Deuterium Retention Studies in Self-ion Damaged Tungsten
Exposed to Neutral Atoms
Sabina Markelj, Anže Založnik, Mitja Kelemen, Primož Vavpetič, Primož Pelicon
Jozef Stefan Institute
Jamova cesta 39
1000, Ljubljana, Slovenia
[email protected], [email protected], [email protected], [email protected],
[email protected]
Thomas Schwarz-Selinger
Max-Planck-Institut für Plasmaphysik
Boltzmannstrasse 2
D-85748 Garching, Germany
[email protected]
Olga V. Ogorodnikova
National Research Nuclear University- MEPhI
Kashirskoe sh. 31
Moscow, Russia
[email protected]
Łukasz Ciupiński and Justyna Grzonka
Warsaw University of Technology
Woloska 141
02-507 Warsaw, Poland
[email protected]
ABSTRACT
Deuterium retention in self-ion damaged tungsten exposed to D atoms was studied
experimentally. In one experimental series D atom loading was done at different temperatures
and in the second series samples were loaded at the same temperature but pre-annealed at
different temperatures. Deuterium depth profile in treated samples was determined by Nuclear
Reaction Analysis and the effect of sample exposure temperature and damage recovery by
sample annealing on deuterium retention was observed. It was found that both exposure and
annealing temperature have strong influence on the maximum deuterium concentration and on
the integrated amount of deuterium in the sample. In both cases the concentration is
decreasing with increasing of the temperature. In case of defect annealing the deuterium
retention is reduced by 60 % at 1200 K as compared to the un-annealed sample. The D
retention decreased by 98 % when comparing D atom exposure at 500 K to exposure at 900
K. For the deuterium exposure at elevated temperatures two processes are active, defect
annealing and deuterium de-trapping from traps produced by the self-damaging. We find that
the second one, thermal de-trapping, dominantly contributes to the reduced retention at
elevated temperatures.
701.1
701.2
1
INTRODUCTION
The influence of irradiation by neutrons produced by the fusion D-T reaction has to be
taken into account in order to predict tritium retention in future fusion reactors (ITER and
DEMO). In that respect, successful neutron resistant materials or self-healing procedures need
to be developed. Present indications show that tungsten is the most suitable ‘baseline’
material as plasma facing component armour (from the EFDA-roadmap). Since neutron
irradiated samples get activated [1] and therefore need special handling, high energy ions are
used as neutron surrogates to produce neutron-like damaged material [2]. It was shown that
fuel retention in damaged tungsten is strongly increased as compared to undamaged tungsten
material and for this reason influence of exposure temperature and damage annealing is being
extensively studied [3]-5].
In order to study the deuterium retention in self-damaged tungsten, the samples for our
experiment were damaged by 20 MeV tungsten ions at room temperature. Such irradiation
results in a 2.4 μm deep damaged layer. It was shown in [6] that self-damaging is a good
surrogate for neutron damage. Namely, 14 MeV neutrons and 20 MeV W ions have similar
spectrum of the primary knock-on atoms and thus creating large collision cascades [7].
However, due to different particle interaction – short range nuclear force for the neutron case
and long-range Coulomb force for the ion case – the mean free path of ions is shorter.
Therefore ions create damage only few μm deep, where as in the neutron case the damage is
spread over the entire depth of the sample. In order to study the influence of the damage on
fuel retention samples are in this experiment exposed to deuterium atoms of 0.3 eV thermal
energy. Atoms with such low energy have small penetration depth and are mainly adsorbed
on the surface or in the subsurface before they can diffuse into the bulk. The advantage of the
atom exposure is that the atoms are trapped at the defects that act as additional strong binding
sites for deuterium atoms in the damage zone, without producing any additional damage.
2
EXPERIMENT
We have performed two kinds of experiments with different exposure and sample
annealing procedure. In both cases we have used polycrystalline 99.997 wt. % W samples
manufactures by Plansee, 12x15 mm2 in size and 0.8 mm thick. Samples were chemomechanically polished to a mirror like finish at Max-Planck-Institut für Plasmaphysik (IPP),
Garching, Germany.
