Home Diffusion Kinetics of Chromium in a Novel Super304H Stainless Steel
Article Open Access

Diffusion Kinetics of Chromium in a Novel Super304H Stainless Steel

  • Shao Bo Ping , Fei Xie , Rui Kun Wang , Zhi Jun Zheng and Yan Gao EMAIL logo
Published/Copyright: April 21, 2016

Abstract

Diffusion behavior and diffusion kinetics of chromium (Cr) from the pure chromium coating in the matrix of Super304H and TP304H steels were investigated in a temperature range of 600–900 ℃ by using energy-dispersive spectrum (EDS) quantitative line scanning and Fick’s second law of diffusion. Pure Cr coating was obtained by electroplating. Experimental results show that the diffusion depth and diffusivity of Cr increase gradually with the increase in temperature. The diffusion activation energy of Cr is found to be lower and the diffusion coefficient of Cr to be higher in the Super304H steel than those in the TP304H steel at the same temperature. Furthermore, the diffusion equation of Cr in the Super304H and TP304H steels is DCr/Super304H=1.08×1015exp(6.08×104/RT) and DCr/TP304H=2.29×1015exp(7.44×104/RT), respectively.

Introduction

A novel fine-grained austenitic heat-resistant stainless steel, Super304H, developed by Sumitomo Metal Industries of Japan in the late 1980s has become one of the preferred materials for superheater and reheater tubes in ultra-supercritical (USC) power plants due to its excellent high-temperature strength, high-temperature oxidation resistance and anti-steam corrosion performance [16]. The chemical composition design of the Super304H steel was based on the traditional TP304H steel by increasing carbon content and adding copper, niobium, nitrogen and boron elements [1, 2]. Due to the higher carbon content in Super304H steel, however, more M23C6 (M = Fe, Cr) is prone to precipitate along grain boundaries and introduce more chromium-depleted zones, which will increase greatly its intergranular corrosion susceptibility (IGCS) when servicing at temperature of 600 –650 °C [79]. In fact, leakage accidents of Super304H superheater and reheater tubes in USC boilers, which have been widely promoted in China, occurred seriously at the stage of water pressure testing [8].

According to the theory of chromium depletion, precipitation of M23C6 phase is the main reason for the IGCS of 18–8 austenitic stainless steel. Furthermore, the fundamental reason of chromium depletion at grain boundaries lies in the low diffusion rate of Cr in austenite, which cannot diffuse quickly enough from the austenite matrix to compensate the chromium-depleted zones caused by carbide precipitation. The diffusion coefficient of carbon in austenite is about 10−14 m2/s at 650 ℃, and that of carbide forming elements (chromium, tungsten, molybdenum and vanadium) in austenite is about 10−20 m2/s [10]. Thorvaldsson [11] used STEM/energy dispersive spectrum (EDS) methods to calculate the diffusion coefficient of Cr in 304 steel at 600–700 ℃ and found it to be about 10−21 m2/s. Since the diffusion coefficient of chromium is six orders of magnitude lower than that of carbon, the precipitation of Cr23C6 phase along grain boundaries will inevitably lead to chromium-depleted zone near the carbides. The diffusion kinetics of chromium is therefore a key factor in affecting the chromium-depleted zones near the carbides. Much attention has been paid to the diffusion kinetics of chromium in some high-performance alloys. Paul [12] calculated the diffusion coefficient of Cr in super Incoloy-800 alloy. Chen and Iijima [13, 14] calculated the diffusion coefficient of Cr in Ni-based alloy by 51Cr tracer method. Brieseck [15] calculated the diffusion coefficient of Cr in WC alloy by using Wavelength Dispersive Spectroscopy-Electron Probe MicroAnalyzer (WDS-EPMA) quantitative line scanning technology. Sabioni [1620] focused on the Cr diffusion kinetics in the oxide film on stainless steel and revealed the differences of Cr diffusion between the oxide film and austenite matrix.

