178
9 Fuel Retention in a Rector with Full …
by laboratory experiments are quite consistent with recent data obtained for plasmafacing W-coated C tiles used in ASDEX-U [26]. One should note that the highest
total retention observed is about 2 × 10
22 m
−2 , which is still far less than the ITER
safety limit of 1.5 × 10
24 m
−2 (assuming 750 g is distributed homogeneously on
100 m
2 plasma-facing surface). Data extrapolation to higher fluences is required to
estimate T retention in ITER and DEMO. It is critically important to know whether
the retention saturates or not. Although simple extrapolation following the fluence
dependences of φ
0.3~1.0 would not give saturation, some data suggests saturation [28].
To make the correct estimation or extrapolation, we have to know depth profiles
in deeper regions, i.e. the micrometer range is not enough but more than cm depth
is required. As demonstrated in modeled depth profiles H implanted in W shown in
Fig. 9.8, the depth profile can be separated into three components: (1) the trapping
at the topmost surface, (2) formation of saturated layers in subsurface, and (3) bulk
retention in depth. The trapping at the topmost surface always exists but would not
exceed a value corresponding to a few times of surface density of W atoms, i.e. ~10
20
m
−2 . Significant amount of H is accumulated or trapped to make the H saturated
layers in the subsurface. Hydrogen concentration in the saturated layers would not
increase but their thickness could continuously increase with fluence, though the
fluence dependence of the growth rate of their thickness is not clear. The saturation
concentration decreases with temperature, while it increases with the H fugacity, or
with flux and incident energy of loaded H. Since both temperature and the fugacity
must have gradients from the surface to the bulk, the concentration in the saturated
layers should not be homogeneous. Thus, to describe retention phenomena correctly,
the concept of fixed trapping energies would not be suitable but multiple trapping
energies [22, 29], or thermodynamic approach using free energy, or entropy effect
are required.
According to the data so far reported, the saturation concentration in W around
400 K, ITER operation temperature, is 0.01–0.001 in H/W atomic ratio [25]. If
the thickness of the saturated layers with H/W = 0.01 was around 10 μm, the total
retention in the saturated layers would be ~6 × 10
21 m
−2 , still two orders of magnitude
less than the ITER safety limit. But longer exposure with higher flux in ITER could
easily increase both the H/W ratio and the thickness of the saturated layers.
The retention caused by deep penetration via diffusion is also not small. Assuming
that the surface concentration of the dissolved H is around 100 appm, the total H
dissolved in the bulk of 10 mm in depth would be ~10
22 m
−2 , which is nearly the
same as that in the above-mentioned saturated layers. Somehow, H dissolved in W
can be released after the exposure and isotopic replacement would also reduce T
retention in near-surface region. However, the deep penetration always accompanies
trapping and only tiny amount of H dissolved in W could be released. Hydrogen
isotopes can reach to the back surface (permeation) which is a concern for tritium
safety.
How does such H uptake modify the structure of near-surface layer of W? The
answer is given in Sect. 6.3.3.2 in Chap. 6. Nevertheless, it should be noted here
again, referring to Fig. 6.11 [27]. Here, H trapped or adsorbed at the topmost surface
was not considered. To form the H saturated subsurface layers, firstly all grains at
9 Fuel Retention in a Rector with Full …
by laboratory experiments are quite consistent with recent data obtained for plasmafacing W-coated C tiles used in ASDEX-U [26]. One should note that the highest
total retention observed is about 2 × 10
22 m
−2 , which is still far less than the ITER
safety limit of 1.5 × 10
24 m
−2 (assuming 750 g is distributed homogeneously on
100 m
2 plasma-facing surface). Data extrapolation to higher fluences is required to
estimate T retention in ITER and DEMO. It is critically important to know whether
the retention saturates or not. Although simple extrapolation following the fluence
dependences of φ
0.3~1.0 would not give saturation, some data suggests saturation [28].
To make the correct estimation or extrapolation, we have to know depth profiles
in deeper regions, i.e. the micrometer range is not enough but more than cm depth
is required. As demonstrated in modeled depth profiles H implanted in W shown in
Fig. 9.8, the depth profile can be separated into three components: (1) the trapping
at the topmost surface, (2) formation of saturated layers in subsurface, and (3) bulk
retention in depth. The trapping at the topmost surface always exists but would not
exceed a value corresponding to a few times of surface density of W atoms, i.e. ~10
20
m
−2 . Significant amount of H is accumulated or trapped to make the H saturated
layers in the subsurface. Hydrogen concentration in the saturated layers would not
increase but their thickness could continuously increase with fluence, though the
fluence dependence of the growth rate of their thickness is not clear. The saturation
concentration decreases with temperature, while it increases with the H fugacity, or
with flux and incident energy of loaded H. Since both temperature and the fugacity
must have gradients from the surface to the bulk, the concentration in the saturated
layers should not be homogeneous. Thus, to describe retention phenomena correctly,
the concept of fixed trapping energies would not be suitable but multiple trapping
energies [22, 29], or thermodynamic approach using free energy, or entropy effect
are required.
According to the data so far reported, the saturation concentration in W around
400 K, ITER operation temperature, is 0.01–0.001 in H/W atomic ratio [25]. If
the thickness of the saturated layers with H/W = 0.01 was around 10 μm, the total
retention in the saturated layers would be ~6 × 10
21 m
−2 , still two orders of magnitude
less than the ITER safety limit. But longer exposure with higher flux in ITER could
easily increase both the H/W ratio and the thickness of the saturated layers.
The retention caused by deep penetration via diffusion is also not small. Assuming
that the surface concentration of the dissolved H is around 100 appm, the total H
dissolved in the bulk of 10 mm in depth would be ~10
22 m
−2 , which is nearly the
same as that in the above-mentioned saturated layers. Somehow, H dissolved in W
can be released after the exposure and isotopic replacement would also reduce T
retention in near-surface region. However, the deep penetration always accompanies
trapping and only tiny amount of H dissolved in W could be released. Hydrogen
isotopes can reach to the back surface (permeation) which is a concern for tritium
safety.
How does such H uptake modify the structure of near-surface layer of W? The
answer is given in Sect. 6.3.3.2 in Chap. 6. Nevertheless, it should be noted here
again, referring to Fig. 6.11 [27]. Here, H trapped or adsorbed at the topmost surface
was not considered. To form the H saturated subsurface layers, firstly all grains at