Samples that we have used in the first series of experiments, i.e. D atom loading at
different temperatures, were made of forged material, the so-called ITER grade tungsten (IW,
d-IW for damaged sample). In these samples grains are elongated perpendicular to the sample
surface for better thermal conductivity. Sample grain size was 5-30 μm [3]. After polishing
they were outgassed at 1200 K for stress relieve.
For the second series of experiments - damage annealing and subsequent D atom
loading - we have used Plansee hot-rolled tungsten samples (PW) having grains oriented
parallel to the surface normal. After polishing, samples were heated for 2 min at 2000 K for
re-crystallization. This procedure enlarges the grain size to 10-50 μm.
Both sets of samples were irradiated at room temperature by W ions in the tandem
accelerator laboratory, TOF beam-line at IPP Garching. In the case of d-IW samples a
homogeneous damage profile of 0.89 dpa down to 2.5 μm was made, using different ion
energies of 20, 8, 4 and 2 MeV with fluences of 1.4×1018 Wm-2, 3.06×1017 Wm-2, 1.97×1017
Wm-2 and 1.38×1017 Wm-2 respectively (SRIM calculation – 90 eV displacement energy, 3 eV
Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, September 14 ̶ 17, 2015
701.3
lattice binding energy, full cascade mode). However, according to recommendations by
Stroller et al. [8] the proper damaging level corresponding to fluence of 1.4×1018 Wm-2 is 0.45
dpa [3]. Due to the simplicity of comparison to previous references, we will refer to the old
damage level calculations throughout the paper.
Since it was shown [5] that increasing the damage level above 0.5 dpa does not increase
the deuterium retention, we have decreased the W ion fluence for the second experiment. The
PW samples were damaged by single energy of 20 MeV W6+ to the fluence of 7.8×1017 m-2,
creating a peak damage level of 0.5 dpa (for the same SRIM calculation procedure). For the
present study the different damaging procedure is of little importance. As shown in [1] the
same radiation-induced defects are created in both samples obtaining the same microstructure.
One way to investigate the created defects is by using the Scanning Transmission
Electron Microscopy (STEM) [10]. A STEM figure of a damaged PW sample is shown in
Fig. 1. One can observe that even though the sample was damaged by single energy 20 MeV
W ions, there is a rather uniform distribution of the created damage down to 2 μm. The
damage can be attributed to large tangled dislocation network being composed of dislocations
and dislocation loops. Each dislocation loop is related to interstitial atoms or vacancy clusters
being an edge of an extra or missing part of crystallographic plane. From 2 μm to 2.5 μm the
damaged zone is composed of small dislocation loops densely distributed and beyond that
there is an undamaged zone where almost no defects are visible.
2 μm
Figure 1: STEM figure of d-PW sample, damaged by 20 MeV W ions
with fluence 7.8×1017 m-2.
Self-ion damaged samples were mounted on a holder with temperature controlled
sample heater and exposed to a well characterised beam of thermal deuterium atoms (~0.3 eV
thermal energy). The D atoms are produced by flowing hydrogen gas through a hot tungsten
capillary at 2100 K. The atom beam profile was determined by erosion of amorphous
hydrogenated carbon film [11, 12], yielding maximum atom flux of 3.5×1019 D/m2s. In the
first experiment the sample exposure was performed at different temperatures and different
exposure times, sufficient to saturate the damage induced by self-damaging. In the second
experiment the samples were first annealed for one hour at different temperatures and then
after cooling to room temperature, heated to 500 K and exposed for 144 h to D atom flux of
approximately 2.6×1019 D/m2s, all in the same vacuum chamber. The resulting D atom
fluence of 1.3×1025 D/m2s was sufficient to saturate the damaged layer at 500 K.
The deuterium concentration depth profile was measured after D atom loading ex-situ
by Nuclear Reaction Analysis (NRA) utilizing the D(3He,α)p nuclear reaction [13]. In order to
obtain a depth profile down to 7 μm we have recorded proton energy spectra obtained with
five different 3He energies from 690 keV to 4000 keV for d-IW samples. NRA was performed
Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, September 14 ̶ 17, 2015
701.4
at IPP. In the case of d-PW samples the NRA analysis was performed at JSI, also with five
different energies from 780 keV to 4300 keV.