As a novel heat-resistant material with fine grains, the diffusion kinetics of Cr in Super304H steel has not been touched yet. Compared with the traditional TP304H steel (grain grade 5–6), Super304H has more alloying elements and refined grains (grain grade 7–8). The discrepancy in composition and grain size is bound to give rise to different diffusion kinetics of Cr in Super304H and TP304H steels. Grain refinement not only provides more diffusion channels for Cr atoms but also increases lattice distortion at grain boundaries, which may affect the activation energy of Cr diffusion. Wang [21] comparatively analyzed the diffusion coefficient of Cr in coarse-grained Fe and nanocrystalline Fe by secondary ion mass spectrometry (SIMS) and revealed that grain refinement to nanometer range significantly enhanced the diffusion coefficient of Fe by 5–6 times. Up to now, Cr diffusion kinetics in 304 austenitic stainless steel has been studied quite a lot [1011, 18], but no attempt has been made on the diffusion kinetics of Cr in the fine-grained Super304H steel.

Therefore in this paper, the diffusion kinetics of Cr in Super304H steel was taken as the research object, with traditional TP304H steel as the reference base. The diffusion depth and diffusion coefficient of Cr element from the electroplated Cr coating into the matrix of these two kinds of steels were detected by EDS quantitative line scanning. What is more, the diffusion coefficient equations of Cr were established and the differences of Cr diffusion behavior in the two steels were analyzed. The results will provide fundamental data for comprehensively interpreting and understanding the evolution of intergranular corrosion sensitization caused by M23C6 precipitation and desensitization process (healing of chromium-depleted zones) in Super304H steel.

Materials and methods

Materials and preparation of Cr coating

Test materials used in this paper were Super304H steel tubes with specifications of Ф 50 × 3.5 mm (diameter × wall thickness) and TP304H steel rods with a diameter of Ф 12 as reference base. Chemical compositions of the as-received specimens were measured by Vacuum Spark Emission Spectrometer (Foundry-Master, WAS, Germany) and nitrogen element was especially measured by photoelectric emission spectrometer. The results are shown in Table 1, which conforms to ASME specifications. Super304H steel tubes were solution treated at 1,150 ℃ for 30 min and TP304H steel rods at 1,050 ℃ for 30 min, with water quenching.

Table 1:

Chemical compositions of Super304H and TP304H steels (mass %).

Materials and StandardChemical Compositions
CSiMnCrNiSPCuNbN
Super304H0.0950.250.6318.008.71<0.0050.0323.140.450.10
ASTM Super304H0.07–0.13≤0.03≤1.0017.00–19.007.50–10.50≤0.010≤0.0402.50–3.500.20–0.600.05–0.12
TP304H0.0680.361.4918.168.070.0170.035
ASTM TP304H0.04–0.10≤0.75≤2.0018.00–20.008.00–11.00≤0.030≤0.040

After solution treatment, pure chromium coatings with thicknesses of 18 and 45 mm, respectively, were electroplated on the surface of Super304H and TP304H steels in aqueous solution of 250 g/L CrO3 anhydride, 2.5 g/L H2SO4 and 1–4 g/L NaF additive. Electroplating process parameters were as follows: 55 ℃, current density 27 A/dm2 and plating time 2 h.

Preparation of chromium diffusion couples

Two kinds of steel samples with Cr coating were cut into size of 10 × 10 mm. The diffusion couples were made with Cr coating as contacting surface and bundled with molybdenum wire. The couples were then put into quartz tubes that were sealed with internal vacuum and filled with 30 kPa pure Ar gas and then isothermally diffusion annealed at 600–900 ℃ (with 50 ℃ as temperature interval) for 50 h and air cooled.

Measurement of concentration profiles

Concentration profiles of Cr on the cross section of Cr diffusion couples were measured by backscattered electron (BSE) imaging and EDS quantitative line scanning using S-3400N Hitachi scanning electron microscope (SEM). EDS quantitative line scanning spectra were measured at an accelerating voltage of 20 kV, a beam current of ~10 nA and a step width of 0.3 μm, with a measuring time of 5 s at each point.

