Preparation of Mg and Al phosphate coatings on ferritic steel by wet-chemical method as tritium permeation barrier

Preparation of Mg and Al phosphate coatings on ferritic steel by wet-chemical method as tritium permeation barrier

Fusion Engineering and Design 85 (2010) 1090–1093 Contents lists available at ScienceDirect Fusion Engineering and Design journal homepage: www.else...

459KB Sizes 0 Downloads 68 Views

Fusion Engineering and Design 85 (2010) 1090–1093

Contents lists available at ScienceDirect

Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes

Preparation of Mg and Al phosphate coatings on ferritic steel by wet-chemical method as tritium permeation barrier Kun Zhang ∗ , Yuji Hatano Hydrogen Isotope Research Center, University of Toyama, 3190 Gofuku, Toyama, Toyama 930-8555, Japan

a r t i c l e

i n f o

Article history: Available online 27 March 2010 Keywords: Tritium permeation barrier Wet-chemical method Sol–gel method Electrolytic deposition Phosphate

a b s t r a c t Layers of Mg or Al phosphate were prepared on type 430 ferritic stainless steel (SS430) by wet-chemical methods as tritium permeation barrier. Disk-type specimens of SS430 were first coated with ZrO2 /ZrO2 or ZrO2 /Al2 O3 layers (100 nm) by sol–gel and electrolytic deposition techniques. Then, the phosphate layers were prepared by dip-coating method; the total thickness of coating was 200 nm. The permeation rate of hydrogen was measured at 300–600 ◦ C under driving pressure of 0.1 MPa. The Mg or Al phosphate layers provided significant barrier effects, and the permeation reduction factor observed under the present conditions was 200–3000 against that of Pd coated 430 specimen. Further investigation, however, was required to improve the stability at high temperatures. © 2010 Elsevier B.V. All rights reserved.

1. Introduction Development of permeation barrier coatings on structural materials is a feasible solution to limit uncontrolled tritium transport in fusion reactor. Ceramic coatings have been studied extensively due to economical advantages and small permeability of hydrogen isotopes [1–3]. Wet-chemical methods, such as sol–gel method, slurry coating and electrochemical deposition, are suitable to prepare ceramic coatings on structural metals, because of their capability for coating complicated geometries and low processing temperatures. Several researchers have examined barrier effects of coatings prepared by wet-chemical methods [4–12]. Terai et al. [4] and Nakamichi et al. [5,6] have prepared chemically densified Cr2 O3 –SiO2 coatings on type 316 stainless steel (SS316) and measured permeation rate of tritium [4] or deuterium [5,6]. They reported that this type of coating provided only limited barrier effects, and permeation reduction factor (PRF), which is defined as the ratio of permeation rate of uncoated specimen in steady state to coated one, was 10–100 at 600 ◦ C. Nevertheless, Nakamichi et al. [5,6] have found that addition of CrPO4 by impregnation treatment in acid solution containing CrPO4 and subsequent heat treatment led to significant improvement in barrier effects; PRF obtained for Cr2 O3 –SiO2 coating including CrPO4 on SS316 was 1000 under the similar conditions. These observations indicated that post-treatment with appropriate materials improves barrier effects of coatings derived by wet-chemical method. Kulsartov et

∗ Corresponding author. Tel.: +81 76 4456928; fax: +81 76 4456931. E-mail address: [email protected] (K. Zhang). 0920-3796/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.fusengdes.2010.02.012

