DEUTERIUM RETENTION IN POLYCRYSTALLINE TUNGSTEN · 2013-10-18 · Deuterium Retention in Polycrystalline Tungsten Zhe Tian Master of Applied Science Graduate Department of Aerospace
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DEUTERIUM RETENTION IN POLYCRYSTALLINE TUNGSTEN
By
Zhe Tian
A thesis submitted in conformity with the requirements for the degree of
5.1 Fluence dependence of D retention in PCW…………………………. 21
5.1.1 D retention in Rembar PCW at 300 K……………………..……. 21
5.1.2 D retention in Plansee PCW at 300 K……………..…………….. 23
5.1.3 D retention in Plansee PCW at 500 K…………..……………….. 24
5.1.4 Discussion of fluence dependence………………………………... 25
5.2 Temperature dependence of D retention in PCW……....………….... 27
5.3 Ion energy dependence of D retention in PCW…………..…..…....… 28
6. Conclusions…………………………………….………………. 29
6.1 Fluence dependence………………………………….…………...…… 29
6.2 Temperature dependence……………………….…………………...... 30
6.3 Ion energy dependence………………………….………………......… 30
References………………………………………….…...……..… 32
Figures……………………………………………………...…... 36
vi
List of figures: Figure 2-1: Depth profiles of D trapped as D atoms (a) and D2 molecules (b) in single crystal
and hot-rolled W implanted with 6 keV D ions at 300 K determined by the SIMS/RGA method.
Figure 2-2: Retained vs. cumulative-fluence for 1 keV/D+ implantations at 500 K. Data are shown for specimens W2 (1023 D/m2 probe-fluence only), W1 (9×1023 D/m2 damage-fluence and probe-fluence), and W3 (1025 D/m2 damage-fluence and probe-fluence).
Figure 2-3: Retained vs. cumulative-fluence for 500 eV/D+ implantations at 500 K. Data are shown for specimens W5 (1023 D/m2 probe-fluence only), W4 (9×1023 D/m2 damage-fluence and probe-fluence), W6 (3×1024 D/m2 damage-fluence and probe-fluence), W7 (1025 D/m2 damage-fluence and probe-fluence), and W9 (3×1025 D/m2 damage-fluence and probe-fluence).
Figure 2-4: Fluence dependence of D retention in PCW at 300 K under various D ion energies
Figure 2-5: NRA measurements of the near-surface D depth profiles. (a) 1 keV and 500 eV D+ (1024 D+/m2 incident fluence) implanted into W at 300 K. Implantation profiles for 1 keV and 500 eV D+ as calculated by TRVMC are shown for comparison (normalized to the peak height of the measured profiles). (b) 500 eV D+ implanted at 500 K into W (1024 D+/m2) and W-1% La2O3 (3.3×1024 D+/m2).
Figure 2-6: Fluence dependence of D retention in W at elevated temperatures using ion beams.
Figure 2-7: Fluence dependence of D retention in W at elevated temperatures using plasma devices and tokamaks.
Figure 2-8: Temperature dependence of D retention in W and W-1%La2O3. (a) 1 keV/D+ at fluences of 1023 and 1024 D+/m2, (b) 500 eV D+ at fluence of 1023 D+/m2.
Figure 2-9: Temperature dependence of D retention in M-SCW with an incident fluence of 1024 D+/m2.
Figure 2-10 (a): Deuterium retention in single-crystal and polycrystalline fine-grain tungsten exposed to low-energy (200 eV/D+) and high flux (about 1×1021 D/m2s) D plasmas as a function of exposure temperature. For comparison, the temperature dependence of the D retention in polycrystalline coarse-grained W irradiated with 200 eV/D+ ions and flux of 4×1019 D m−2 s−1 to a fluence of 1×1024 Dm−2 is also shown. Note that the deuterium retention was calculated from deuterium depth profiles measured up to a depth of 7μm.
Figure 2-10 (b): Deuterium retention in polycrystalline tungsten exposed to low-energy (98–100 eV/DT) and high flux ((8.7–10)×1021 D(T)m−2 s−1) D or (D+ T) plasmas as a function of exposure temperature.
Figure 2-11: Deuterium retention as a function of incident D+ flux at three fluences (1021, 1022, and 1023 D+/m2) at room temperature.
Figure 3-1: Schematic of single-beam ion accelerator. Figure 3-2: Implantation specimen holder. Figure 3-3: Schematic of the TDS system.
vii
Figure 4-1: SEM images of irradiated Plansee PCW (after TDS test). (a) On-spot area; (b) Off-spot area.
Figure 4-2: SEM images of irradiated Rembar PCW (after TDS test). (a) On-spot area; (b) Off-spot area.
Figure 4-3: Cross-sectional SEM photograph of Plansee PCW specimen. Figure 4-4: Signals of H2, HD, and D2 in two thermal desorption runs. (a) 200 eV/D+, 500 K,
Plansee PCW; (b) 200 eV/D+, 500 K, Plansee PCW. Figure 4-5: Example calculation of deuterium retention in Plansee PCW (200 eV/D+, 500 K).
The vertical bar indicates the estimated HD contribution (i.e., integration of HD signals over the two different time spans).
Figure 5-1: 500 eV/D+ ion implantation on PCW at 300 K. Figure 5-2: 200 eV/D+ ion implantation on PCW at 300 K. Figure 5-3: Deuterium retention in different types of PCW at 300 K. Figure 5-4: Fluence dependence of D retention in PCW at elevated temperatures. Figure 5-5: Collection of fluence dependence data on deuterium retention in tungsten at
room temperature. Figure 5-6: Fluence dependence of D retention in Plansee PCW for different energy and
temperature combinations. Figure 5-7: Collection of fluence dependence data on D retention in tungsten at elevated
temperatures. Figure 5-8: Irradiation temperature effects on D retention in PCW. Figure 5-9: Incident ion energy dependence of D retention in Plansee PCW. Figure 5-10: Energy dependence of D retention in Rembar PCW.
viii
List of tables: Table 1: Impurity content of polycrystalline tungsten specimens…………………………12
1
1. Introduction
1.1 Fusion Energy
The global energy demand is growing rapidly due to the increase in world population
and the increasing energy use per capita. The present global energy system relies heavily on
fossil fuels, which supply almost 80% of the world’s energy demand [1]. However, the
current use of fossil fuels is facing many challenges and will be restricted due to concerns
about atmospheric pollution and resource depletion [2]. In order to address our present and
future energy demand, there is an urgent need to develop a sustainable energy mix. Nuclear
fusion is considered to be one of the alternative energy options because of several
advantages, such as limitless fuel supply, no greenhouse gas emission, suitability for
large-scale electricity production and low levels of radioactive waste [3].
Nuclear fusion involves the reaction between two light nuclei which combine to form
a heavier nucleus, accompanied by a release of energy. Hydrogen isotopes, namely
deuterium and tritium, have been regarded as the most likely elements suitable for a
terrestrial fusion reactor [4]. Among the possible fusion reactions, the D-T reaction is
currently under investigation because of the large reaction cross-section at relatively low
energy [5]. In order for fusion reactions to occur, one must provide enough kinetic energy
to overcome the Coulomb repulsion force between the two positively charged hydrogen
nuclei. Heating the fuel particles to temperatures on the order of 100 million ºC will speed
up the hydrogen ions to sufficiently high velocities to undergo fusion; this method is
thermonuclear fusion. However, atoms with high thermal velocities will easily escape even
a large reaction volume in a short period of time, so it is necessary to find some means of
confining the hot fusion fuel, or plasma. During the past 50 years, two main confinement
techniques have been developed: inertial and magnetic confinement. For inertial
confinement, powerful lasers are employed to heat and compress sub-millimeter sized
capsules of fusion fuel to high densities and temperatures to reach the necessary conditions
for fusion reactions to occur [3]. In magnetic confinement, magnetic fields are used to
confine and stabilize the plasma.
The most advanced type of magnetic confinement concept is known as the tokamak,
2
derived from Russian words meaning “toroidal chamber magnetic”. A tokamak employs
both toroidal and poloidal magnetic fields to confine the plasma [5]. Large experimental
tokamak devices have been studied, such as the Tokamak Fusion Test Reactor (TFTR) with
11 MW peak fusion power [6], and the Joint European Torus (JET) with fusion power of 16
MW [7]. However, such reactors are still far from the requirement of a commercial fusion
power station. The most recent major fusion device, now under construction in France, is
the International Thermonuclear Experimental Reactor (ITER) [8]. The goal of ITER is to
demonstrate fusion ignition conditions, and to maintain a burning plasma without external
heating sources.
1.2 Plasma-facing materials
In all magnetic confinement fusion devices, including tokamaks, some plasma
particles can escape from the magnetic confinement region and impact the surrounding
wall. Such plasma-wall interactions critically affect tokamak operation in many ways [9].
Plasma erosion determines the lifetime of plasma-facing components and creates a source
of impurities, which can play a role in cooling and diluting the plasma. Deposition of
eroded impurities onto plasma-facing materials alters their surface composition and via
co-deposition may lead to long-term accumulation of large in-vessel tritium inventories.
Retention and recycling of hydrogen affect fuelling efficiency, plasma density control and
the density of neutral hydrogen in the plasma boundary, which in turn, impacts particle and
energy transport.
The first wall will be responsible for withstanding the intense heat load and particle
flux from the core plasma, over months or years of operation, with little or no maintenance.
Thus, the choice of plasma-facing materials is a critical issue for the development of ITER.
The ideal candidate would have [9]: low erosion, thereby extending the lifetime of
components; high thermal conductivity to dissipate high heat loads; thermal stress
resistance to prevent fracture due to temperature changes; resistance to the detrimental
effects of neutron irradiation; and low tritium retention for safety and fuelling issues.
Although it is impossible to find a single material which meets all of the above
requirements, the use of the divertor magnetic geometry may provide the opportunity to
employ different materials for different regions of the reactor wall. Currently, a
3
combination of beryllium, carbon and tungsten has been selected as plasma-facing
materials for ITER [10]. Beryllium, with a low atomic number, excellent oxygen gettering
capabilities but low melting temperature and high sputtering yield [11], has been chosen as
the armor material for ~80% of the total surface exposed to the plasma (primary wall, upper
baffle and as the primary option for the port limiters) [11]. Another low Z material, carbon,
which has the advantages of good power handling and thermal shock resistance, has been
selected for the primary high heat flux plasma-contact surfaces in the divertor. A particular
advantage of carbon is that it does not melt and thus can preserve its shape under high
temperature circumstances. But, carbon also has shortcomings, such as a high chemical and
physical erosion rate as well as high deuterium retention, primarily in co-deposited layers
in the divertor region [9]. Tungsten is favorable as another plasma-facing material
candidate due to its high melting temperature, low physical sputtering yield, high threshold
for physical sputtering, and no chemical sputtering [9]. However, such a high Z material
could cause high radiative losses if sputtered tungsten impurities reach the core plasma.