Results and discussions

SEM observation of original Cr electroplating

Figure 1 exhibits the BSE micrographs of the cross section of original Cr electroplating on Super304H and TP304H steels. The thickness of Cr layer is about 18 and 45 μm, respectively, on Super304H and TP304H steels. The bonding interface is flat with no defects in both samples. In addition, there are obviously Nb-rich precipitates in the matrix of Super304H and no Nb-rich precipitates in TP304H.

Figure 1: Cross-sectional BSE micrographs of original Cr electroplating on Super304H (a) and TP304H (b) steels.
Figure 1:

Cross-sectional BSE micrographs of original Cr electroplating on Super304H (a) and TP304H (b) steels.

Determination of diffusion coefficient model

In the present work, the adopted diffusion couple model corresponding to the diffusion profile is schematically shown in Figure 2 [22]. The components of both matrix and plated Cr layer are not affected by diffusion conditions and the corresponding diffusion coefficient may be calculated by Fick’s second law of diffusion as follows:

(1)C/t=D2C/x2

with boundary conditions:

whent=0{x>0,C=C0x<0,C=C1

and

whent0{x=,C=C0x=,C=C1

where C0 is the Cr concentration of matrix in the samples, C1 is the Cr concentration of Cr electroplating and CS represents the Cr concentration at the origin position of bonding interface in the diffusion couples. Application of these conditions to eq. (1) yields a solution:

(2)C(x,t)=(C0+C1)/2+erf(Z)(C0C1)/2

with

(3)Z=x/(2Dt)

where C(x,t) is the Cr concentration at the depth x away from the bonding interface after diffusion annealing time t. The expression erf(Z) is the Gaussian error function. The diffusion coefficient D is determined by the fitting eq. (4) as follows, where k is the slope of the fitting lines of Z versus x based on eq. (3):

(4)D=1/(4k2t)
Figure 2: Schematic diagram of component–distance relation in the diffusion couple model [22].
Figure 2:

Schematic diagram of component–distance relation in the diffusion couple model [22].

Detailed characterization of diffusion couples

Quantitative line scan spectrum and the corresponding Cr diffusion profiles (in at. %) perpendicular to the bonding interface of the Super304H diffusion couples after vacuum heat treatment at different temperatures for 50 h are shown in Figure 3(a)–(g). It is obvious that the Cr diffusion depth increases gradually with the rising of temperature. A different gradient diffusion layer of Cr is observed at annealing temperature of 900 ℃ compared to the straight diffusion layer of Cr when the annealing temperature is between 600 and 850 ℃.

Figure 3: Left: Quantitative line scan spectrum across the bonding interface of Super304H diffusion couples after vacuum heat treatment at different temperatures for 50 h; Right: Corresponding Cr diffusion profiles (in at. %) across the bonding interface (a) 600 ℃; (b) 650 ℃; (c) 700 ℃; (d) 750 ℃; (e) 800 ℃; (f) 850 ℃; and (g) 900 ℃.
Figure 3:

Left: Quantitative line scan spectrum across the bonding interface of Super304H diffusion couples after vacuum heat treatment at different temperatures for 50 h; Right: Corresponding Cr diffusion profiles (in at. %) across the bonding interface (a) 600 ℃; (b) 650 ℃; (c) 700 ℃; (d) 750 ℃; (e) 800 ℃; (f) 850 ℃; and (g) 900 ℃.

Figure 4(a)–(g) exhibits the quantitative line scan spectrum and the corresponding Cr diffusion profiles (in at. %) perpendicular to the bonding interface of the TP304H diffusion couples after vacuum heat treatment at different temperatures for 50 h. Similar to the Cr diffusion couples on Super304H, the Cr diffusion depth of the diffusion couples on TP304H increases gradually with the rising of temperature and a gradient diffusion layer of Cr is also observed at annealing temperature of 900 ℃. The reason for this abnormal Cr diffusion gradient structure at 900 ℃ is not known at the moment.