al. [7] and Nakamichi et al. [8] have studied the effects of similar coating (Cr2 O3 –SiO2 + CrPO4 ) prepared on low-activation ferritic steel F82H. Although low-activation ferritic steels are primary candidates of structural materials of fusion blankets [13], the barrier effects obtained for F82H was smaller than that for SS316. The permeation reduction factor (PRF) for F82H was 400 [7] and 292 (in-pile test) [8] at 600 ◦ C and much smaller at lower temperatures. Chikada et al. [10] and Yao et al. [11] have used a sol–gel method (spin coating) to prepare Er2 O3 films on SS316, but they have not applied this technique to ferritic steels. Films prepared by wet-chemical methods are prone to be porous due to volatilization of solvent during drying processes and cracking caused by different thermal expansions between coating materials and substrate metals during subsequent heat treatments. In our previous study [12], an electrolytic deposition technique, so-called electrochemical deposition-pyrolysis (ECDP) [14,15], was applied to seal sol–gel derived ZrO2 coatings on a ferritic steel. Thin films of ZrO2 (50 nm) were prepared on type 430 stainless steel (SS430) by a conventional sol–gel method, and pores in this film were sealed with ZrO2 or Al2 O3 by ECDP technique. The significant reduction in permeation rate was observed after sealing process by ECPD technique, and PRF of 1000 was obtained at 600 ◦ C as the maximum value. The properties of coating, however, were not well reproducible, and further improvement was required. The purpose of the present study is to examine barrier effects of Mg and Al phosphates against permeation of hydrogen through ferritic steel. Because of their environment-friendliness, refractoriness, and chemical stability, these phosphates have been used as binders of refractory ceramics [16–19], sealants of plasma-sprayed oxide coatings [20,21], and anti-corrosion/oxidation coatings [18,22,23]. The permeability of hydrogen isotopes, however, has

K. Zhang, Y. Hatano / Fusion Engineering and Design 85 (2010) 1090–1093

1091

been scarcely examined. In this study, capping layers of Mg and Al phosphates were prepared by dip-coating on ZrO2 /ZrO2 or ZrO2 /Al2 O3 films formed on SS430 by sol–gel and ECDP techniques, and hydrogen permeation rate was measured at 300–600 ◦ C. The preparation of phosphate layer resulted in significant reduction in permeation rate, and PRF reached 3000 at 500 ◦ C. 2. Experimental Disk-type specimens of SS430 (20 mm diameter, 0.5 mm thickness) were polished with diamond powder and cleaned in ultrasonic bath with ethanol and de-ionized water. Thin films of ZrO2 (50 nm) were first prepared on the specimens with a conventional sol–gel method, and these ZrO2 films were sealed with ZrO2 or Al2 O3 by ECDP technique in the manner described elsewhere [12]. The thickness of coating was 100 nm after ECDP processes. The layer of ZrO2 prepared by sol–gel method is hereafter denoted as ZrO2 (SG), and that sealed with ZrO2 or Al2 O3 by ECDP technique is ZrO2 (ED)/ZrO2 (SG) or Al2 O3 (ED)/ZrO2 (SG). Then, the phosphate capping layers were prepared by dipcoating method. For preparing Mg phosphate layers, 2.5, 5 and 10 mass% Mg(H2 PO4 )2 aqueous solution was prepared. At first, a specimen was dip-coated with 2.5 mass% solution at extraction rate of 0.5 mm s−1 and dried at 200 ◦ C in air for 20 min. After heating at 500 ◦ C in air for 1 h, the specimen was processed with 5 mass% solution for two times and heat-treated at 400 ◦ C. At last, the specimen was processed with 10 mass% solution for one time and heat-treated at 300 ◦ C in air for 1 h. For preparing Al phosphate layer, simplified processes were adopted, i.e. only 5 mass% Al(H2 PO4 )3 aqueous solution and 500 ◦ C heat treatment were applied. Depth profiles of constituent elements of the coatings were measured by glow discharge optical emission spectroscopy (GD-OES). Cross-section morphology of the coatings was observed by a field emission scanning electron microscope (FE-SEM). Permeation rate of hydrogen through the specimen was measured with a vacuum device evacuated to 10−6 Pa by a turbomolecular pump and an oilsealed rotary pump. The device was separated into two chambers, i.e. up- and downstream chambers, by the specimen sealed with metal gaskets (U-Tightseal, Usui Kokusai Sangyo Kaisha, Japan). The pressure of hydrogen in the upstream chamber was kept at 0.1 MPa, and that of hydrogen permeating through the specimen into the downstream chamber was measured by a quadrupole mass analyzer and a capacitance manometer; the pressure in the downstream chamber was maintained to be sufficiently low in comparison with the upstream. Detection limit of permeation rate controlled by pressure of residual gases was 1 × 10−8 m−2 s−1 . Each permeation measurement was started at the lowest temperature (300, 400 or 500 ◦ C), and then the specimen temperature was raised in a step-by-step manner to 600 ◦ C.