Therefore, the use of tungsten must be avoided in regions where exposure to significant
fluxes of energetic particles might occur.
1.3 Tungsten
This study focuses on deuterium retention in tungsten. Therefore, a brief description
about tungsten, including tungsten’s physical and chemical properties as well as their
connections to fusion use, is given in this section.
Known for its high temperature characteristics, such as the highest melting
temperature of 3683 K, lowest vapor pressure of 6.5×10-7 Pa at 2300 K, and highest tensile
strength at elevated temperatures among all metals, together with good thermal
conductivity (178 W/m·K) [12] and low sputtering yield [13] for minimizing impurity
generation, tungsten is regarded as a favored armor material for selected wall components
of future tokamaks.
The use of tungsten was initially suggested in the 1970s when it was briefly
considered an alternative to stainless steel as plasma-facing material [14]. To prevent high
radiative losses by tungsten impurities leading to unacceptable plasma cooling, it was
suggested that tungsten only be used in low ion energy regions (< 40 eV) to avoid erosion.
4
However, the limitation of tokamak technology at that time made it quite difficult to
produce sufficiently low energy plasmas. Therefore, for several decades, the research focus
shifted to low Z materials, such as carbon and beryllium.
The interest of using tungsten as a plasma-facing material has been renewed due to
advances in tokamak operation over the past 20 years, which resulted in the production of
edge plasmas with energies of 3-100 eV. As a result, recent studies have concentrated on
updating the database on hydrogen interaction with tungsten. A study on the assessment of
tungsten for use in the ITER plasma facing components indicated that tungsten armor
structures were capable of surviving as many as 200 cycles at 16 MW/m2 without any
damage [15]. In the meantime, research on low Z plasma-facing materials, such as carbon
and beryllium, suggested the high erosion rates may require more frequent reactor
maintenance and downtime to replace and repair the first wall [16].
1.4 Thesis objective
The objective of this thesis is to extend the database to improve our understanding on
D retention in polycrystalline tungsten by examining the effects of material structure and,
very importantly, extending the irradiation fluence with our ion beam accelerator to the
high fluences typically achieved in plasma devices (≥ 1026 D+/m2). Discrepancies in the
existing retention results in the literature, especially between measurements by different
research groups, have led this study to investigate deuterium retention in polycrystalline
tungsten as a function of incident fluence, irradiation temperature, ion energy, and surface
structure.
2. Background: deuterium retention in tungsten
Plasma-wall interactions are among the critical issues in fusion devices. In particular,
hydrogen retention in plasma-facing components can affect fuelling efficiency, plasma
density control, and the density of neutral hydrogen in the plasma boundary [9]. This in
turn affects particle and energy confinement. Tritium inventory and permeation through the
wall or into coolant channels are also concerns for the safety of reactors. Therefore, the
5
underlying mechanism of hydrogen trapping in plasma-facing materials has been widely
studied, but with inconsistent results. Generally, two types of tungsten materials have been
studied, namely single-crystal tungsten (SCW) and polycrystalline tungsten (PCW). It has
been suggested that the concentration and nature of trapping sites in the bulk material
would largely determine the transport and retention of hydrogen [17]. Since PCW is
planned for use in ITER [10], deuterium retention in polycrystalline tungsten has been
selected as the primary topic to be investigated in this thesis.
Previous investigations on hydrogen retention in tungsten have been performed using
various experimental techniques and under different irradiation conditions. Generally, ion
implantation has been used to simulate the ions and neutrals incident on plasma-facing
materials in tokamaks. Common ion sources include accelerators, ion guns and linear
plasma devices (e.g. PISCES [18]). The hydrogen remaining in a specimen following ion
implantation can be determined by thermal desorption spectroscopy (TDS) or nuclear
reaction analysis (NRA). The depth profile of hydrogen in tungsten could be obtained from
secondary ion mass spectroscopy (SIMS) or nuclear reaction analysis (NRA). The
microstructure of the materials could be observed by scanning electron microscopy (SEM).
In addition, several simulation codes, such as TMAP, have been developed to interpret
hydrogen trapping mechanism in tungsten [19]. The TAMP-7 code was used by Poon et al.
to derive D trapping energies in SCW [19a].
2.1 Irradiation effects on structure evolution of tungsten materials
The minimum energy of hydrogen and deuterium ions for displacement damage
production is ~2050 eV for H+ [20] and ~960 eV D+ [21]. Irradiation with H or D ions
above the threshold leads to elastic collisions resulting in nucleation and growth of
dislocation loops. When the incident D+ ion energy is below the displacement threshold in
W, ion-induced damage due to D+ does not occur, but it is still possible to produce damage
by knock-on collisions between D+ ions and surface impurities [22]. At essentially all
incident energies, hydrogen accumulation, extending to depths much greater than the ion
range, may cause lattice distortion, even leading to the formation of vacancy clusters or
nano-bubbles, micro-voids and in some cases blistering [21].
Sakamoto et al. [20] studied hydrogen ion bombardment effects on tungsten
6
microstructure. For the energy range of 0.5-8 keV H+ and implantation temperatures of
300-1073 K, TEM observations showed that for H+ < 2 keV, an incident fluence of
1.5×1022 H+/m2 did not lead to any change in the single-crystal tungsten’s inner structure at
room temperature. However, when using 3 keV H+, which is above the threshold energy for
displacement damage, hydrogen clusters and dislocation loops were created. For 8 keV H+
implanted into single-crystal tungsten, a temperature dependence study showed dislocation
loops coalesced to form dislocation networks at 373 K and below; between 473 K and 773
K, dislocation loops were formed but without networks; at 873 K or above, fine hydrogen
bubbles were observed [20]. This study also showed that defect accumulation was more
prominent in polycrystalline tungsten than in single-crystal tungsten, suggesting higher D
retention levels would be found in polycrystalline tungsten.
Alimov et al. [23] compared deuterium retention and lattice damage in two types of
W, namely SCW and hot-rolled PCW at two ion energies (6 keV/D+ and 10 keV/D+) and
two temperatures (300 K and 650 K). They found at least two types of ion-induced defects
causing trapping of deuterium: (1) micro-voids filled with D2 located in the implantation
zone; (2) dislocations distributed from the surface to depths far beyond 1 μm. SIMS/RGA
depth profiles (Figure 2-1) revealed that 6 keV/D+ irradiation of W (both SCW and PCW)
at 300 K led to the accumulation of D atoms both in the ion stopping zone and at depths of
up to several microns. The existence of D trapped as D2 molecules in micro-voids was
inferred from the release of D2 during sputter-SIMS measurements. After D+ implantation
at 650 K, D2 molecules were not observed and deuterium was accumulated solely in the
form of D atoms at depths of up to 500 nm. In the stopping zone of SCW irradiated with 6
and 10 keV/D+ at 300 K, the maximum concentration of D accumulated in both states
reaches a value of 0.10 D/W while the maximum value is 0.015 D/W for 6 keV/D+
irradiation at 650 K [23]. RBS/C analysis indicated that micro-voids and dislocation-type
defects were two existing types of hydrogen trapping, the former within the ion
implantation zone and the latter extending to depths far beyond 1 micron. Heating up to
900 K during implantation, the D2-filled micro-voids completely disappear while the
dislocation density decreases at depths beyond the implantation zone and increases in the
ion stopping zone [23].
Haasz et al. [24] used both 3 keV D3+ ions (1 keV/D+: above displacement threshold)
7
and 1.5 keV D3+ ions (500 eV/D+: below threshold) to investigate the effect of ion-induced
damage on W. In this study, specimens were subjected to different damage histories by
varying the incident D+ fluence prior to exposing specimens to a “probe” fluence (TDS was
performed after each “damage” fluence). Figure 2-2 shows the effect of such an
implantation history on D trapping. For specimen W1, the retention values for
damage-fluence implantations of 1 keV/D+ and 9×1023 D/m2 showed a small increase with
cumulative fluence, but the values appeared to level off after ~8×1024 D/m2, approximately
a factor of 2 above the virgin damage implantation value. The damage and probe fluences
used for the 500 eV/D+ ion implantations were identical to those used for the 1 keV/D+ case,
with the addition of 3×1024 D/m2 and 3×1025 D/m2 damage-fluence runs. The results from
these specimens are shown in Figure 2-3. Retention values for a particular set of
damage-fluence runs increased as a function of cumulative fluence, but the dependence
appeared to be weak. It is noted that the retention values corresponding to the 3×1024 D/m2
damage-fluence runs (W6) were about twice the values obtained for the 9×1023 D/m2 runs
(W4), but increasing the damage fluence from 3×1024 D/m2 (W6) to 1025 D/m2 (W7) or
3×1025 D/m2 (W9) did not significantly increase the retention levels [24]. This suggests that
for 500 eV D+ implantation at 500 K, the amount of ion induced damage saturates for
incident-fluences above 3×1024 D/m2 [24].
The results for 500 eV/D+ ions for the “probe” fluence showed increased D retention
in the specimens, which was suggested to be due to swelling-induced stresses and/or
precipitation of W hydrides leading to dislocation creation and grain cracking [24].
Evidence of surface blisters at high fluences was also seen [24]. Such effects appear to be
dependent on both ion energy and incident fluence. Furthermore, the damage was not
removed during TDS to 2100 K so that more trapping sites are present in subsequent
implantations. Identical experiments using 1 keV/D+ ions did not show any significant
change in “probe” fluence retention due to prior implantations [24].
2.2 Fluence dependence of deuterium retention in tungsten
Since polycrystalline tungsten has been selected as one of the candidates for
plasma-facing components in ITER, deuterium retention in tungsten as a function of the
incident ion fluence has been widely investigated. It is anticipated that the typical
8
irradiation fluence in the ITER divertor is ~1028- 1029 DT/m2 [9]. Linear plasma devices,
such as PISCES [18], may be able to reach such high-fluence irradiation due to their
capability to produce high flux plasmas (~1022 D+/m2s). By comparison, accelerator beams
are limited to 1017-1020 ions/m2s and hence, incident fluences of 1017-1026 ions/m2.
Therefore, estimation of hydrogen inventories in ITER wall materials from ion beam data
requires extrapolating the lower fluence results. While the fluxes of ion beams are lower
than those of plasma device, ion beam accelerators do provide better control over ion
species, ion energy and ion flux. A well-characterized beam could help understand the
deuterium trapping mechanism in tungsten. Since this study is only focusing on deuterium
retention issues in polycrystalline tungsten, experimental results reviewed here are mainly
on PCW.
