. [20] L.Z. Tan, Y. Yang, T. R. Allen, Porosity prediction in supercritical water exposed ferritic/martensitic steel HCM12A, Corros. Sci 48 (2006) 4234–4242. 2003.

AC C

[21]A 335/A 335M-03 Standard Specification for Seamless Ferritic Alloy-Steel pipe for High-Temperature Service, [22] J. Pesicka, R. Kuzel, A. Dronhofer, et al., The evolution of dislocation density during heat treatment and creep of tempered martensite ferritic steel, Acta Materialia, 51 (2003) 4847–4862. [23] M. Hayakawa, T. Hara, S. Matsuoka, et al., Microstructural observation of tempered martensite in mediumcarbon low-alloy steel by atomic force microscopy, J. Jpn. Inst. Met. 64(6) (2000)460–466. [24] M. Hayakawa, S. Matsuoka, K. Tsuzaki, Observations of prior austenite grain boundaries and carbides in the same area of tempered martensite in medium-carbon steel by atomic force microscopy, J Jpn. Inst. Met. 65 (2001) 734–741. [25] G129-00 Standard Practice for Slow Strain Rate Testing to Evaluate the Susceptibility of Metallic Materials to Environmentally Assisted Cracking, 2013. [26] H.W. Zhou, Y.Z. He, H. Zhang, et al., Influence of dynamic strain aging pre-treatment on the low-cycle fatigue behavior of modified 9Cr-1Mo steel, Int. J Fatigue. 47 (2013) 83–89. 24

ACCEPTED MANUSCRIPT [27] B. Fourniera, F. Dalle, M. Sauzay, et al., Comparison of various 9-12 % Cr steels under fatigue and creep-fatigue loadings at high temperature, Mater. Sci. Eng.A 528 (2011) 6934–6945. [28] N.C.Santhi Srinivas, P. Verma, V. Singh., Dynamic strain aging behavior of modified 9Cr-1Mo steel under monotonic and cyclic loading, Procedia Eng. 183 (2017) 765–772. [29] T. Sakthivel, K. Laha, M. Nandagopal et al., Influence of temperature on tensile flow and work hardening behaviour of modified 9Cr–1Mo steel. Mater. High Temp. 31 (2014) 60–68. [30] Y. Katada, N. Nagata, The effect of temperature on fatigue crack growth behavior of a low-alloy pressure

RI PT

vessel steel in a simulated BWR environment, Corros. Sci. 25 (1985) 693–704.

[31] J. D. Atkinson, J. Yu, Role of dynamic strain-ageing in the environment assisted cracking observed in pressure vessel steels, Fatigue Fract. Eng Mater. Struct., 20 (1997) 1–12.

[32] K. Arioka, T. Yamada, T. Miyamoto, et al., Dependence of Stress Corrosion Cracking of Alloy 690 on Temperature, Cold Work, and Carbide Precipitation—Role of Diffusion of Vacancies at Crack Tips, Corros. 67

SC

(2011) 035006-1-035006-18.

[33] L.F. Zhang, K. Chen, D.H. Du, et al., Characterizing the effect of creep on stress corrosion cracking of cold worked alloy 690 in supercritical water environment, J Nucl. Mater. 492 (2017) 32–40.

M AN U

[34] A. Miyasaka, H. Hogawa: CORROSION/90, Paper no. 67, Houston, TX, NACE (1990). [35] H. Mughrabi, Cyclic slip irreversibilities and the evolution of fatigue damage. Metall. Mater. Trans. A 40 (2009) 1257–1279.

[36] M. Schütze, Deformation and cracking behavior of protective oxide scales on heat-resistant steels under tensile strain, Oxid. Met. 24 (1985) 199–232.

[37] M. Schütze, Protective Oxide Scales and Their Breakdown (Wiley, Chichester, 1997). [38] X.Y. Zhong, X.Q. Wu, E.H. Han, Effects of exposure temperature and time on corrosion behavior of a ferritic-martensitic steel P92 in aerated supercritical water, Corros. Sci, 90 (2015) 511–521.

