7.4 Reflection, Reemission, and Retention
121
amount of the retained flux gradually increases and becomes quite large after long
time injection. Because H injection and neutron irradiation introduce various defects
which work as H trapping sites. It is quite difficult to evaluate W(t) accurately after
the long time exposure. In principle, the integration of the flux balance gives the W(t)
as given by Eq. (7.3). However, the accuracy of the measurements of φ(t), only 2–3
digits, does not allow the reliable evaluation of W(t).
As indicated in the figure, the total retention during the injection includes two
components; “dynamic retention” and “static retention”. The former is directly
connected to the H recycling and the latter to the T inventory. The latter decreases
with increasing the wall temperature, because trapped T can be detrapped thermally.
As mentioned above, irradiation of energetic ions and neutrons produces various
defects such as interstitials and vacancies, their clusters or dislocation loops, bubbles
and so on, which work as additional trapping sites to increase the static retention or
in-vessel T inventory, especially for W [6].
Although the permeation flux is tiny compared to the injected flux, the permeated
tritium into the cooling water is a safety concern and free permeation is not allowed.
Hence, formation of permeation barrier is required and is an important R&D task [7].
As discussed in Chap. 4, some of injected H are directly reflected and the flux
ratio of the injected H and reflected one is defined as the reflection coefficient.
Generally, the reflection coefficient is higher for heavier metals and lower incident
energy. Reflected particles are either ions or neutrals, with less ions for lower energy
injection. Because reflected particles do not fully loose energy, they carry energy
resulting in energy reflection. Incident energy dependence of reflection coefficients
of H and energy for H injection to carbon (C) and tungsten (W) [8] are shown
in Figs. 4.5 and 4.6, respectively. W shows larger reflection coefficients for both
particles and energy than C. The figures indicate that more than half of both incident
particles and energy injected into W are reflected. This means that compared to C,
plasma-wall interaction could be less on W.
It should be noted that at ITER or reactor divertor, incident H flux is expected to
be 10
23 –10
25 particles·m
−2 ·s
−1 , which is 10
4 –10
6 times large than areal density of
atoms in the materials (around 10
18 –10
19 atoms·m
−2 ). This means that after 1 s of
injection, 10
4 –10
6 atomic layers, or, 10
3 –10
5 nm from the surface are fully occupied
or saturated with H. If the incident energy is 100 eV or less, surface region within
10
3 nm in depth is immediately saturated. Until now we do not have such high H
flux sources. There may be some new physics or chemistry. Actually, as shown in
Fig. 4.12 [9], chemical sputtering of C by energetic H significantly decreases with
H flux of over 10
22 m
−2 ·s
−1 for which reason is not clarified yet (see Chap. 4).
In permeation driven by energetic H injection, as discussed later, recombinational
release of incident H is often believed to be the rate-limiting process of the reemission. For extremely high flux discussing here, recombination must be spontaneous
because the surface H concentration immediately comes up to saturation and the saturation concentration must be very close to 1 in H atom to metal atom ratio (H/M).
For such densely hydrogen saturated surface, the reemission of injected H is not
necessarily due to molecular emission but atoms. Moreover, reflection coefficient at
the H saturated layers would be modified from that of no H including surface.
121
amount of the retained flux gradually increases and becomes quite large after long
time injection. Because H injection and neutron irradiation introduce various defects
which work as H trapping sites. It is quite difficult to evaluate W(t) accurately after
the long time exposure. In principle, the integration of the flux balance gives the W(t)
as given by Eq. (7.3). However, the accuracy of the measurements of φ(t), only 2–3
digits, does not allow the reliable evaluation of W(t).
As indicated in the figure, the total retention during the injection includes two
components; “dynamic retention” and “static retention”. The former is directly
connected to the H recycling and the latter to the T inventory. The latter decreases
with increasing the wall temperature, because trapped T can be detrapped thermally.
As mentioned above, irradiation of energetic ions and neutrons produces various
defects such as interstitials and vacancies, their clusters or dislocation loops, bubbles
and so on, which work as additional trapping sites to increase the static retention or
in-vessel T inventory, especially for W [6].
Although the permeation flux is tiny compared to the injected flux, the permeated
tritium into the cooling water is a safety concern and free permeation is not allowed.
Hence, formation of permeation barrier is required and is an important R&D task [7].
As discussed in Chap. 4, some of injected H are directly reflected and the flux
ratio of the injected H and reflected one is defined as the reflection coefficient.
Generally, the reflection coefficient is higher for heavier metals and lower incident
energy. Reflected particles are either ions or neutrals, with less ions for lower energy
injection. Because reflected particles do not fully loose energy, they carry energy
resulting in energy reflection. Incident energy dependence of reflection coefficients
of H and energy for H injection to carbon (C) and tungsten (W) [8] are shown
in Figs. 4.5 and 4.6, respectively. W shows larger reflection coefficients for both
particles and energy than C. The figures indicate that more than half of both incident
particles and energy injected into W are reflected. This means that compared to C,
plasma-wall interaction could be less on W.
It should be noted that at ITER or reactor divertor, incident H flux is expected to
be 10
23 –10
25 particles·m
−2 ·s
−1 , which is 10
4 –10
6 times large than areal density of
atoms in the materials (around 10
18 –10
19 atoms·m
−2 ). This means that after 1 s of
injection, 10
4 –10
6 atomic layers, or, 10
3 –10
5 nm from the surface are fully occupied
or saturated with H. If the incident energy is 100 eV or less, surface region within
10
3 nm in depth is immediately saturated. Until now we do not have such high H
flux sources. There may be some new physics or chemistry. Actually, as shown in
Fig. 4.12 [9], chemical sputtering of C by energetic H significantly decreases with
H flux of over 10
22 m
−2 ·s
−1 for which reason is not clarified yet (see Chap. 4).
In permeation driven by energetic H injection, as discussed later, recombinational
release of incident H is often believed to be the rate-limiting process of the reemission. For extremely high flux discussing here, recombination must be spontaneous
because the surface H concentration immediately comes up to saturation and the saturation concentration must be very close to 1 in H atom to metal atom ratio (H/M).
For such densely hydrogen saturated surface, the reemission of injected H is not
necessarily due to molecular emission but atoms. Moreover, reflection coefficient at
the H saturated layers would be modified from that of no H including surface.