Figure 4: Left: Quantitative line scan spectrum across the bonding interface of TP304H diffusion couples after vacuum heat treatment at different temperatures for 50 h; Right: Corresponding Cr diffusion profiles (in at. %) across the bonding interface (a) 600 ℃; (b) 650 ℃; (c) 700 ℃; (d) 750 ℃; (e) 800 ℃; (f) 850 ℃; and (g) 900 ℃.
Figure 4:

Left: Quantitative line scan spectrum across the bonding interface of TP304H diffusion couples after vacuum heat treatment at different temperatures for 50 h; Right: Corresponding Cr diffusion profiles (in at. %) across the bonding interface (a) 600 ℃; (b) 650 ℃; (c) 700 ℃; (d) 750 ℃; (e) 800 ℃; (f) 850 ℃; and (g) 900 ℃.

Calculation of diffusion coefficient and activation energy

According to the results of Figures 3 and 4, nine sets of data of Cr concentration C(x,t) near the boding interface, including oS, four data along the diffusion direction into the matrix and four data against the diffusion direction into the Cr layer, were selected to accurately fit and calculate the Cr diffusion coefficient.

Figure 5: Linear fitting lines of Z versus x of the diffusion couples of Super304H (a) and TP304H (b) steels.
Figure 5:

Linear fitting lines of Z versus x of the diffusion couples of Super304H (a) and TP304H (b) steels.

The linear fitting lines of Z versus x based on the selected nine sets of data of the diffusion couples of Super304H and TP304H are drawn in Figure 5(a) and (b), respectively. The data of slope k of the fitting linear lines are listed in Table 2. The Cr diffusion coefficients determined by eq. (4) through the data of slope k in Table 2 are listed in Table 3.

Table 2:

Fitting slope k of the linear lines of Z versus x at different temperatures in Super304H and TP304H steels.

Temperature/℃600650700750800850900
Slope kSuper304H2.230841.924331.509201.356081.164910.913640.74962
TP304H3.757583.325742.357562.183241.860831.306960.96918
Table 3:

Diffusion coefficient of Cr at different temperatures in Super304H and TP304H steels.

Temperature/℃600650700750800850900
D × 10−19/(m2/s)Super304H2.793.756.107.5510.2316.6424.72
TP304H0.981.262.502.914.018.1314.79

The results of diffusion coefficient of chromium in Table 3 show that the diffusivity of chromium in the Super304H steel is obviously 2–3 times larger than that in the TP304H steel at the same temperature. The reason may be attributed to the finer grains of Super304H (grain grade of 7–8) than the TP304H (grain grade of 5–6) steels [13], which provide more grain boundaries as diffusion channels and facilitates more rapid Cr diffusion along grain boundaries. The aim of the fine grain design of Super304H by high-temperature softening (around 1,250 ℃) and relatively low-temperature solution treatment (around 1,150 ℃) is to improve the oxidation resistance by Cr diffusion enhancement at grain boundaries [13] and has been proved soundly by the calculation of Cr diffusion coefficient D in the present study.

Moreover, the linear relationship between ln D and 1/T of Super304H and TP304H steels based on the data in Table 3 is exhibited in Figure 6(a) and (b), respectively. The preexponential factor D0 and activation energy Q are calculated according to the Arrhenius relationship of eq. (5), where R is the gas constant:

(5)D=D0exp(Q/RT)

The diffusion activation energies Q of chromium in Super304H and TP304H stainless steels in the temperature range of 600–900 ℃ are 60.8 and 74.4 kJ/mol, respectively, in the present investigation. From published literatures, little attention has been paid to the diffusion activation energy of Cr in the novel Super304H steel. However, an empirical equation for estimating the activation energy of chromium in the 304 stainless steel was reported in the literature [22], where the self-diffusion of chromium in the 304 stainless steel was investigated using 51Cr tracer diffusion method and the diffusion activation energy of chromium was 67.5 kJ/mol in the temperature range of 943–1,038 ℃. Considering the temperature difference, the result in Ref. [22] is in agreement with our result of diffusion activation energy of chromium in 304 stainless steel.

The Cr diffusion coefficient equations for the two steels in the temperature range of 600–900 ℃ are given as DCr/Super304H=1.08×1015exp(6.08×104/RT) and DCr/TP304H=2.29×1015exp(7.44×104/RT), respectively.