Fig. 1. Cross-section of Mg phosphate/Al2 O3 (ED)/ZrO2 (SG) coating.

coatings, peak of Cr was observed at the interface between substrate and oxide coatings as shown in Fig. 2(b). This Cr peak indicated the formation of Cr oxide at the interface during the heat treatment in air. The profile of Cr in Fig. 2(a) has no peak and shows the absence of Cr oxide at the interface. These observations showed that Mg phosphate/ZrO2 (ED)/ZrO2 (SG) coating had better barrier effects against penetration of oxygen to the interface than ZrO2 (ED)/ZrO2 (SG) and Al2 O3 (ED)/ZrO2 (SG); porosity in Mg phosphate/ZrO2 (ED)/ZrO2 (SG) coating should be lower than that in other types of coating. Similar profiles were also observed for Al phosphate/ZrO2 (ED)/ZrO2 (SG) and Mg phosphate/Al2 O3 (ED)/ZrO2 (SG) coatings.

3. Results and discussion Fig. 1 shows micrograph of cross-section of the specimen with Mg phosphate layer on Al2 O3 (ED)/ZrO2 (SG) as a typical example. The specimens for the cross-sectional observation were prepared by a focused ion beam (FIB) technique, and a layer of tungsten (W) was deposited to avoid the sputtering of coating during the specimen preparation. The coating has uniform thickness of about 200 nm. No crack is observed at the interface between coating and substrate. Depth profiles of constituent elements in Mg phosphate/ZrO2 (ED)/ZrO2 (SG) coating are shown in Fig. 2(a). The coating had two-layer structure of phosphate layer and oxide layer even after the heat treatment in air. Before preparing phosphate layers on ZrO2 (ED)/ZrO2 (SG) and Al2 O3 (ED)/ZrO2 (SG)

Fig. 2. Depth profiles of constituent elements in (a) Mg phosphate/ZrO2 (ED)/ZrO2 (SG) and (b) ZrO2 (ED)/ZrO2 (SG) coatings measured with GD-OES, where Zr × 3 means that the plot of Zr was made after multiplying the values of emission intensity by 3 because original intensity was too small.

1092

K. Zhang, Y. Hatano / Fusion Engineering and Design 85 (2010) 1090–1093

Fig. 3. Permeation rates of hydrogen through phosphate coated SS430 specimens and reference specimen covered with Pd. For comparison, the data for specimens with ZrO2 (SG), ZrO2 (ED)/ZrO2 (SG) and Al2 O3 (ED)/ZrO2 (SG) coatings are also shown. Driving pressure of hydrogen was 0.1 MPa.