2.1
Sample Exposure at Different Temperatures
First, deuterium retention was studied on d-IW damaged sample after long-term D
exposure at different sample temperatures. Samples were exposed to a constant flux of
3.5×1019 D/m2s at sample temperatures from 500 K to 900 K. The obtained depth profiles are
shown in Fig. 2. Samples were subjected to the same atom fluence of 3.8×1023 D/m2. As can
be observed the depth profiles have the highest concentration from surface to about 2 μm deep
and then the concentration decreases. This means that fluence of Γ = 3.8×1023 D/m2 was
sufficient to saturate the damaged layer by deuterium atoms, except for the exposure at 600 K.
The maximum concentration decreases with increasing temperature from 0.3 atomic % for
600 K down to 0.004 atomic % for exposure at 900 K. In other words, retention decreases
almost by two orders of magnitude with 300 K temperature increase. The decrease at higher
temperature can be attributed to two effects: first, thermal de-trapping of D atoms from the
high energy traps created by damaging and second, defect annealing that also takes place
during the exposure.
Figure 2: Deuterium depth profiles at different exposure temperatures for the same D fluence
on d-IW samples.
In order to saturate the whole damaged layer at 600 K a higher fluence was needed.
Comparison of the depth profile obtained for two different fluences 3.8×1023 D/m2 and
2.0×1024 D/m2 is shown in Fig. 3. As can be observed the deuterium first saturates the traps
near the surface and once they are saturated D atoms slowly proceed deeper into the layers. In
Fig. 3 we also show the damaged depth profile obtained for the multiple energy irradiation
and for single energy irradiation by 20 MeV ions. The same effect was observed for 500 K
where even a fluence of 8.0×1024 D/m2 was not enough to completely saturate the damaged
layer in the last 1 μm of the damaged layer.
Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, September 14 ̶ 17, 2015
701.5
Figure 3: Deuterium depth profile at different atom fluences at 600 K on d-IW samples. The
calculated damage depth profiles for the two cases are also shown: flat damage profile by
dashed line and dotted profile for irradiation by single energy of 20 MeV W ions.
Concentration measured in undamaged sample under indicated exposure conditions is also
shown for comparison.
2.2
Sample Pre-annealing and D Atom Loading at 500 K
In the second experiment d-PW samples were first annealed in vacuum for one hour at
different temperatures (600 K – 1200 K) and then exposed to D atoms at 500 K to a fluence of
Γ = 1.3×1025 D/m2. The depth profiles obtained by NRA are shown in Fig. 4a. There is a
decrease in maximum concentration, from 0.42 atomic % for the un-annealed sample to 0.12
atomic % for the sample annealed at 1200 K. The maximum concentration has decreased for
3.5 times, being in good agreement with a similar study where samples were loaded by
plasma [3]. In Fig. 4b a comparison of these depth profiles to the depth profile obtained for an
undamaged W sample is shown, exposed at the same temperature and fluence. One can
clearly see that although annealing to 1200 K reduced the damage level significantly retention
is still more than an order of magnitude larger as compared to the undamaged case.
Samples that underwent the same preparation, damaging and annealing procedure were
also analysed by STEM. It was observed that with increase of the annealing temperature
larger dislocation loops can be observed but with reduced dislocation density [14]. The
dislocation density dropped by 66 % for the highest annealing temperature as compared to the
un-annealed sample. This is in good agreement with the total D amount decrease shown in Fig
5b.
Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, September 14 ̶ 17, 2015
701.6
(a)
(b)
Figure 4: a) Deuterium depth profile obtained for samples first annealed for 1 hour at the
indicated temperatures and then exposed to D atoms at 500 K. b) The same depth profile as
above but in logarithmic scale together with undamaged reference sample.
3
DISCUSSION
We have performed two experiments where in one case the samples were loaded at
different temperatures and in the second case they were loaded at the same temperature but
pre-annealed at different temperatures. A comparison of the maximum concentration in the
damaged area for the two experimental procedures is shown in Fig 5a. One can observe that
for the sample exposure at different temperatures marked as “Exposure” in Fig 5, the
maximum concentration decreases by a factor of 100 when going from 500 K to 900 K. In the
case of annealing at different temperatures and loading at 500 K, marked as “Pre-annealing”
in Fig. 5, the decrease is only by a factor of 3.5 when going from 500 K to 1200 K. Even
though the damaging was performed at room temperature, the data point corresponding to
“un-annealed” case in Fig. 4, is shown at 500 K due to D atom loading at this temperature. It
was shown in [4] that there is already some change in maximum D concentration and total D
amount when increasing the temperature of annealing from 400 K to 550 K.
Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, September 14 ̶ 17, 2015
701.7
(b)
(a)
Figure 5: Comparison of: a) deuterium maximum concentration in the damaged area in
logarithmic scale and b) total deuterium amount in analysing depth of 7 μm as a function of
exposure or annealing temperature for the two experiments in linear scale.
A similar trend is observed also in Fig 5b, where the comparison of the total amounts
within 7 μm is shown for the two cases. In the case of annealing a 60 % decrease in the total
deuterium amount was observed when annealing from 500 K to 1200 K. In the case of
exposure at different temperatures, the total amount decreases by 98 % from 500 K to 900 K.
In the case of exposure at 900 K the decrease is not so drastic as compared to the change
between 700 and 800 K, since there is a background level of D being trapped in natural traps
beyond the damaged region. Moreover, there are two data points for the exposure at 500 K.
Namely, as mentioned above, in the d-IW case the last 1 μm in the damaged layer was not
completely saturated for atom fluence of 8.0×1024 D/m2, when comparing to the depth profile
obtained in the annealing study on “un-annealed” sample. The lower data point corresponds to
the “unsaturated” d-IW sample and the higher point is from the annealing study on d-PW
sample. However, a good agreement in the maximum concentrations between “Exposure” and
“Pre-annealing” was obtained at 500 K, as shown in Fig 5a. From the thermal desorption
study conducted on these samples for cases: the exposure study [15] and pre-annealing study
[14] the deuterium is completely desorbed at 1200 K. Therefore the dominating process that
contributed to the drastic reduction of the total amount when increasing the exposure
temperature is the deuterium de-trapping from the defects induced by the W ion irradiation.
On the other hand partial recovery of the defects was observed in the annealing study.
However, the deuterium retention for sample annealed at 1200 K is still substantial, 20.6×1019
D/m2, as compared to the undamaged sample obtaining total D amount of 1×1019 D/m2 in the
analysing depth of 7 μm.
4
CONCLUSION
Deuterium retention was studied experimentally on self-ion damaged W as a function of
D atom loading temperature and defect pre-annealing temperature W. Such an experimental
procedure enabled us to separate the effect of damage annealing from thermal de-trapping.
We have observed both a reduction of the maximum D-concentration in the damaged zone
and the total deuterium amount within damaged layer in both cases with increase of the
temperature. However the decrease is much more pronounced in a narrower temperature
window for the exposure at higher temperatures. From this study one can expect that elevated
temperatures of the walls of the reactor will be beneficial in terms of fuel retention. This is
especially important when comparing to neutron damage where damage level will be similar
to our case (about 1 dpa in ITER) but spread over the entire depth of the targets. In such a
Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, September 14 ̶ 17, 2015
701.8
case diffusion of D atoms through damaged material is very important. As it was shown the
diffusion is much slower at lower temperatures meaning that longer times and fluences will be
needed to saturate the damage layer. Therefore both processes are important, deuterium
trapping energy and diffusion. There is another aspect that was not considered in this study,
namely, in the reactor both implantation of D ions and neutrals as well as neutron irradiation
will take place at the same time. It was observed in some materials that impurities such as
hydrogen, change the behaviour of defect creation and recovery, e.g. on vacancy migration
during the recovery stage [16]. For this reason we have started in our laboratory at JSI an
experiment where both W ion irradiation and D atom loading takes place simultaneously.
ACKNOWLEDGMENTS
This work has been carried out within the framework of the EUROfusion Consortium
and has received funding from the Euratom research and training programme 2014-2018
under grant agreement No. 633053. Work was performed under WP PFC. The views and
opinions expressed herein do not necessarily reflect those of the European Commission.
Research Project was part of the CRP F43021 carried out under the sponsorship of the IAEA.
The authors acknowledge Dr. Iztok Čadež for constructive and valuable discussions.
REFERENCES
[1]
Y. Hatano, M. Shimada, T. Otsuka, Y. Oya, et al., Deuterium trapping at defects created
with neutron and ion irradiations in tungsten. Nucl. Fusion 53, 2013, 073006.