Figure 6: The linear relationship between ln D and 1/T of the Super304H (a) and TP304H (b) steels.
Figure 6:

The linear relationship between ln D and 1/T of the Super304H (a) and TP304H (b) steels.

In the crystals, the diffusion of atoms depends mainly on the thermal vibration and migration of atoms at the equilibrium position. It is known that chromium diffusion in stainless steels is dominated by the exchange and vacancy mechanism [23]. The vacancy formation energy is very important for atom diffusing besides the atom migration energy for the atom jumping from one balance position to another in the process of diffusion. From the above-calculated results, the diffusion activation energy of Cr in Super304H steel is lower than that in TP304H steel. The main reason may be that there are more vacancies near the grain boundaries in the Super304H with finer grains, which give rise to relatively lower diffusion activation energy of chromium in the Super304H steel.

In the present study, the fine grain structure of the Super304H steel has been proved to give rise to lower diffusion activation energy and higher diffusion coefficient of chromium compared with coarse-grained TP304H steel, which will no doubt improve the oxidation resistance of the Super304H as primarily intended. Meanwhile, the Cr diffusion enhancement at grain boundaries in Super304H will definitely affect the precipitation rate of Cr23C6, the chromium-depleted zones and the healing of chromium-depleted zones at grain boundaries, that is, the evolution process of intergranular corrosion sensitization at servicing temperature of 600–650 ℃ of the material. More attention should be paid to these issues, which may help better interpreting and understanding the high IGCS of the novel Super304H steel and find the solution to solve the problem.

Conclusions

  1. The diffusion depth and diffusion coefficient of Cr in Super304H and TP304H steels increase gradually with the rising of temperature and there is an abnormal gradient diffusion layer of Cr at annealing temperature of 900 .

  2. Higher diffusion coefficient and lower diffusion activation energy of chromium are found in the Super304H steel compared to the TP304H steel, which are attributed mainly to the fine grain structure of the Super304H steel. The more vacancies near grain boundaries in the fine-grained Super304H steel give rise to its lower diffusion activation energy of Cr. The fine-grained structure of the Super304H steel provides more grain boundaries as rapid diffusion channels for Cr and leads to its higher Cr diffusion coefficient.

  3. The diffusion coefficient equation of Cr in the Super304H and TP304H steels is DCr/Super304H=1.08×1015exp(6.08×104/RT) and DCr/TP304H=2.29×1015exp(7.44×104/RT), respectively.

Funding statement: Funding: The authors acknowledge the financial support from the National Nature Science Foundation of China (51471072) and Key Laboratory of Advanced Energy Storage Materials of Guangdong Province.

References

[1] Y. Sawaragi and S. Hirano, New Alloys for Pressure Vessels and Piping, ASME Press, New York (1990), pp. 141–146.Search in Google Scholar

[2] Y. Sawaragi, N. Otsuka and H. Senba, Sumitomo Search, 56 (1994) 34–43.Search in Google Scholar

[3] T. Kan, Y. Sawaragi and Y. Yamadera, Mater. Adv. Power Eng., 4 (1998) 441–450.Search in Google Scholar

[4] D.V. Thornton, K.H. Meyer, R. Viswanathan and J.W. Nutting, eds., European high temperature materials development in advanced heat resistant steels for power generation [C], IOM Communications Ltd., London (1999), pp. 349–365.Search in Google Scholar

[5] T. Sourmail, Mater. Sci. Technol., 17 (2001) 1–14.10.1179/026708301101508972Search in Google Scholar

[6] A. Iseda, H. Okada, H. Semba and M. Lgarashi, Energ. Mater. Mater. Sci. Eng. Energ. Syst., 2 (2007) 199–206.10.1179/174892408X382860Search in Google Scholar

[7] J. Swaminathan, R. Singh, M.K. Gunjan and B. Mahato, Eng. Fail. Anal., 18 (2011) 2211–2221.10.1016/j.engfailanal.2011.07.015Search in Google Scholar