The hydrogen permeation rates through phosphate coated specimens in steady state are plotted in Fig. 3 together with the data for a reference SS430 specimen. Both surfaces of the reference specimen were covered with thin Pd layer (270 nm) to obtain stable and reproducible permeation rates. The results for the specimens with ZrO2 (SG), ZrO2 (ED)/ZrO2 (SG) and Al2 O3 (ED)/ZrO2 (SG) coatings obtained in the previous study [12] are also shown for comparison. The specimens coated with Mg and Al phosphates provided significantly small permeation rates in comparison with the reference specimen. The permeation rate through these specimens showed weak temperature dependence below 500 ◦ C and increased sensitively with temperature above 500 ◦ C. The solid line in this figure was obtained by least-square method with data at/below 500 ◦ C, and the dashed line with data at/above this temperature. From these lines and dotted line for reference SS430 specimen, PRF was evaluated to be 200, 1200 and 400 at 300, 500 and 600 ◦ C, respectively. The smallest value of permeation rate obtained at 500 ◦ C gave the maximum value of PRF, 3000. As described below, the slope of dashed line does not relate to the activation energy of permeation through the coating, because the degradation of coating was observed in the high temperature region. Nevertheless, no significant modification in coating was observed at/below 500 ◦ C. The slope of solid line therefore corresponds to the activation energy of permeation through coating, and it was evaluated to be 18 ± 7 kJ mol−1 . The activation energy of permeation through reference SS430 specimen was evaluated to be 47 ± 3 kJ mol−1 , and this value is comparable to the activation energy of hydrogen permeation through ferritic steels including SS430 given in Ref. [24]. The permeation rates through the phosphate coated specimens were comparable to those through the specimens with ZrO2 (ED)/ZrO2 (SG) and Al2 O3 (ED)/ZrO2 (SG) coatings at 600 ◦ C. Nevertheless, the reproducibility was much better. In addition, the permeation rates observed at other temperatures for phosphate coated specimens were clearly smaller than those for the specimens with ZrO2 (ED)/ZrO2 (SG) and Al2 O3 (ED)/ZrO2 (SG) coatings. The above-mentioned values of PRF were comparable with or larger than the values reported by Kulsartov et al. [7] and Nakamichi et al. [8] for F82H specimens with Cr2 O3 –SiO2 + CrPO4 coating; 400 and 292 at 600 ◦ C and smaller at lower temperatures. The thickness of their coating was 50–80 ␮m, while that of present coating was just 200 nm. It was therefore concluded that Mg and Al phosphates

Fig. 4. Depth profiles of constituent elements in Mg phosphate/ZrO2 (ED)/ZrO2 (SG) coating after permeation experiment: (a) upstream side exposed to 0.1 MPa H2 gas and (b) downstream. As Fig. 2, Zr × 3 means that the plot of Zr was made after multiplying the values of emission intensity by 3.

provide strong barrier effects against permeation of hydrogen isotopes through ferritic steel. The activation energy of hydrogen permeation through phosphate coated specimens, however, was significantly smaller than that through substrate as mentioned above. It appears that a small number of open pores were still present in the coating with phosphate layer, and diffusion through pores made dominant contribution to permeation through coating. As previously mentioned, the permeation rate through coated specimen sensitively increased with increasing temperature in high temperature region. In order to understand the mechanism underlying such increase in permeation rate, the specimen was kept at temperatures higher than 500 ◦ C for several hundreds of hours. The permeation rate increased with time gradually but continuously, and reached the value larger than initial one by 8 times. Hence, the steep increase in the permeation rate with increasing temperature observed above 500 ◦ C was ascribed to the degradation of coatings. Fig. 4 shows depth profiles of constituent elements in Mg phosphate/ZrO2 (ED)/ZrO2 (SG) coating after the long-term permeation experiment. The thickness of phosphate layer clearly decreased. It should be noted that the extent of thickness reduction is high in the upstream side exposed to 0.1 MPa H2 gas in comparison with the downstream side where the pressure of H2 gas was kept below 5 Pa during the permeation experiments. These observations indicate that hydrogen enhanced the degradation. In the present study, relatively high driving pressure (0.1 MPa) was employed because the permeability through the specimens coated by phosphates was low and accurate measurement was difficult at lower driving pressure. This driving pressure, however, is signifi-

K. Zhang, Y. Hatano / Fusion Engineering and Design 85 (2010) 1090–1093

cantly higher than expected partial pressures of hydrogen isotopes in fusion blankets. Hence, the permeation experiments at low driving pressure by using tritium (or deuterium) are under preparation to evaluate the durability under the conditions relevant to blankets. In addition, detailed study on the mechanism underlying the degradation is also required. Similar reduction in thickness was observed also for Al phosphate layers, but the extent of reduction appeared to be smaller than Mg phosphate. The durability of coating should be improved by optimization of compositions, thickness, and fabrication conditions. 4. Conclusions Magnesium and aluminum phosphates were applied to cap ZrO2 /ZrO2 or ZrO2 /Al2 O3 permeation barrier coatings prepared on type 430 ferritic steels. The permeation experiments at 300–600 ◦ C showed that Mg and Al phosphates thus prepared on ferritic steel act as strong barrier against permeation of hydrogen isotopes. The permeation reduction factor was evaluated to be 200–3000 under the present conditions. Further study, however, was required to improve the stability at high temperatures. Acknowledgements This study has been supported in part by Grant-in-Aid for Scientific Research on Priority Areas, 476, Tritium Science and Technology for Fusion Reactor from the Ministry of Education, Culture, Sports, Science and Technology. The authors thank Dr. Xianghua Kong of University of Science and Technology Beijing for fruitful discussion.