[2]
W.R. Wampler, R.P. Doerner, The influence of displacement damage on deuterium
retention in tungsten exposed to plasma. Nucl. Fusion 49, 2009, 115023.
[3]
O.V. Ogorodnikova, Y. Gasparyan, V. Efimov, Ł. Ciupiński, J. Grzonka. Annealing of
radiation-induced damage in tungsten under and after irradiation with 20 MeV self-ions.
J. Nucl. Mater. 451, 2014, pp. 379–386.
[4]
E. Markina, M. Mayer, A. Manhard, T. Schwarz-Selinger,. Recovery temperatures of
defects in tungsten created by self-implantation. J. Nucl. Mater. 463, 2015, pp. 329–
332.
[5]
M.H.J. ’t Hoen, M. Mayer, A.W. Kleyn, H. Schut, P.A.Z. van Emmichoven, Reduced
deuterium retention in self-damaged tungsten exposed to high-flux plasmas at high
surface temperatures, Nucl. Fusion 53, 2013, 043003.
[6]
O.V. Ogorodnikova, V. Gann. Simulation of neutron-induced damage in tungsten by
irradiation with energetic self-ions. J. Nucl. Mater. 460, 2015, pp. 60–71.
[7]
P. Vladimirov and S. Bouffard, Displacement damage and transmutations in metals
under neutron and proton irradiation, C. R. Physique 9, 2008. pp. 303-322.
[8]
R.E. Stoller, M.B. Toloczko, G.S. Was, A.G. Certain, S. Dwaraknath, F.A. Garner, On
the use of SRIM for computing radiation damage exposure, Nuclear Nucl. Instr. and
Meth. in Phys. Res. B 310, 2013, pp. 75-80.
[9]
V.K. Alimov, Y. Hatano, B. Tyburska-Püschel, K. Sugiyama, I. Takagi, Y. Furuta, J.
Dorner, M. Fußeder, K. Isobe, T. Yamanishi, M. Matsuyama, Deuterium retention in
Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, September 14 ̶ 17, 2015
701.9
tungsten damaged with W ions to various damage levels., J. Nucl. Mater. 441, 2013, pp.
280–285.
[10] L. Ciupinski, O. V. Ogorodnikova, T. Plocinski, M. Andrzejczuk, M. Rasinksi, M
Mayer, K.L. Kurzydlowski, TEM observations of radiation damage in tungsten
irradiated by 20 MeV W ions, Nucl. Instr. and Meth. in Phys. Res. B 317, 2013, pp.
159-164
[11] T., Schwarz-Selinger, A., Von Keudell, W., Jacob,. Novel method for absolute
quantification of the flux and angular distribution of a radical source for atomic
hydrogen. Journal of Vacuum Science & Technology A 18, 2000, pp. 995–1001.
[12] S. Markelj, P. Pelicon, I. Čadež, T. Schwarz-Selinger and W. Jacob, In situ study of
erosion and deposition of amorphous hydrogenated carbon films by exposure to a
hydrogen atom beam, J.Vac. Sci. Tech. A 30, 2012, pp. 041601-1.
[13] V.Kh. Alimov, M. Mayer and J. Roth, Differential cross-section of the D(3He, p)4He
nuclear reaction and depth profiling of deuterium up to large depths, Nucl. Instr. and
Meth. in Phys. Res. B 234, 2005, pp. 169–175.
[14] A. Zaloznik, S. Markelj, T. Schwarz-Selinger, Ł. Ciupiński, J. Grzonka, P. Vavpetič, P.
Pelicon, The influence of the annealing temperature on deuterium retention in selfdamaged tungsten, PFMC 15 proceedings submitted to Phys. Scripta
[15] Y.M. Gasparyan, O.V. Ogorodnikova, V.S. Efimov, A. Mednikov, E.D. Marenkov,
A.A. Pisarev, S. Markelj, I. Čadež, 2015, Thermal desorption from self-damaged
tungsten exposed to deuterium atoms, J. Nucl. Mater. 463, 2014, pp. 1013–1016.
[16] H. Schultz, Defect parameters of b.c.c metals: group-specific trends, Mater. Sci. And
Eng. A141, 1991, pp. 149-167.
Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, September 14 ̶ 17, 2015