To examine the tritium exposure, four types of B₄C/Mo-based composite coatings on the substrate of stainless steel (15 mm×15 mm×2 mm) were prepared by vacuum plasma spraying (VPS). The spraying parameters were shown in Table 1. Mo coating was pure metal coating, its spraying parameters was different from the B₄C/Mo-

based composite coatings. All samples were rinsed in acetone, and then baked at 120℃ for 5 h, 200℃ for 2 h, 250℃ for 2 h, 300℃ for 4 h, 350℃ for 2 h and 400℃ for 4 h under high vacuum conditions (<3×10^-4 Pa). The vacuum pressure eventually reached an order of 10^-6 Pa.

Table 1

Table 1

Table 1 Vacuum plasma spraying parameters Parameters B4C BM5(95vol%B4C-5vol%Mo) BM15(85vol%B4C -15vol%Mo) Mo Current/A 620 600 Ar gas flow/slpm 37 40 H2 gas flow/slpm 13 12 Carrier gas flow/slpm 2.5 3 Powder feed rate/(r·min-1) 22 18 Spray distance/mm 220 180 Pressure/Pa 40000 50000

Table 1 Vacuum plasma spraying parameters

1.2.1 Exposure and BIXS measurement

All coating samples were exposed to tritium gas (tritium concentration 5% T) under given temperature (673 K), pressure (2.6 kPa) and time (4 h). The amount of tritium retained in the surface layers of the samples was determined by BIXS with argon as the working gas^{[ 18]}.

1.2.2 Imaging plate

The imaging plate (IP) technique, which was developed for the detection of low energy β-rays such as those generated by tritium, was performed to characterize the surface tritium activities of the four coating samples. The IP (BAS-TR2025, Fuji Photo Film Co. Ltd, Japan) was first placed in contact with each coating sample for 15 min in a dark shielded room, and then processed by an imaging plate reader (BAS-2500, Fuji Photo Film Co. Ltd, Japan) to obtain the digitized photon-stimulated-luminescence (PSL) intensity pattern as well as a two-dimensional mapping. The PSL intensity is a measure of the adsorbed energy in the IP, which cannot be directly converted to absolute tritium level but is nearly proportional to the tritium surface concentration. Details of tritium detection by IP technique are available elsewhere^{[ 29, 30, 31]}.

Figure 1 shows the distribution of tritium in the four types of coating samples. A higher tritium level is shown in red and a lower level in blue^{[ 31]}. As can be seen, the surface of the B₄C coating showed a red color, indicating substantial tritium adsorption. The surface of the Mo coating showed a blue color, demonstrating the smallest amount of adsorbed tritium comparing with the other coatings. For all the samples investigated, tritium adsorption followed the order of B₄C>BM15>BM5>Mo.

Fig. 1

	Figure Option ViewDownloadNew Window
	Fig. 1 IP image of the coating samples

Figure 2 shows the X-ray spectrum for the B₄C coating and associated changes peaks of the elements within the B₄C coating with time up to 1 h. Five X-ray peaks were observed in the spectrum, one of which corresponding to Ar(Kα) was sharp and intense, with the maximum intensity appeared at 2.99 keV. Cr(Kα), Fe(Kα) and Ni(Kα) peaks were also observed, which were common since the substrate was stainless steel. However, the Ti(Kα) peak was not expected. Also the X-ray energies for C and B were too small to be observed. The considerably broad, strong spectrum of bremsstrahlung X-rays indicated that the tritium adsorbed at the surface of B₄C coating was potentially diffused into the bulk material^{[ 18]}.

Fig. 2

	Figure Option ViewDownloadNew Window
	Fig. 2 X-ray spectrum of the B4C coating

Figure 3 shows the X-ray spectra of BM5, BM15 and Mo coatings. For these three spectra, only sharp peaks were observed with no clear bremsstrahlung X-ray peaks. This indicated that most tritium was probably only adsorbed on the surface of these coatings.

Fig. 3

	Figure Option ViewDownloadNew Window
	Fig. 3 X-ray spectra for coatings (a) BM5; (b) BM15; (c) Mo

Figure 4 shows the cross-sectional morphologies of the four types of coatings. As can be seen, there were a large number of pores in the B₄C coating, ranging from several to 30 μm in size. Comparing with the B₄C coating, the BM5 and BM15 coatings were much denser,since molybdenum occupied most of the big pores (Mo was indicated by the bright area, as shown in Fig.4(b) and (c)). Since the porosity of the Mo coating was the lowest (30%, 8.2%, 6.4% and 2.4% for B₄C, BM15, BM5 and Mo coatings, respectively), which also showed the smallest amount of tritium retention, it could be inferred that it must be the dense coating regardless of the physical and chemical properties that counts for the low adsorption of tritium.

Fig. 4

	Figure Option ViewDownloadNew Window
	Fig. 4 Cross-sectional morphologies of coatings (a) B4C; (b) BM5; (c) BM15; (d) Mo

Table 2 summarizes the intensities of Ar(Kα) and Ti(Kα) peaks in the spectra of the four coatings. The results showed that the net counting rates of Ar(Kα) for the BM5, BM15 and Mo coatings were much lower than that of the B₄C coating, which was in accordance with the results from IP images, indicating that the retaining amount of tritium on the surface of these three coatings were dramatically smaller than that of the B₄C coating. Interestingly, the intensities of Ar(Kα) peaks in the X-ray spectra were almost the same for both of the BM5 and BM15 coatings, while that of the Ti(Kα) peak for the BM5 coating was approximately 1/30 of that for the BM15 coating only. This demonstrates that the containing of Ti contamination is also one of the key factors that determine the absorption of tritium into the coatings.

Table 2

Table 2

Table 2 Net counting rate of four composite coatings Net counting rate / (counts·min-1) B4C BM5 BM15 Mo Ar(Kα) 1344.54 113.74 118.73 29.86 Ti(Kα) 23.27 3.62 109.47 31.63

Table 2 Net counting rate of four composite coatings

The B₄C coating has already been successfully applied as first wall materials, indicating that B₄C possesses many good properties. However, in the present study the B₄C coating showed the highest tritium retention, which should be attributed to the high porosity of the coating. It is known that tritium can be in contact with coatings in the form of gas. Thus, tritium can easily penetrate into the coating through pores and finally be adsorbed in them. Similarly, a study by Wu^{[ 18]} has demonstrated that the intensity of Ar(Kα) peak for the Pc-W coating with compact structure is only 1/4 of that for the VPS-W coating with inevitable voids or pores. The big pores, especially those above 10 μm in size can provide channels for tritium gas permeation and space for tritium storage. Besides, as for the case of BM5 and Mo coatings, the porosity of which is less than 5%, the nano-sized cracks are still inevitable due to the thermal spraying process, which may also provide rapid absorption passageway for tritium. However, since the speed of tritium absorption through nano-sized cracks is much lower, the net counting rate of Ar(Kα) for B₄C coating was substantially higher than those of the other coatings.

In addition to cracks and pores, other factors may also determine the amount of tritium adsorbed. For the coatings whose porosity is less than 5%, the content of Ti contamination should be another important factor that determines the absorption of tritium. Ti is very active to react with tritium:Ti + H₂ → TiH₂ (1)

For the coatings investigated, the intensities of main peaks (Kα and Kβ) of titanium were strong even when the Ti content was low, indicating that the solubility of tritium in Ti was relatively high.