Several studies on PCW irradiated at 300 K by various deuterium ion energies are
shown in Figure 2-4. Haasz et al. [25] employed 500 eV and 1 keV D ions to irradiate
specimens cut from a 25 μm thick and 99.95 wt% pure polycrystalline tungsten foil made
by Rembar Corporation. They found that for both energies the retained amount of
deuterium saturated for fluence above 1023 D+/m2, with similar retention levels (~6×1020
D/m2). The similar deuterium inventory was verified by NRA measurements in the same
study [25], which showed D was trapped to depths of 500 nm, far beyond the ion range for
both energies (see Figure 2-5(a)). In another study, Ogorodnikova et al. [26] observed a
linear increase of deuterium retention in PCW at 300 K with no sign of saturation for
incident fluences up to 2×1024 D+/m2. It should be noted that the PCW specimens used by
Ogorodnikova were cut from a reduced-rolled, 0.5 mm thick, 99.96 wt% pure tungsten foil
made by Plansee Corporation. The disagreement was possibly due to the different surface
structure of the two types of specimens, which could be caused by different specimen
preparation techniques. The resolution of this disagreement is also one of the objectives of
this thesis project.
Further studies of D retention by 200 eV/D+ ions at 300 K by Ogorodnikova et al. [27]
(measured by TDS) and Alimov et al. [28] (measured by NRA) also yielded conflicting
results; see Figure 2-4. Alimov’s results [28] illustrated a trend of saturation for incident D
ion fluence above 5×1023 D+/m2, while Ogorodnikova’s [27] showed an increasing trend
with increasing fluence, without saturation. In the case of single-crystal tungsten, Poon et al.
9
[29] found the fluence dependence of D retention in SCW for 500 eV D+ irradiations at 300
K to be similar to that of PCW [25]. The retention initially increased with increasing D+
fluence and then tended to saturate at 6×1020 D/m2 for fluences above 1023 D+/m2.
Fluence dependence of deuterium retention in Rembar PCW has also been studied at
elevated temperatures. For 500 K implantation with both 500 eV/D+ and 1 keV/D+ ions,
Haasz et al. [25] observed no sign of saturation (see Figure 2-6), which was interpreted as
indication of a diffusion-limited trapping mechanism [25]. The NRA analysis of the front
and back surfaces of the tungsten specimen both showed ~0.05 at.% concentrations,
suggesting that the trapped D is uniformly distributed thorough the bulk of the material (see
Figure 2-5(b)). For 200 eV/D+ irradiation of Plansee PCW at 380-470 K, Ogorodnikova et
al. [27] also observed the retained D to increase with increasing incident fluence; however,
in her case the magnitude and slope of the curve differed from the results of Haasz et al.
[25]. Single-crystal tungsten irradiated by 500 eV/D+ ions (produced by accelerator) was
investigated by Poon et al. [29] at 500 K case. The fluence dependence observed in [29]
also indicated an increasing trend with increasing fluence without reaching saturation even
at 1025 D+/m2; see Figure 2-6. However, PCW and SCW differed in the slope of the curves
and the magnitude of the retained amount of deuterium, with the SCW showing less
retention [25,29], which suggests that there are significantly fewer trapping sites in
single-crystal tungsten [29].
Deuterium retention as a function of incident fluence has also been investigated in
linear plasma devices [30-36] and tokamaks (e.g. [37]) at elevated temperatures; see Figure
2-7. Doerner et al. [30] reported that 100 eV D+ irradiation on ITER grade PCW at 575 K
in PISCES showed an increasing trend for deuterium retention with incident fluence
ranging from 1025 D+/m2 to 1026 D+/m2. Kolasinski et al. [32] used 70 eV D plasmas to
irradiate Plansee PCW at 623 K and determined a saturation level for an incident fluence of
1026 D+/m2. Alimov et al. [36] implanted 38 eV D plasmas into W at 530 K and found
similar retention values for fluences between 1026 and 1027 D+/m2, suggesting the
occurrence of saturation for incident fluences above 1026 D+/m2. Differences in materials
and implantation conditions must still account for the large variation in trends and absolute
retention values observed at high fluences. From the collection of data in Figure 2-7, it is
not evident whether the retention levels off or continues to increase for fluence > 1026
10
D+/m2.
2.3 Temperature dependence of D retention in tungsten
Various studies on different types of tungsten (PCW or SCW) irradiated by different
deuterium sources (D plasmas or D+ ions) at elevated temperatures have shown that D
retention in tungsten is dependent on the exposure temperature. Haasz et al. [25] and Poon
et al. [29] used deuterium ion beams with different energies to investigate temperature
effects on D retention in both PCW and SCW. It was found, for PCW [25], that diffusion of
D into the bulk increased in importance with increasing irradiation temperatures. In Figure
2-8, Haasz et al. [25] found an enhancement showing a maximum of the retained amount of
D at ~500 K compared with the 300 K case for both 500 eV/D+ and 1 keV/D+ irradiation.
In Figure 2-9, Poon et al. [29] observed that deuterium retention in SCW decreased with
increasing temperature, with no sign of a peak or plateau in the temperature dependence, as
seen with polycrystalline tungsten (Figure 2-8). No retention was observed in specimens
implanted at temperatures above 700 K. Also, it was suggested that the large scatter in the
data in Figures 2-8 and 2-9 might be a result of varying levels of background impurities
during implantation [29].
Plasma devices have also been employed to study the temperature dependence of D
retention in W; see Figure 2-10. In the study by Alimov et al. [38], D retention in both
SCW and PCW was 1×1020 D/m2 at 300 K for a deuterium plasma with energy of 200
eV/D+, a flux of 1×1021 D+/m2s and incident fluence of 2×1024 D+/m2. As the temperature
increased, the retained D rose to the maximum value of (4~6)×1020 D/m2 at 460 to 490 K,
then decreased by an order of magnitude to 2×1019 D/m2 at 530 K for SCW and 640 K for
PCW; see Figure 2-10 (a). Results of other investigations [39-42] on the temperature
dependence of D retention in PCW by D plasmas are plotted in Figure 2-10 (b). A
maximum in the retention was observed, but it appears to be shifted to higher temperatures,
compared to the studies by Alimov et al. [38] or Haasz et al. [25].
2.4 Flux dependence of deuterium retention in tungsten
High flux (1×1020 D+/m2s) and high fluence (above 1×1024 D+/m2) ion irradiation was
found to be able to create μm size surface blisters on polycrystalline tungsten [24]. It is
11
evident that hydrogen bubble formation and growth in metals depends on the local
hydrogen concentration and vacancy concentration. Thus, a study was performed by Poon
et al. [43] to change the local hydrogen concentration in single-crystal tungsten by varying
incident ion fluxes in order to observe possible flux effects on D retention.
D retention in SCW implanted with 500 eV/D+ ions at 300 K was found to vary
significantly at low flux (<1018 D+/m2s) and low fluence (<1021 D+/m2) (see Figure 2-11).
The amount of retained D was sharply decreased for fluxes below 1018 D+/m2s, indicating a
possible flux threshold for D retention. But, for higher fluences of 1022 and 1023 D+/m2, D
retention did not show significant dependence on incident flux. Poon et al. [43] suggested
that a steady state was established between the incident ions and the D diffusing out of the
implantation zone, which implies that the local mobile D concentration was dependent on
the incident flux. Hydrogen trapping occurred only when the local D concentration
exceeded a threshold, and high enough to cause lattice distortion [43]. This explains why
there was no trapping for very low fluxes. Although flux higher than the threshold led to
high concentration of mobile D concentration and induced D trapping, the trapping rate at
the high fluxes did not depend on the fluxes [43].
3. Experimental apparatus
3.1 Polycrystalline tungsten specimens
Two types of polycrystalline tungsten have been investigated in this study. One was
produced by Rembar Corporation and hot-rolled to a thickness of 25 μm. Specimens
measuring 5×10 mm2 were cut from this 99.95 wt% pure foil. The other type of PCW
tungsten (99.97 wt%) was made by Plansee Corporation without the reduced-rolled process.
Specimens were cut for us by Plansee in dimensions of 5×5×1 mm3. The impurity content
as specified by the manufacturers can be found in table 1.
12
Table 1: Impurity content of polycrystalline tungsten specimens
Impurity Rembar (99.95 wt% pure W) Plansee (99.97 wt% pure W)
Hydrogen (H) < 5 PPM < 5 PPM
Carbon (C) < 30 PPM < 30 PPM
Nitrogen (N) < 10 PPM < 5 PPM
Oxygen (O) < 30 PPM < 20 PPM
Molybdenum (Mo) < 100 PPM < 100 PPM
Silver (Ag) < 5 PPM < 10 PPM
Iron (Fe) < 30 PPM < 30 PPM
Aluminum (Al) < 15 PPM < 15 PPM
3.2 Ion beam implantation system
3.2.1 Single-beam ion accelerator
Deuterium ion implantation in PCW specimens was performed by employing the
single-beam ion accelerator facility in the University of Toronto Institute for Aerospace
Studies. A schematic layout is shown in Figure 3-1. For further details on the accelerator
facility, see Ref. [44]. Since the ion implantation needs to be conducted under high vacuum
conditions, three Leybold-Heraeus turbo-molecular pumps were used to provide differential
pumping for the system. The three vacuum sections comprise the first stage including the
first lens and bending magnet chamber, the second stage for the second lens chamber, and
the third stage for the target chamber.
The single-beam ion accelerator is capable of delivering a mass analyzed light ion
beam with energies from 0.1 to 10 keV. High purity deuterium was fed through a high
pressure regulator to a sapphire-seal variable leak valve and delivered into the
duoplasmatron ion source with a controlled flow rate. Nylon gas lines were connected to
the variable leak to provide electrical isolation. The pressure in the nylon line was kept
above atmospheric pressure to prevent possible contamination from atmosphere.
A platinum mesh filament was positioned in the center of the duoplasmatron ion
source chamber. The filament is coated with an electron emitting barium carbonate
compound and acts as the cathode for the plasma discharge. The plasma so generated is
13
confined by an axial magnetic field in the source chamber. Then a 10 kV extraction voltage
extracts the ions from the duoplasmatron source through a 0.25 mm diameter pinhole
aperture. With the acceleration caused by the extraction voltage, the ions pass through a
second orifice and arrive at the first Einzel lens. Focused by the first lens, the ions’ vertical
position is corrected when passing through a set of steering plates. A 30° bending magnet is
employed to select the desired species by mass selection; the bending magnet is capable of
selecting masses up to 16 amu. The selected ions are then decelerated by passing through a
double gridded deceleration gap to reach the final energy. Finally, a single-grid, second
Einzel lens is used for final focusing of the beam through a 3 mm diameter collimation
aperture onto the target specimen.