1093

References [1] G.W. Hollenberg, E.P. Simonen, G. Kalinin, A. Terlain, Fusion Eng. Des. 28 (1995) 190. [2] T. Terai, J. Nucl. Mater. 248 (1997) 153. [3] D.L. Smith, J.H. Park, I. Lyublinski, V. Evtikhin, A. Perujo, H. Glassbrenner, et al., Fusion Eng. Des. 61–62 (2002) 629. [4] T. Terai, T. Yoneoka, H. Tanaka, H. Kawamura, M. Nakamichi, K. Miyajima, J. Nucl. Mater. 212–215 (1994) 976. [5] M. Nakamichi, H. Kawamura, T. Teratani, J. Nucl. Sci. Technol. 38 (2001) 1007. [6] M. Nakamichi, H. Kawamura, T. Teratani, Fusion Sci. Technol. 41 (2002) 939. [7] T.V. Kulsartov, K. Hayashi, M. Nakamichi, S.E. Afanasyev, V.P. Shestakov, Y.V. Chikhray, et al., Fusion Eng. Des. 81 (2006) 705. [8] M. Nakamichi, T.V. Kulsartov, K. Hayashi, S.E. Afanasyev, V.P. Shestakov, Y.V. Chikhray, et al., Fusion Eng. Des. 82 (2007) 2246. [9] M. Nakamichi, H. Nakamura, K. Hayashi, I. Takagi, J. Nucl. Mater. 386–388 (2009) 692. [10] T. Chikada, A. Suzuki, Z. Yao, T. Kobayashi, D. Levchuk, H. Maier, et al., Proceedings of 2nd Japan–China Workshop on Blanket and Tritium Technology, Sendai, May 9–10, 2008, p. 96. [11] Z. Yao, A. Suzuki, D. Levchuk, T. Chikada, T. Tanaka, T. Muroga, et al., J. Nucl. Mater. 386–388 (2009) 700. [12] K. Zhang, Y. Hatano, J. Nucl. Mater. (submitted for publication). [13] R.J. Kurtz, A. Alamo, E. Lucon, Q. Huang, S. Jitsukawa, A. Kimura, et al., J. Nucl. Mater. 386–388 (2008) 411. [14] Y. He, F.H. Stott, Corros. Sci. 36 (1994) 1869. [15] X. Lu, R. Zhu, Y. He, Surf. Coat. Technol. 79 (1996) 19. [16] W.D. Kingery, J. Am. Ceram. Soc. 33 (1950) 239. [17] J.E. Hughes, M.F. Mosser, K.B. Eddinger, R.E. Myers, US Patent No. 6368394B1 (2002). [18] D.D.L. Chung, J. Mater. Sci. 38 (2003) 2785. [19] A.S. Wagh, S.Y. Jeong, US Patent No. 6776837B2 (2004). [20] E.M. Leivo, M.S. Vippola, P.P.A. Sorsa, P.M.J. Vuoristo, T.A. Mäntylä, J. Therm. Spray Technol. 6 (1997) 205. [21] S. Ahmaniemi, P. Vuoristo, T. Mäntylä, Surf. Coat. Technol. 151–152 (2002) 412–417. [22] G.W. Jernstedt, Bloomfield, J.C. Lum, US Patent No. 2331196 (1943). [23] S. Sambasivan, K.A. Steiner, K.K. Rangan, US Patent No. 7311944B2 (2007). [24] P. Jung, J. Nucl. Mater. 238 (1996) 189.