[8] Y. Gao, C.L. Zhang, X.H. Xiong, Z.J. Zheng and M. Zhu, Eng. Fail. Anal., 24 (2012) 26–32.10.1016/j.engfailanal.2012.03.004Search in Google Scholar

[9] C.L. Zhang, X.H. Xiong, S.B. Ping and Y. Gao, Corros. Eng., Sci. Technol., 49 (2014) 624–632.10.1179/1743278214Y.0000000164Search in Google Scholar

[10] S.H. Zhang, The Alloy Steel, Metallurgical Industry Press, Beijing (1981).Search in Google Scholar

[11] T. Thorvaldsson and A. Salwen, Scr. Metall., 18 (1984) 739–742.10.1016/0036-9748(84)90331-4Search in Google Scholar

[12] A.R. Paul, K.N.G. Kaimal, M.C. Naik and S.R. Dharwadkar, J. Nucl. Mater., 217 (1994) 75–81.10.1016/0022-3115(94)90306-9Search in Google Scholar

[13] T.F. Chen, G.P. Tiwari, Y. Iijima and K. Yamauchi, Mater. Trans., 44 (2003) 40–46.10.2320/matertrans.44.40Search in Google Scholar

[14] Y. Iijima, H. Nitta, R. Nakamura, K. Takasewa, A. Inoue and S. Takemoto, J. Jpn. I. Met., 69 (2005) 321–331.10.2320/jinstmet.69.321Search in Google Scholar

[15] M. Brieseck, M. Bohn and W. Lengauer, J. Alloy. Compd., 489 (2010) 408–414.10.1016/j.jallcom.2009.09.137Search in Google Scholar

[16] A.C.S. Sabioni, A.M. Huntz, A.M.J.M. Daniel and W.A.A. Macedo, Philos. Magaz., 85 (2005) 3643–3658.10.1080/14786430500323795Search in Google Scholar

[17] A.C.S. Sabioni, A.M. Huntz, M.F. Salgado, A. Pardini, E.H. Rossi, R.M. Paniago and V. Ji, Mater. High Temp., 27 (2010) 89–96.10.3184/096034010X12710805379897Search in Google Scholar

[18] A.C.S. Sabioni, R.P.B. Ramos, V. Ji, F. Jomard, W.A.A. Macedo, P.L. Gastelois and V.B. Trindade, Oxid. Met., 78 (2012) 211–220.10.1007/s11085-012-9301-ySearch in Google Scholar

[19] A.C.S. Sabioni, E.A. Malheiros, V. Ji, F. Jomard, W.A.A. Macedo and P.L. Gastelois, Oxid. Met., 81 (2014) 407–419.10.1007/s11085-013-9451-6Search in Google Scholar

[20] A.C.S. Sabioni, J.N.V. Souza, V. Ji, F. Jomard, V.B. Trindade and J.F. Carneiro, Solid State Ion., 276 (2015) 1–8.10.1016/j.ssi.2015.03.027Search in Google Scholar

[21] Z.B. Wang, N.R. Tao, W.P. Tong, J. Lu and K. Lu, Acta Mater., 51 (2003) 4319–4329.10.1016/S1359-6454(03)00260-XSearch in Google Scholar

[22] H.S. Daruvala and K.R. Bube, Mat. Sci. Eng., 41 (1979) 293–295.10.1016/0025-5416(79)90151-4Search in Google Scholar

[23] G.X. Hu, X. Cai and Y.H. Rong, Fundamentals of Materials Science, 3rd edition, Shanghai Jiao Tong University Press, Shanghai (2010).Search in Google Scholar

Received: 2015-10-16
Accepted: 2016-2-20
Published Online: 2016-4-21
Published in Print: 2017-2-1

©2017 by De Gruyter

This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Downloaded on 21.10.2025 from https://www.degruyterbrill.com/document/doi/10.1515/htmp-2015-0227/html?lang=en&srsltid=AfmBOorsVgJ5ivoxMYbxCVMnG14aZ0lkWf5hazCs5DjrjGj5-aLhkpKV
Scroll to top button