3.2.2 Specimen holder
During ion implantation, the target tungsten specimen was positioned in a specimen
holder (Figure 3-2). Following the procedure of Poon [45], the requirements for a specimen
holder include: allowing measurement of the beam current on both the mask and the
specimen, heating and measuring the temperature of the specimen, and specimen alignment
with the beam. To meet all the requirements, a series of layers were designed and
constructed on a stainless steel (SS) base. The base part has a 5 mm diameter aperture in
the center. The first layer on the base was a 10×15 mm2 piece of 25 μm, 99.96 wt% pure
polycrystalline tungsten foil (produced by Rembar Corporation), which was spot-welded
onto the stainless steel plate. A 1.5-mm diameter aperture was drilled through this foil in
order to define the ion exposure area of the specimen, so that it could serve to mask the
incident ion beam, allowing the central high-flux part of the beam to bombard the specimen.
On top of the tungsten foil, a 50 μm thick mica sheet with a slightly bigger aperture than
the one on the foil was used as the second layer to electrically isolate the specimen from the
tungsten foil mask, in order to make it possible to measure the beam current on the
specimen directly. With such a multi-layer set-up, the total beam current can be determined
by summing the current on the mask and the specimen.
Because tungsten irradiation was also performed at elevated temperatures, a
nickel-chromium/nickel-aluminum (chromel-alumel, type K) thermocouple was inserted
between the mica layer and the specimen to measure the temperature. A ceramic heater was
14
placed on top of the specimen and fixed by bolting through the holes on the stainless steel
plate. A graphite washer was placed between the heater and stainless steel washers to
provide a good electrical contact during heating. A thin strip of stainless steel was inserted
between the center of the heater and the back side of the specimen for measuring the beam
current on the specimen. All layers were held together by the bolts mounting the ceramic
heater to the stainless steel base plate. Three electrically isolated sections were used in the
specimen holder assembly [45], namely, (1) the specimen, thermocouple and current
measurement strip; (2) the stainless steel plate and tungsten foil mask; and (3) the bolts and
ceramic heater. In the target chamber, the specimen holder was placed vertically to face the
beam from the accelerator and was isolated from the rest of the vacuum vessel.
3.3 Thermal desorption system
Thermal desorption spectrometry (TDS) was performed in a separate (i.e., other than
the implantation chamber) vacuum system. For TDS, the specimen was placed and heated
on a 25 μm thick tungsten foil cradle made of Rembar PCW. The cradle was resistively
heated and the power to the cradle was controlled by programmed computer software.
During TDS, the temperature of the specimen was also measured by a chromel-alumel
thermocouple. In this case the thermocouple was spot-welded to the specimen; fine
platinum wires were used between the thermocouple and the specimen in order to
strengthen the bond. The released species during heating were monitored by a Hiden
quadruple mass spectrometer (QMS). The QMS was calibrated prior to each TDS run using
calibrated leak bottles of H2 and D2. The heating process was usually conducted under high
vacuum conditions, with a background pressure of ~5×10-8 Torr in the chamber. The
schematic of the thermal desorption system is shown in Figure 3-3. This system was also
used for high temperature annealing of specimens prior to ion implantation.
A Sorenson DCR 20-80B power supply, capable of providing 0-80 A current and 0-20
V voltage, was used to heat the cradle on which the specimens were placed. The power
supply was controlled by running a National Instruments LabWindows software program to
produce a smoothly linear temperature increase. This heating control program was based on
iterations and had to be adjusted for different type of specimens.
15
3.4 Scanning electron microscopy (SEM)
Previous studies [25,29] have indicated that surface preparation would influence the
surface structure of tungsten specimens, thus affecting the deuterium retention in tungsten.
Also, the surface structure of the specimen will be modified when exposed to ion beam or
plasma irradiation. As a result, the surface morphology is a good indication to interpret the
mechanism of plasma surface interactions. A Hitachi S-5200 scanning electron microscopy
(SEM) at the University of Toronto Center for Nanostructure Imaging was employed in this
study to provide high resolution images for the surface structure of tungsten specimens.
The SEM is a type of electron microscope that images the specimen surface by scanning it
with a high-energy electron beam. The electrons then interact with the atoms at or near the
surface on the specimen, producing signals that contain information about surface
topography, composition and other properties. The types of signals produced by SEM
include secondary electrons, back scattered electrons (BSE), and characteristic X-rays.
The basic operation of SEM is as follows [46]. An electron beam is emitted from an
electron gun fitted with a filament cathode. The beam typically has an energy ranging from
a few hundred eV to 40 keV. Being focused by condenser lenses to a spot about 0.4 nm to 5
nm in diameter, the beam passes through pairs of scanning coils or pairs of deflector plates
in the electron column, typically in the final lens, which deflect the beam in the x and y
axes so that it scans over a rectangular area of the sample surface. When the primary
electron beam interacts with the sample, the electrons lose energy by repeated random
scattering and absorption within a teardrop-shaped volume of the specimen known as the
interaction volume, which extends from less than 100 nm to around 5 µm into the surface.
The energy exchange between the electron beam and the sample results in the reflection of
high-energy electrons by elastic scattering, emission of secondary electrons by inelastic
scattering and the emission of electromagnetic radiation, each of which can be detected by
specialized detectors. The beam current absorbed by the specimen can be detected and used
to create images of the distribution of specimen current.
16
4. Experimental procedure
This study was undertaken to investigate deuterium retention in polycrystalline
tungsten materials, focusing on the fluence, temperature and energy dependencies, as well
as the effect of structure on D retention. A procedure consisting of four steps has been
followed in the experiment. The first step was specimen preparation, namely, high
temperature annealing. In the second step, the specimen was mounted in the specimen
holder and placed in the accelerator target chamber for irradiation by energetic deuterium
ions; the ion energy, ion flux, ion fluence and implantation temperature, were controlled so
that one could gain an understanding of the effects of these parameters on deuterium
retention. The third step was to conduct thermal desorption of the implanted specimens,
and the final step was the analysis of the thermal desorption profiles to derive D retention
levels for various irradiation conditions.
4.1 Specimen preparation and anneal
Prior to D+ implantation, the tungsten specimens were first annealed at 850-950 K for
30 minutes in the TDS chamber at a pressure of ~10-7 Torr, followed by cooling down at a
slow rate to avoid quenching. Up to 5 specimens could be annealed at once, and removed
from the heating cradle for implantation when the temperature returned to room
temperature. (We note that the maximum temperatures reached in this study (~950 K) are
much lower than those obtained in previous D retention experiments performed by the
UITAS Fusion group, which were typically ~1500 K. This happened inadvertently by using
the chromel-alumel thermocouple instead of the high-temperature tungsten-5% rhenium /
tungsten-26% rhenium.)
Referring to previous studies regarding thermal annealing effects, van Veen [47]
determined vacancy annihilation occurred at ~800 K but depended on temperature ramping
rates. Anderl et al. [48] indicated that PCW annealed to 1673 K led to a reduction of
dislocation density by a factor of 7. The study also suggested that annealing to 1500 K
increased the grain size of PCW by a factor of 3 to 5, which suggested a decrease of grain
boundaries [48]. Also, high temperature treatment played an important role in desorbing
hydrogen and other impurity gases trapped in the specimens during manufacturing.
17
However, one should note that annealing at around 900 K may not be able to sufficiently
remove the effects of irradiation history. Haasz et al. [24] found that sequential
implantations on the same location of the same specimen led to a significant increase of
deuterium retention in PCW. Therefore, in the present study, virgin specimens were used
for each implantation to avoid any cumulative effect of repeated irradiations on the same
spot.
In order to show the annealing effects and how incident deuterium ions interact with
the tungsten specimens, scanning electron microscopy (SEM) was employed to provide
images of the surface structure of both Rembar and Plansee polycrystalline tungsten
specimens after TDS tests. For each type of tungsten, SEM pictures were taken on both the
irradiated (on-spot) and the surrounding (off-spot) areas. Such characterization provided
information on the micro-structure of these two different types of tungsten, as well as the
evolution of the surface structure due to irradiation by deuterium ions.
SEM images for the Plansee PCW are shown in Figure 4-1, where image (a) indicates
the irradiated area (on-spot) and image (b) represents the surrounding area (off-spot). When
comparing Figure 4-1 (a) and (b), it is hard to find obvious differences in surface structure
of the Plansee PCW on-spot and off-spot areas. SEM photographs for the Rembar PCW are
shown in Figure 4-2; again, the image (a) represents the irradiated area (on-spot) and (b)
the surrounding area (off-spot). In image (b), the grain boundary and some unknown
impurity dots on the surface of the off-spot area are observed. However, the irradiated area
in (a) differs significantly from the surrounding off-spot area in (b). This implies that the
Rembar tungsten specimen surface has been modified by D+ ion bombardment. Overall, the
Plansee specimen appears to have greater surface roughness than the Rembar material.
Cross-sectional SEM photograph of Plansee PCW specimen shows the surface roughness is
about 5-10 μm; see Figure 4-3.
4.2 Deuterium ion implantation
After annealing, prepared specimens were placed into the specimen holder and
installed in the implantation target chamber. Following about 12 hours of pumping, the
pressure of the target chamber reached ~10-8 Torr, while the 1st accelerator stage achieved a
vacuum of ~5×10-7 Torr. Before introducing deuterium into the duoplasmatron, the filament
18
in the duoplasmatron ion source was heated to outgas the chamber to remove impurity
gases. Then, by gradually opening a leak valve, high-purity D2 gas was fed into the system
until the pressure of the first stage reached 1.5×10-5 Torr, which was found to be the
optimum for producing the highest beam current. A 150 V arc voltage (associated with a
2.5 μA arc current) was applied between the filament and plasma cavity to initiate the
plasma discharge. The 30° bending magnet was used to choose the desired species by mass
selection. A current of 1.55 A was applied to the bending magnet to generate D3+ ions. The
selected ions were then decelerated by passing through a double gridded deceleration gap
to reach the final energy.
The accelerator was operated such that the beam energy at the exit aperture was 3 keV.
Thus, to generate a 1.5 keV D3+ ion beam, a 1500 V bias voltage was applied on the
specimen. During implantation, both the total beam current and the current on the specimen
were monitored and adjusted periodically so that the variation of the currents was kept
within ±20% of the mean value. A total beam current of 25-40 μA associated with 4-7 μA
(4-7×1019 D/m2s) on the specimen was obtained during normal operations. To achieve an
ion fluence > 3×1024 D/m2, it was necessary to operate the accelerator for more than ~12 h.
For these cases, the implantation was broken up over a period of two or more days. The
longest implantation, 8×1025 D/m2, required 30 days of normal accelerator operation (8-10
hours per day).