TE D

[39] Y. Chen, K. Sridharan, T.R. Allen, et al., Microstructural examination of oxide layers formed on an oxide dispersion strengthened ferritic steel exposed to supercritical water, J Nucl. Mater. 359 (2006) 50–58. [40] J. Bischoff, A. T. Motta, R. J. Comstock, Evolution of the oxide structure of 9CrODS steel exposed to supercritical water, J. Nucl. Mater. 392 (2009) 272–279.

EP

[41] A.C.S. Sabioni, A.M. Huntz, F. Silva, et al., Diffusion of iron in Cr2O3: polycrystals and thin films, Mater. Sci. Eng. A 392 (2005) 254–261.

[42] J. Bischoff, A.T. Motta, C. Eichfeld, et al., Corrosion of ferritic–martensitic steels in steam and supercritical

AC C

water, J. Nucl. Mater. 441 (2013) 604–611. [43] J. Bischoff, A. T. Motta, EFTEM and EELS analysis of the oxide layer formed on HCM12A exposed to SCW, J. Nucl. Mater. 430 (2012) 171–180. [44] P. Ampornrat, G.S. Was, Oxidation of ferritic–martensitic alloys T91, HCM12A and HT-9 in supercritical water, J. Nucl. Mater. 371 (2007) 1–17. [45] S.E. Ziemniak, M.E. Jones, K.E.S. Combs, Magnetite solubility and phase stability in alkaline media at elevated temperatures, J. Solution Chem. 24 (1995) 837–877. [46] K. Sue, T. Adschiri, K. Arai, Predictive model for equilibrium constants of aqueous inorganic species at subcritical and supercritical conditions, Ind. Eng. Chem. Res. 41 (2002) 3298–3306. [47] L. Martinelli, F. Balbaud-Célérier, G. Picard, et al., Oxidation mechanism of a Fe–9Cr–1Mo steel by liquid Pb –Bi eutectic alloy (Part III), Corros. Sci. 50 (2008) 2549–2559. [48] Y. Fukai, Superabundant vacancies formed in metal-hydrogen alloys, Phys. Scripta. T103 (2003) 11–14. [49] Z.L. Zhu, H. Xu, D.F. Jiang, et al., The role of dissolved oxygen in supercritical water in the oxidation of 25

ACCEPTED MANUSCRIPT ferritic–martensitic steel, J Supercritical Fluids. 108 (2016) 56–60. [50] A. Atkinson, Transport processes during the growth of oxide films at elevated temperature, Rev. Mod. Phys. 57 (1985) 437–470. [51] S. Chevalier, F. Desserrey, J.P. Larpin, Oxygen transport during the high temperature oxidation of pure nickel, Oxid. Met. 64 (2005) 219–234. [52] D. Laverde, T. Gómez-Acebo, F. Castro, Continuous and cyclic oxidation of T91 ferritic steel under steam, Corros. Sci. 46 (2004) 613–631. implications for practical application, Corros. 70 (2014) 112–129.

RI PT

[53] J. Zurek, E. De Bruycker, S. Huysmans, et al., Steam oxidation of 9–12%Cr steels: critical evaluation and [54] C.T. Fujii, R.A. Meussner, Oxide structures produced on iron-chromium alloys by a dissociative mechanism, J. Electrochem. Soc. 110 (1963) 1195–1204.

[55] C.T. Fujii, R.A. Meussner, The mechanism of the high-temperature oxidation of iron-chromium alloys in

SC

water vapor, J. Electrochem. Soc. 111 (1964) 1215–1221.

[56] N. Birks, G.H. Meier, F.S. Pettit, Introduction to the High-Temperature Oxidation of Metals, 2nd ed., Cambridge University Press, New York, 2006. Corros. Sci. 48 (2006) 3123–3138.

M AN U

[57] L. Tan, Y. Yang, T.R. Allen, Oxidation behavior of iron-based alloy HCM12A exposed in supercritical water, [58] R.P. Gangloff, Hydrogen assisted cracking of high strength alloys, in: I. Milne, R.O. Ritchie, B. Karihaloo (Eds.) 6 (2003) 31–101.

[59] S. Lynch, Hydrogen embrittlement phenomena and mechanisms, Corros. Rev. 30 (2012) 105–123. [60]A. Nagao, C.D. Smith, M. Dadfarnia, et al., The role of hydrogen in hydrogen embrittlement fracture of lath

AC C

EP

TE D

martensitic steel, Acta Mater. 60 (2012) 5182–5189.

26