4.3 Thermal desorption spectroscopy (TDS)
After implantation, the irradiated tungsten specimen was removed from the ion
accelerator, spot-welded to nickel-chromium/nickel-aluminum (chromel-alumel, type K)
thermocouple wires, and then installed in the cradle in the TDS system. Typically,
irradiated tungsten specimens were exposed to atmosphere for less than 30 minutes to
minimize the effect of time delay after implantation. Quastel et al. [22] found that thermal
desorption performed after 8 weeks after deuterium ion irradiation led to about a factor of 2
reduction in the measured total D retention compared with the <1 h time delay cases,
indicating that during D+ irradiation some mobile D diffuses deep into the bulk, requiring
relatively long post-irradiation times at room temperature to diffuse to the surface and be
released.
19
With the specimen loaded onto the heating cradle and inserted into the TDS system,
the TDS chamber was pumped down to ~1×10-7 Torr. Before heating the specimen by TDS,
mild baking of the test chamber to about 330-350 K was performed for 2 hours. In the same
study by Quastel et al. [22], it was suggested that mild baking of the test chamber to ~360
K resulted in the escape of about 40% of the trapped D. The loss of this weakly-trapped
deuterium corresponded to the near elimination of the 400 K desorption peak as well as a
reduction in the total amount of D retained. This suggested the existence of additional
lower energy traps, containing about 40% of the trapped D. For the present study, the
shallow traps were emptied by the 330-350 K baking, and only the more deeply trapped D
was measured; see Figure 4-4. The mild baking was also able to improve the vacuum in the
TDS chamber to below 5×10-8 Torr.
After the TDS chamber cooled down to room temperature from mild baking and a
good vacuum (< 5×10-8 Torr) was reached, TDS was ready to start. A complete thermal
desorption cycle is described as follows. First, various QMS signals were calibrated by the
corresponding leak bottles. QMS steady-state signals were measured with the leak bottle on
and off. Then the leak bottle flow rate was divided by the difference of the signals, yielding
the absolute numbers of molecules/s for a given QMS signal. In this study, D2 and H2 leak
bottles were used for quantifying these two species. The calibration factor for HD was
assumed to be the average of these two. Next step was to initiate the computerized heating
program run under LabWindows/CVI, which is a manually modified voltage profile to
control the ramping rate for a linear increase of specimen temperature. With a typical ramp
rate of 1.5-2 K/s, all the specimens were linearly heated up to 900-1000 K and dwelled
there for 2-3 minutes. During the whole process of heating and dwelling, the QMS
monitored the signals of masses 2 (H2), 3 (HD), and 4 (D2). The last step was to cool down
the system with a relatively slow cooling rate of 2 K/s in order to avoid possible quenching
of the specimen.
4.4 TDS analysis
Integration of the QMS signals produced during TDS yielded the total trapped
deuterium in a specimen. Details of the calculation for each result are shown below. The
release of D2 and HD from the specimen was considered to contribute to the deuterium
20
retention in tungsten; this is slightly different from Poon’s method [45] where only D2 was
considered. For the D2 release, the QMS mass 4 signal was used. These signals were
converted to the absolute number of D2 molecules by multiplying them with the calibration
factor. Then, the area confined by the background levels and desorption peaks was
integrated and divided by the beam spot area to yield the D2 release rate, in the unit of
m-2s-1. Finally, the retained amount of D2 was obtained by integrating the release rate by the
desorption time under the peak. This number was multiplied by 2 to yield the number of D
atom released in the form of D2 molecules; it is the number of D atoms, which is plotted in
the retention figures.
Interpretation of the HD signals was more difficult, as background contribution could
be significant. Thus, not the whole area below the HD peaks and above the background
level was integrated. Instead, only those traces that clearly tracked the D2 counterparts were
considered to contribute to D retention from the HD signal. In order to assess the
uncertainty from selecting HD signals, an ‘upper limit’ was also obtained by including a
larger time span for the HD signals. Figure 4-3 shows typical traces of H2, D2, HD QMS
signals in two thermal desorption runs, namely, 200 eV/D+ irradiation on Plansee PCW at
both 500 K (Figure 4-4 (a)) and 300 K (Figure 4-4 (b)). In Figure 4-4 (a), HD signals from
300 s to 500 s were regarded as HD’s main contribution part for D retention and extracted
to calculate the ‘best-guess’ value. To obtain the ‘upper limit’, HD signals from 200 s to
700 s were considered to cover all possible HD signals that contributed to D retention.
Similar calculation was used in Figure 4-4 (b). In the 300 K case shown in Figure 4-4 (b),
no noticeable desorption peak was observed at 400 K, implying that mild baking of the test
chamber to about 330-350 K before TDS has released the majority of the weakly-trapped
deuterium from low-energy trapping sites. Normally only less than 20% of the calculated
total D retention was contributed by HD signals. To give a clear illustration of this
calculation, results of deuterium retention in Plansee tungsten under 200 eV/D+ irradiation
at 500 K are plotted in Figure 4-5. Comparison between trend lines of “D from D2 only”
and “upper limit” values shows that the QMS signal (mass 4) D2 is the major part (almost
80%) of the retained deuterium. Also, these two trend lines tend to be parallel, indicating a
consistent ratio of D2 and HD for the different TDS runs.
21
5. Results and discussion
This chapter describes the experimental results on deuterium retention in
polycrystalline tungsten measured in this study. Three different controlling parameters were
investigated, namely incident ion fluence, specimen temperature, and ion energy. Among
these, fluence dependence is the main focus of this thesis, leaving temperature and ion
energy dependence as brief investigations. For each controlling parameter, the objective,
experimental conditions and the results are given below, followed by discussions of the
experimental results.
5.1 Fluence dependence of D retention in PCW
Two types of PCW materials with different surface structures and preparation, namely
Rembar and Plansee, were tested in this study. The deuterium retention in these two
different types of tungsten made it possible to assess the effects of surface condition on D
retention in PCW. The irradiations were performed at 300 K and 500 K. D+ energies of 200
and 500 eV/D+ were selected; these energies are well below the 940 eV displacement
threshold to avoid deuterium-induced damage in tungsten. A significant objective of this
work was to increase the incident fluence an order of magnitude higher than fluences
achieved in previous ion-beam studies. Our fluence range was 1×1021 to 8×1025 D+/m2,
with the upper end being close to fluences achieved in plasma devices, allowing for
comparisons between the current beam results and those from linear plasma devices and
tokamaks [30-37].
5.1.1 D retention in Rembar PCW at 300 K
At room temperature (~300 K), D retention measurements were made with both the
Rembar and Plansee PCW specimens. For the Rembar PCW, a systematic study of 500
eV/D+ and 1 keV/D+ irradiations has already been performed by Haasz et al. [25]. Here we
only use energies below the threshold. This study started with 500 eV/D+ ion implantation
in Rembar PCW for repeating the experiments shown in [25]. Then new irradiations with
200 eV/D+ were performed to explore the fluence dependence in PCW at this energy.
22
Deuterium retention in Rembar PCW by 500 eV/D+ at 300 K as a function of incident
fluence is plotted in Figure 5-1. The incident fluence for this study ranges from 1×1022 to
1×1025 D+/m2. The ion flux was controlled at 5-7 ×1019 D+/m2s for these runs. Published D
retention in Rembar PCW [25] and Plansee PCW [26] under the same irradiation
conditions (500 eV D+, 300 K) are included in Figure 5-1 for comparison. The results of
Haasz et al. [25] for D retention in Rembar PCW agree well with the present Rembar PCW
results. For low fluences (<1×1023 D+/m2), D retention increases with increasing incident
D+ fluence. When the incident fluence reaches ~1×1023 D+/m2, the retention tends to
saturation at a level of 4-5×1020 D/m2, which is similar to the saturation level obtained by
Haasz et al. [25]. However, both these sets of results differ significantly from the results of
Ogorodnikova et al. [26], where D retention levels are higher, with no sign of saturation
(albeit she only has two data points). The speculation was that such a disagreement could
be caused by the different types of polycrystalline tungsten specimens used for these
studies, namely Rembar for Haasz et al. [25] and the present study, while Plansee for
Ogorodnikova et al. [26]. In order to further explore the effect of different materials, this
thesis has also investigated D retention measurement in Plansee PCW; see sections 5.1.2
and 5.1.3.
Now, we address new experiments with 200 eV/D+ irradiation of Rembar PCW at 300
K. D retention results are plotted in Figure 5-2 for the fluence range 1×1022 to 2×1025
D+/m2. Comparison is made with results obtained by Ogorodnikova et al. [27] and Alimov
et al. [28] under similar irradiation conditions. For the 200 eV/D+ case in the present study,
D retention at 300 K still shows a trend to saturate at a level of 5-6×1020 D/m2 for incident
fluences above 2×1024 D+/m2. For this 200 eV case, saturation occurs at an incident fluence
10 times higher than that for the 500 eV case; the retention value is also higher than the 500
eV case. When comparing the present 200 eV results to the results of Ogorodnikova et al.
[27] and Alimov et al. [28], we observe the following: Ogorodnikova et al. [27] observed
an increasing D retention with increasing incident D+ fluence without a clear sign of
saturation, although, with scatter in the data above ~1024 D+/m2, the start of saturation
cannot be ruled out. The incident fluence of Ogorodnivoka’s study was in the range 3×1021
to 6×1024 D+/m2, and for the highest fluence (6×1024 D+/m2), the retention level reached as
high as 3×1021 D/m2, almost 5 times higher than the saturation level observed in this thesis.
23
While the results of Alimov et al. [28] show a clear trend to saturation (Figure 5-2) for
incident fluences above 2×1024 D+/m2, which is similar to our current study, the saturation
level is ~2×1020 D/m2, about 3 times lower than the one observed in this thesis. A possible
explanation might be the use of different techniques for trapped D measurement – TDS
here and NRA by Alimov el al. [28].
5.1.2 D retention in Plansee PCW at 300 K
The motivation for performing experiments with Plansee PCW arose because of the
different fluence-dependence results of Haasz et al. [25] and Ogorodnikova et al. [26,27].
The fluence dependence of D retention in Rembar PCW irradiated by 500 eV/D+ ions at
room temperature by Haasz et al. [25] showed a clear trend to saturation. However, the
fluence dependence of D retention in Plansee PCW irradiated by 200 and 500 eV/D+ at 300
K by Ogorodnikova et al. [26,27] showed no sign of saturation. Also, different amounts of
trapped deuterium were reported by Alimov el al. [28] and Haasz et al. [25], although, the
two studies both observed a saturation trend; see Figure 5-2. With an attempt to resolve the
above discrepancies, a new set of D+ irradiation experiments with 200 and 500 eV/D+ ions
at 300 K were performed on Plansee PCW. The specific focus was on whether D retention
continues to increase with increasing fluence, as seen by Ogorodnikova et al. [26,27] or
does it saturate as seen by Haasz et al. [25]. In cases where saturation does occur, what
controls the “onset” incident fluence and the amount of retained D at saturation, what are
the reasons for the different levels seen by Haasz et al. [25] and Alimov et al. [28]?
The present D retention measurements in Plansee PCW irradiated by 200 and 500
eV/D+ ions at room temperature are shown in Figure 5-3, as a function of fluence from
3×1022 to 3×1024 D+/m2. The ion flux for these runs was well controlled at 3-4 ×1019
D+/m2s. In order to observe the effects (if any) of the different types of polycrystalline
tungsten, D retention data for Rembar PCW (shown in Figure 5-1 and 5-2) are also plotted
in Figure 5-3. Plansee PCW irradiated by 500 eV/D+ ions tends to saturate at about
3.7×1020 D/m2, which is slightly lower than the saturation level (~5×1020 D/m2) of Rembar
PCW irradiated at the same conditions. However, a significant difference in D retention is
seen between Plansee and Rembar PCW particularly at lower incident fluences (< 1×1023
D+/m2), where retention values are a factor of 4 to 5 lower in Plansee than in Rembar PCW.
24
High incident fluences of 200 eV/D+ ions implanted into Plansee PCW at 300 K lead
to a saturation level of 4×1020 D/m2, which is close to the 3.7×1020 D/m2 for the 500 eV
case (shown in Figure 5-3). The behavior of fluence dependence for 200 eV/D+ depicted in
Figure 5-3 tracks the 500 eV curve very well, showing that every 200 eV data point is close
to the corresponding 500 eV point at the same incident fluence. From Figure 5-3, we note
that the retention values for the Plansee PCW differ from those of the Rembar PCW,
although both of them show clear evidence of a saturation trend. It is possible that the
different retention levels for the Rembar and Plansee PCW are caused by the different
surface structures and preparation conditions for these two types of specimens. The SEM
photographs presented in section 4.1 do show different surface structures (Figures 4-1 and
4-2). For the Plansee PCW, ion irradiation appears to cause some changes on the specimen
surface, while for the Rembar PCW, the effects of ion bombardment are clearly observed –
the surface has become much smoother due to ion irradiation. Consequently, the difference
on the surface structure could play a role in determining different retention levels in these
two types of polycrystalline tungsten specimens.
5.1.3 D retention in Plansee PCW at 500 K
In contrast to the 300 K cases, where D retention in tungsten tends to saturate,
independent of the incident ion energy and type of tungsten materials, deuterium trapping
in tungsten at 500 K is reported to show an increasing trend with incident fluence
[25,27,29,30]. As for the 300 K studies in sections 5.1.1 and 5.1.2, a systematic study of the
fluence dependence of D retention in Plansee PCW at 500 K was also performed in this
thesis. Most of the experiments were performed with 200 eV/D+, however, to investigate
the dependence of D retention on incident ion energy, some runs were also done with 100
and 500 eV/D+; see section 5.3. A plot of D retention in Plansee PCW at 500 K is shown in
Figure 5-4. The 200 eV/D+ irradiation results at 380-470 K for Plansee PCW by
Ogorodnikova et al. [27] are also plotted for comparison. In order to discern possible
different deuterium trapping behavior in different types of polycrystalline tungsten, D
retention data in Rembar PCW at 500 K by Haasz et al. [25] are also included in Figure 5-4.
Further, since the incident fluence in this thesis has been extended to approach ~1×1026
D+/m2 to provide overlap with results from linear plasma devices, some results from
25
PISCES [30] are also shown.
For the present 500 K study, the incident 200 eV/D+ ion fluence spreads 5 orders of
magnitude, ranging from 1×1021 to 8×1025 D+/m2, and the ion flux was controlled at 5-7
×1019 D+/m2s. The D retention in Plansee PCW, as a function of incident fluence in Figure
5-4 illustrates that D retention increases with increasing incident fluence with a slope of
0.35-0.4, with no indication of saturation. Irradiation studies, also on Plansee PCW, under
the similar irradiation conditions (380-470 K and 200 eV/D+) by Ogorodnikova et al. [27]
also showed a constantly increasing trend in D retention, but with a slope above 0.5,
indicating a higher increasing rate in D retention. The highest incident D+ fluence reported
by Ogorodnikova et al. [27] was ~3×1024 D+/m2, more than an order of magnitude lower
than the highest fluence reached in the present study. However, for 3×1024 D+/m2, the
retention in [27] has already reached ~6×1020 D/m2, which is higher than the retained
amount for 8×1025 D+/m2 obtained in the present study. Although both [27] and the present
study used the same type of W material and similar irradiation conditions, the retention
data still differ in both the retained amount of D and the slope of the increasing trend line.
The differences must now be explained by different specimen preparation techniques or
different experimental procedures.
D retention in Rembar PCW at 500 K studied by Haasz et al. [25] is also shown in
Figure 5-4. The comparison between Plansee and Rembar PCW shows that the retained
deuterium for 500 eV irradiation in the Plansee PCW at 500 K at a fluence of ~1023 D+/m2
is about 3-4 times lower than in the Rembar PCW, which is possibly due to the different
structure and preparation procedures for the two types of tungsten specimens. As discussed
in 5.1.1, the difference in the surface structure could play a role in determining different
retention levels in these two types of polycrystalline tungsten specimens. In comparing the
current results with those from PISCES [30], we note that magnitudes of the amounts
retained are very similar, but there is a steeper fluence dependence slope in the PISCES
data.
5.1.4 Discussion of fluence dependence
A plot consisting of the existing published data and the present results for the fluence
dependence of deuterium retention in tungsten at room temperature (near 300 K) is shown
26
in Figure 5-5. Both polycrystalline (Plansee PCW and Rembar PCW) and single-crystal
tungsten (JM-SCW from Johnson-Matthey and M-SCW from State Institute of Rare Metals,
Moscow [29]) are included in this plot. It is noted that the fluence dependence data
obtained at the University of Toronto at 300 K consistently show a saturation trend at high
fluences, independent of the type of tungsten (Plansee or Rembar PCW, JM or M SCW)
and ion energy (200, 300 or 500 eV/D+). Similar saturation trend was also seen by Alimov
el al. [28]. In contrast, no saturation trend was found at room temperature by Ogorodnikova
et al. [26,27] at both 200 and 500 eV/D+, nor by Golubeva et al. at 323 K [49].
In assessing the D retention results from the various studies regarding the existence of
a saturation trend, the highest incident ion fluence achieved in the different experiments is
an important factor. For example, in the case of 200 eV/D+ irradiation of the Rembar PCW
performed in the present study, the saturation trend was not fully confirmed until an
incident fluence of 2×1025 D+/m2. However, for the experiments of Ogorodnikova et al
[26,27] and Golubeva et al. [49], the maximum fluences reported were 5×1024 D+/m2 and
7×1023 D+/m2, respectively. Thus, the absence of an observed saturation trend may simply
be due to the fact that a high enough fluence has not yet been reached.
A possible explanation for the saturation trend at 300 K was suggested by Haasz et al.
[25] based on NRA depth profiles measured on both the front and back surfaces of a
Rembar PCW specimen. The depth profile of the trapped D on the front surface of the
irradiated specimen showed that the trapped D extended well beyond the implantation zone
(~ 40 nm) to ~500 nm, while on the back surface no D was detected, implying that
diffusion may be too slow at 300 K to allow trapping sites deep in the bulk to be reached.
The D retention results at 500 K irradiations on Plansee PCW obtained in the present
study show that the amount of trapped D linearly increases with increasing incident D ion
fluence with a slope of 0.35-0.4, showing no sign of saturation; see Figure 5-4. A plot of D
retention in Plansee PCW as a function of incident fluence at different energy/temperature
combinations is shown in Figure 5-6. It is observed that at fluences below ~1024 D+/m2, the
retained amount of D at 500 K and 200 eV/D+ is less than that at room temperature, but at
the highest fluence measured (8×1025 D+/m2), the D retention value at 500 K is about 2
times larger than the saturated level for the 300 K cases. A similar trend was also observed
in the study of D retention in Rembar PCW at 500 K by Haasz et al. [25], which suggested
27
diffusion taking greater importance at elevated temperatures and the results were consistent
with a diffusion-limited trapping mechanism (slope = 0.5).
A composite plot containing most of the existing published and present data for the
fluence dependence of D retention in W at ~500 K is shown in Figure 5-7. For the ion beam
studies [25,27], D retention was seen to increase with increasing incident ion fluence, with
no sign of saturation. However, for irradiations performed in plasma devices or tokamaks,
no consistent trends were observed. For instance, the results of Kolasinski et al. [32] with a
70 eV D plasma used to irradiate Plansee PCW at 623 K show a saturation level for
incident fluences of ~1026 D+/m2. Alimov et al. [36] using a 38 eV D plasma at 530 K
found similar retention amounts for 1026 and 1027 D+/m2 irradiations, suggesting the
occurrence of saturation for an incident fluence above 1026 D+/m2. Differences in materials
and implantation conditions must still account for the large variation in trends and absolute
retention values observed at high fluences.
To explain the non-saturation trend of fluence dependence at 500 K, the study by
Haasz et al. [25] provides a good reference. The NRA profiles of a specimen implanted at
500 K by 500 eV D+ (1×1024 D+/m2) showed that a nearly uniform concentration of D was
found through the specimen ~0.05 at.% on both the front and back surfaces, indicating that
almost all of the traps in the bulk are accessible to the diffusing D [25]. According to such
results, it is suggested that one must go to higher fluences to reach bulk saturation, or there
is a source of traps produced by the irradiation process [25]. With the specimens of the
current study being 40 times thicker (1 mm compared to 25 μm) than those in [25], we
would not expect to see signs of saturation until considerably higher fluences than what we
were able to reach (i.e. 8×1025 D+/m2).
5.2 Temperature dependence of D retention in PCW
Here we have briefly examined the effect of specimen temperature during
implantation on the deuterium retention in Plansee PCW. As indicated in section 2.3, large
discrepancies exist in the current data base on the temperature dependence, motivating
further study.
At the University of Toronto, several previous studies have investigated the
temperature effects on deuterium retention in both single-crystal and polycrystalline
28
tungsten materials. For PCW, investigations using 500 eV/D+ on Rembar PCW [25] found
a localized peak in retention at an implantation temperature of 450 K. Lower amounts of
retained D are observed at both lower and higher temperatures. D retention at or above 700
K was below detectable levels. For single-crystal tungsten, Poon et al. [27] found that the
highest D retention was observed for 300 K irradiations, and the retention decreased with
increasing irradiation temperature, with no local maxima observed over the studied
temperature range. Alimov et al. [38] also observed a decreasing trend of D retention with
increasing irradiation temperature from 350 to 570 K. The purpose of the present study was
to determine the behavior of D trapping in another type of polycrystalline tungsten, namely
Plansee PCW.
An ion energy of 200 eV/D+ was selected to perform the implantations at fluxes of
4-6×1019 D+/m2s and a fluence of 1×1023 D+/m2. The implantation temperature was varied
between 300 K and 500 K. The measured D retention is shown as a function of irradiation
temperature in Figure 5-8. Results for 500 eV/D+ ion irradiation on Rembar PCW by Haasz
et al. [25], 500 eV/D+ ion irradiation on SCW by Poon et al. [29], and 200 eV/D+ ion
irradiation on Plansee PCW by Alimov et al. [38] are also included here for comparison.
For the new results obtained in this study, the highest retained deuterium amount was found
at room temperature (300 K) with a value of 1.7×1020 D/m2, and then D retention decreases
with increasing irradiation temperature, reaching 5×1019 D/m2 at 500 K. Clearly, the
observed D retention behavior for the Plansee PCW is significantly different from that of
the Rembar PCW. Interestingly, the temperature dependence of the Plansee PCW D
retention does show a similar trend to that observed for single-crystal tungsten [29].
We note that the different fluence dependencies observed at different temperatures –
e.g., saturation at 300 K, no saturation of 500 K – mean that any temperature dependence
profile will depend on incident fluence.
5.3 Ion energy dependence of D retention in PCW
Three different D+ energies were used for irradiation, namely 100, 200 and 500 eV/D+.
Although energy transfer from 500 eV D+ to a W atom (21 eV) is insufficient to create a
displacement [20], the energy transfer from 500 eV D+ to oxygen and carbon are 200 eV O+
and 250 eV C+, assuming two-body collisions. Due to their larger mass, the energy transfer
29
from O+ or C+ to W is more efficient than from D+. Thus, it is possible for 500 eV D+ to
create vacancies in W through intermediate recoil collisions with O and C impurities on the
surface of W [22]. But 200 eV and 100 eV D+ cannot induce such knock-on effect because
the energy transfer from D+ to O or C is insufficient to create displacement in W.
Irradiations were performed at 100, 200 and 500 eV/D+ energies and 300 K and 500
K temperature; see Figure 5-9. Three different fluence/temperature combinations are shown.
In the first case, 300 K and incident fluence of 1×1023 D+/m2, the retained D shows only a
slight increase as the energy is increased from 100 to 500 eV/D+. The second case, for an
incident fluence of 1×1024 D+/m2 at 300 K, similarly shows only a slight increase with
increasing energy. However, as expected from the fluence dependence curves at 300 K in
Figure 5-3, the D retention levels at a fluence of ~1024 D+/m2 are about a factor of 2 higher
than at ~1023 D+/m2; this difference prevails over the 100-500 eV/D+ energy range. For the
last case, a fluence of ~1023 D+/m2 was used and the irradiation temperature was raised to
500 K; the retention levels are noticeably lower than for the previous two cases, but a factor
of 3 increase in retention is noted as the energy increases from 200 to 500 eV.
Given the scatter in the experimental results, it is only possible to conclude that for
D+ energies below 500 eV/D+, D retention in the Plansee PCW at 300 K depends only
weakly on the incident ion energy. The dependence is stronger at 500 K and it is possible
that this is associated with recoil displacements due to O and C impurities. A similar
conclusion might be drawn from the results of [25], shown in Figure 5-10. Even though the
maximum energy, 1 keV/D+, is already above the displacement threshold of 940 eV/D+, it
is still difficult to observe any noticeable effects of ion energy on D retention [25]. The
spread of the retention measurements at 500 K and 1 keV in [25] has been attributed by the
authors to accumulated damage progressively created by successive implantations on the
same spot [24].
6. Conclusions
6.1 Fluence dependence
30
Irradiation at 300 K: deuterium retention in both Rembar and Plansee PCW foils at
room temperature, as a function of incident D+ ion fluence, indicates a trend to saturation.
For Rembar PCW, the saturation level for 500 eV/D+ irradiation was ~6×1020 D/m2 for
incident fluences above 1×1023 D+/m2, similar to the results of Haasz et al. [25]. Irradiation
performed by 200 eV/D+ ions caused the D retention to level off at ~7.5×1020 D/m2 (only
slightly higher than the 500 eV/D+ case) for incident fluence above 1×1024 D+/m2. For the
Plansee PCW, irradiations by 500 and 200 eV/D+ ions demonstrated similar trapping
behavior, with the retention reaching saturation levels of ~4×1020 D/m2 for incident
fluences above 5×1023 D+/m2.
Irradiation at 500 K: only Plansee PCW was studied here. D retention results for 200
eV/D+ irradiations suggest that D retention in the Plansee PCW increases linearly with
increasing incident D+ fluence without any indication of saturation. Even when the incident
fluence was increased to 8×1025 D+/m2, which is in the range of plasma devices, there was
still no sign of saturation. The Retention value corresponding to the highest incident
fluence (8×1025 D+/m2) was found to be 5.2×1020 D/m2 – only slightly higher than the
saturation levels in Plansee PCW irradiated with 500 and 200 eV/D+ at 300 K. However,
we note again that at 500 K, the retained D is seen to increase at least over the incident
fluence range 1021 to 1026 D+/m2. The retained amount of D obtained at the highest fluence
reached in our ion-beam experiment (~8×1025 D+/m2) is similar to some retention results
obtained in plasma devices. The remaining question is whether the increasing trend
continues above ~1026 D+/m2 into the ITER fluence regime.
6.2 Temperature dependence
Our temperature dependence study of D retention in the Plansee PCW irradiated with
200 eV/D+ ions showed that D retention decreases with increasing irradiation temperature
in the 300 to 500 K range. The highest retention level was found to be ~1.7×1020 D+/m2 at
300 K, followed by a linear decreasing trend, reaching ~5×1019 D+/m2 at 500 K.
6.3 Ion energy dependence
D retention in the Plansee PCW irradiated at 300 K to a fluence of ~1×1023 D+/m2
was seen to increase only slightly with ion energy from 100 eV/D+ to 500 eV/D+; the
31
retention values were all within the range 1-2×1020 D/m2. Similar results were found for the
Plansee PCW irradiated at 300 K over the same energy range to an incident fluence of
1×1024 D+/m2; the retained levels were slightly higher in the range 3-3.5×1020 D/m2. For
irradiations at 500 K, the D retention increased noticeably (a factor of ~3) as the energy
increased from 100 to 500 eV/D+. Based on these limited results, one can draw the
conclusion that the energy of incident D ions plays a minor role in affecting D trapping in
polycrystalline tungsten.
32
References [1] Energy Information Administration, U.S. Department of Energy, "World consumption of primary energy by energy type and selected country groups, 1980-2004", July 31st 2006, retrieved on January 20th 2007. [2] R.M. Bell and D.A.J. Rand, “Clean Energy”, The Royal Society of Chemistry 2004. [3] R.G. Watts, “Innovative Energy Strategies for CO2 stabilization”, Cambridge University Press 2002. [4] R.A. Serway, C.J. Moses, C.A. Moyer, “Modern Physics”, 2nd Edition, Saunders College Publishing 1997. [5] T.J. Dolan, “Fusion Research”, Pergamon Press 1982. [6] R.J. Hawryluk, “Results from D-T experiments on TFTR and implications for achieving an ignited plasma”, Phil. Trans. R. Soc. Lond. A 357 (1999) 443. [7] M. Keilhacker, A. Gibson, C. Gormezano, P.J. Lomas et al., “High fusion performance from deuterium-tritium plasmas in JET”, Nucl. Fusion 39 (1999) 209. [8] R. Aymar, “The ITER reduced cost design”, Fusion Eng. Design 49&50 (2000) 13. [9] G. Federici, C.H. Skinner, J.N. Brooks, J.P. Coad et al., “Review: Plasma-material interactions in current tokamaks”, Nuclear Fusion, Vol. 41, No. 12R, 2001. [10] C.H. Skinner, A.A. Haasz, V.Kh. Alimov, N. Bekris et al., “Recent advance on hydrogen retention in ITER’s plasma-facing materials: beryllium, carbon, and tungsten”, Fusion Science and Technology 54 (2008) 891-945. [11] R.A. Anderl, R.A. Causey, J.W. Davis, R.P. Doerner et al., “Hydrogen isotope retention in beryllium for tokamak plasma-facing applications”, J. Nucl. Mater. 273 (1999) 1-26. [12] W.D. Callister, “Materials Science and Engineering: An Introduction”, Wiley & Sons (1985), Toronto. [13] W. Eckstein, “Physical sputtering and radiation enhanced sublimation”, Physical Processes of the Interaction of Fusion Plasmas with Solids, p. 93. [14] S.K. Erents, Proc. 8th Symp. On Fusion Tech. (SOFT), Utrecht, Netherlands, 1974, pg. 895. [15] J.W. Davis, V.R. Barabash, A. Makhankov, L. Plochl, K.T. Slattery, “Assessment of tungsten for use in the ITER plasma facing components”, J. Nucl. Mater. 258-263 (1998)
33
308-312. [16] R. Behrisch, G. Federici, A. Kukushkin, D. Reiter, “Material erosion at the vessel walls of future fusion devices”, J. Nucl. Mater. 313–316 (2003) 388–392. [17] A.A. Haasz, M. Poon, R.G. Macaulay-Newcombe, J.W. Davis, “Deuterium retention in single crystal tungsten”, J. Nucl. Mat. 290-293 (2001) 85-88. [18] R. Doerner, R.W. Conn, D. Phelps, L.M. Waganer, “The PISCES-upgrade facility for fusion plasma-materials interactions research”, Proceedings of the 18th Symposium on Fusion Technology, p 775-8 vol.1, 1995. [19] G.R. Longhurst, D.F. Holland, J.L. Jones, B.J. Merrill, “TMAP4: Tritium Migration Analysis Program, Description and User’s Manual”, INEL report, EGG-FSP-10315, EG & Idaho Inc. (1992). [19a] M. Poon, A.A. Haasz, J.W. Davis, “Modelling deuterium release during thermal desorption of D+-irradiated tungsten”, J. Nucl. Mater. 374 (2008) 390-402. [20] R. Sakamoto, T. Muroga, N. Yoshida, “Microstructural evolution induced by low energy hydrogen ion irradiation in tungsten”, J. Nucl. Mater. 220–222 (1995) 819. [21] A.D. Quastel, J.W. Davis, A.A. Haasz, R.G. Macaulay-Newcombe, “Effect of post-D+-irradiation time delay and pre-TDS heating on D retention in single crystal tungsten”, J. Nucl. Mater. 359 (2006) 8–16. [22] M. Poon, R.G. Macaulay-Newcombe, J.W. Davis, A.A. Haasz, “Effects of background gas impurities during D+ irradiation on D trapping in single crystal tungsten”, J. Nucl. Mater. 337-339 (2005) 629-633. [23] V.Kh. Alimov, K. Ertl, J. Roth, “Deuterium Retention and Lattice Damage in Tungsten Irradiated with D ions”, Physica Scripta. T94 (2001) 34-42. [24] A.A. Haasz, M. Poon, J.W. Davis, “The effect of ion damage on deuterium trapping in tungsten”, J. Nucl. Mater. 266-269 (1999) 520-525. [25] A.A. Haasz, J.W. Davis, M. Poon, R.G. Macaulay-Newcombe, “Deuterium retention in tungsten for fusion use”, J. Nucl. Mater. 258-263 (1998) 889-895. [26] O.V. Ogorodnikova, J. Roth, M. Mayer, “Deuterium retention in tungsten in dependence of the surface conditions”, J. Nucl. Mater. 313-316 (2003) 469-477. [27] O.V. Ogorodnikova, J. Roth, M. Mayer, “Pre-implantation and pre-annealing effects on deuterium retention in tungsten”, J. Nucl. Mater. 373 (2008) 254-258. [28] V. Kh. Alimov, private communication, Institute of Physical Chemistry of the Russian
34
Academy of Sciences, Moscow, Russia, 2008. [29] M. Poon, A.A. Haasz, J.W. Davis, R.G. Macaulay-Newcombe, “Impurity effects and temperature dependence of D retention in single crystal tungsten”, J. Nucl. Mater. 313-316 (2003) 199.
[30] R.P. Doerner, private commucation, University of California at San Diego, La Jolla, CA, USA, 2008. [31] W.R. Wampler, R. Doerner, L.-N. Luo, “The effect of displacement damage on deuterium retention in plasma exposed tungsten”, the 9th International Workshop on Hydrogen Isotopes in Fusion Reactor Materials, Salamanca, Spain, 2008. [32] R.D. Kolasinski, private communication, Hydrogen & Metall. Sci. Dept., Sandia Nat. Labs., Livermore, CA, USA, 2009. [33] G. Maddaluno, C. Alessandrini, G. Giacomi, L. Verdini, “Thermal desorption measurement of deuterium retention in fusion relavant materials”, Associazione EURATOM-ENEA sulla Fusione, Centro Ricerche Frascati, Frascati, Rome, Italy, 2008. [34] G. Wright, private communication, University of Wisconsin at Madison, 2009. [35] R. Causey, K. Wilson, T. Venhaus, W. R. Wampler, “Tritium retention in tungsten exposed to intense fluxes of 100 eV tritons”, J. Nucl. Mater. 266-269 (1999) 467. [36] V.Kh. Alimov, “Surface modification and deuterium retention in tungsten and molybdenum exposed to low-energy, high-flux deuterium plasmas”, in the 1st International Conference on New Materials for Extreme Environments, San Sebastián, SPAIN, 2008. [37] V. Philipps, private communication, the ITPA SOL/DIV Avila, Spain, 2008. [38] V.Kh. Alimov and J. Roth, “Hydrogen isotope retention in plasma-facing materials: review of recent experimental results”, Phys. Scr. T128 (2007) 6–13. [39] G.-N. Luo, W.M. Shu and M. Nishi, “Influence of blistering on deuterium retention in tungsten irradiated by high flux deuterium 10-100eV plasmas”, Fusion Eng. Des. 81 (2006) 957-962. [40] R. Causey, K. Wilson, T. Venhaus, W.R. Wampler, “Tritium retention in tungsten exposed to intense fluxes of 100 eV tritons”, J. Nucl. Mater. 266–269 (1999) 467. [41] T. Venhaus, R. Causey, R. Doerner, T. Abeln, “Behavior of tungsten exposed to high fluences of low energy hydrogen isotopes”, J. Nucl. Mater. 290–293 (2001) 505. [42] K. Tokunaga, M.J. Baldwin, R.P. Doerner, N. Noda et al., “Blister formation and deuterium retention on tungsten exposed to low energy and high flux deuterium plasma”, J. Nucl. Mater. 337–339 (2005) 887.
35
[43] M. Poon, R.G. Macaulay-Newcombe, J.W. Davis, A.A. Haasz, “Flux dependence of deuterium retention in single crystal tungsten”, J. Nucl. Mater. 307–311 (2002) 723–728. [44] A. Quastel, “Effects of impurities on deuterium retention in single crystal tungsten”, M.A.Sc. thesis, University of Toronto (2005). [45] M. Poon, “Deuterium trapping in tunsten”, Ph.D. thesis, University of Toronto (2004). [46] L. Reimer, “Scanning electron microscopy: physics of image formation and microanalysis”, Berlin; New York: Springer, c1998. 2nd completely rev. and updated ed. [47] A. van Veen, “Vacancies and interactions in metals and alloys”, C. Abromeit, H. Wollenberger (Eds.) (Berlin 1986), 3. [48] R.A. Anderl, R.J. Pawelko, S.T. Schuetz, “Deuterium retention in W, W1%La, C-coated W and W2C”, J. Nucl. Mater. 290-293 (2001) 38-41. [49] A.V. Golubeva, M. Mayer, J. Roth, V.A. Kurnaev, O.V. Ogorodnikova, “Deuterium retention in rhenium-doped tungsten”, J. Nucl. Mater. 363–365 (2007) 893–897.
36
Figures
Figure 2-1: Depth profiles of D trapped as D atoms (a) and D2 molecules (b) in single-crystal and hot-rolled W implanted with 6 keV D ions at 300 K determined by the SIMS/RGA method (Figures from Ref. [23]).
37
Figure 2-2: Retained vs. cumulative-fluence for 1 keV/D+ implantations at 500 K. Data are shown for specimens W2 (1023 D/m2 probe-fluence only), W1 (9×1023 D/m2 damage-fluence and probe-fluence), and W3 (1025 D/m2 damage-fluence and probe-fluence). (Figure from Ref. [24].)
Figure 2-3: Retained vs. cumulative-fluence for 500 eV/D+ implantations at 500 K. Data are shown for specimens W5 (1023 D/m2 probe-fluence only), W4 (9×1023 D/m2 damage-fluence and probe-fluence), W6 (3×1024 D/m2 damage-fluence and probe-fluence), W7 (1025 D/m2 damage-fluence and probe-fluence), and W9 (3×1025 D/m2 damage-fluence and probe-fluence). (Figure from Ref. [24].)
38
Figure 2-4: Fluence dependence of D retention in PCW at 300 K under various D ion energies. [25-28]
39
Figure 2-5: NRA measurements of the near-surface D depth profiles. (a) 1 keV and 500 eV D+ (1024 D+/m2 incident fluence) implanted into W at 300 K. Implantation profiles for 1 keV D+ and 500 eV D+ as calculated by TRVMC are shown for comparison (normalized to the peak height of the measured profiles). (b) 500 eV D+ implanted at 500 K into W (1024 D+/m2) and W-1% La2O3 (3.3×1024 D+/m2). (Figures from Ref. [25].)
40
Figure 2-6: Fluence dependence of D retention in W at elevated temperatures using ion beams. [25,27,29]
41
Figure 2-7: Fluence dependence of D retention in W at elevated temperatures using plasma devices and tokamaks. [30-37]
42
Figure 2-8: Temperature dependence of D retention in W and W-1%La2O3. (a) 1 keV/D+ at fluences of 1023 and 1024 D+/m2, (b) 500 eV/D+ at fluence of 1023 D+/m2. (Figure from Ref. [29].)
43
Figure 2-9: Temperature dependence of D retention in M-SCW with an incident fluence of 1024 D+/m2. (Figure from Ref. [29].)
44
Figure 2-10: (a) Deuterium retention in single-crystal and polycrystalline fine-grain tungsten exposed to low-energy (200 eV/D+) and high flux (about 1×1021 D/m2s) D plasmas as a function of exposure temperature. For comparison, the temperature dependence of the D retention in polycrystalline coarse-grained W irradiated with 200 eV D ions and flux of 4×1019 D m−2 s−1 to a fluence of 1×1024 Dm−2 is also shown. Note that the deuterium retention was calculated from deuterium depth profiles measured up to a depth of 7μm. (b) Deuterium retention in polycrystalline tungsten exposed to low-energy (98–100 eV/DT) and high flux ((8.7–10)×1021 D(T)m−2 s−1) D or (D+ T) plasmas as a function of the exposure temperatures. [38-42]
45
Figure 2-11: Deuterium retention as a function of incident D+ flux at three fluences (1021, 1022, and 1023 D+/m2) at room temperature. (Figure from Ref. [43].)
46
Figure 3-1: Schematic of single-beam ion accelerator. (Figure from Ref. [44].)
47
Figure 3-2: Implantation specimen holder.
Figure 3-3: Schematic of the TDS system.
48
Figure 4-1: SEM images of irradiated Plansee PCW (after TDS test). (a) On-spot area; (b) Off-spot area.
49
Figure 4-2: SEM images of irradiated Rembar PCW (after TDS test). (a) On-spot area; (b) Off-spot area.
50
Figure 4-3: Cross-sectional SEM photograph of Plansee PCW specimen.
51
Figure 4-4: Signals of H2, HD, and D2 in two thermal desorption runs. (a) 200 eV/D+, 500 K, Plansee PCW; (b) 200 eV/D+, 300 K, Plansee PCW.
52
Figure 4-5: Example calculation of deuterium retention in Plansee PCW (200 eV/D+, 500 K). The vertical bar indicates the estimated HD contribution (i.e., integration of HD signals over the two different time spans).
53
Figure 5-1: 500 eV/D+ ion implantation on PCW at 300 K. [25,26]
Figure 5-2: 200 eV/D+ ion implantation on PCW at 300 K. [27,28]
54
Figure 5-3: Deuterium retention in different types of PCW at 300 K. [25]
Figure 5-4: Fluence dependence of D retention in PCW at elevated temperatures [25,27,30-32, 36]. (The red circle indicates the highest D+ fluence achieved in the present study.)
55
Figure 5-5: Collection of fluence dependence data on D retention in tungsten at room temperature. [17,25-29,49]
56
Figure 5-6: Fluence dependence of D retention in Plansee PCW for different energy and temperature combinations.
57
Figure 5-7: Collection of fluence dependence data on D retention in tungsten at elevated temperatures [25,27,31-37]. (The red circle indicates the highest D+ fluence achieved in the present study.)
58
Figure 5-8: Irradiation temperature effects on D retention in PCW. [25,29,38] (The specimen of the present data point (200 eV/D+, 450 K, 1×1023 D+/m2, 8×1020 D/m2) was annealed at 1500 K.)
59
Figure 5-9: Incident ion energy dependence of D retention in Plansee PCW.
Figure 5-10: Energy dependence of D retention in Rembar PCW. [25]