Thermal kinetics of micro‐defects in He‐ion implanted W and W5Re alloys

To investigate the thermal evolution of vacancy-type defects in He-ion irradiated W and W5Re alloy, different isochronal annealing treatments from 373 to 1273 K were conducted on the irradiated materials. Positron annihilation spectroscopy including positron annihilation lifetime spectroscopy and Doppler broadening spectroscopy were mainly used to characterize the micro-defects evolution. The results showed that the thermal evolution characteristics of defects in both W and W5Re were similar. After He-ion irradiation, mono-vacancies with positron annihilation lifetime of ~ 190 ps were detected in W, together with a large amount of dislocation loops with positron annihilation lifetime of ~ 150 ps in W5Re alloys. The coarsening of vacancy clusters at the expense of small vacancy clusters was the main thermal evolution feature of vacancy-type defects in both W and W5Re when annealing temperature increased to 1073 K. In this progress, the positron annihilation lifetime increased to ~ 350 ps (clusters composed of 4 –8 mono-vacancies) in both W and W5Re. As the temperature increased to 1273 K, the positron annihilation lifetime decreased to ~ 240 ps, which was attributed to a significant population reduction of the dislocation loops, the dissociation of large HenVm complexes and the annealing of micro-voids in both W and W5Re. The vacancy-type defects in W5Re were more susceptible to the annealing temperature because of the formation of vacancy cluster-Re complexes. Re clusters in irradiated W5Re alloy could serve as the nucleation sites of He bubbles, which promoted the swelling and protrusion formation on the surface.


Introduction
Tungsten (W) is widely considered as the candidate for plasma facing materials (PFMs) in fusion reactor [1][2][3]. However, under the deuterium-tritium fusion reaction, PFMs-W has to withstand tremendous damages produced by high energy neutron (~ 14 MeV) and high thermal flux (~ 10 MW·m −2 ). Besides, a continuous production of helium (He) through (n, α) transmutation reactions in the bulk of PFMs-W are easy to be captured by irradiation-induced defects (such as vacancies, clusters, voids and dislocations) to form He-defect complexes even He bubbles during the diffusion process, leading to the degradation of the material mechanical properties including irradiation hardening and swelling [4,5]. Therefore, it is significant to understand the physical fundamental behind these changes introduced by fusion neutrons. However, the enough high flux of 14 MeVneutron irradiation is not available currently. Instead of the neutron irradiation, multiple He-ion irradiation has become one of the main methods to study the irradiation damage and He behavior in PFMs simultaneously. The fusion neutrons also transform PFMs-W into transmutation rhenium (Re) element through neutron capture reactions. It is reported by many researchers that the accumulation of Re significantly threatens the lifetime of PFMs-W because of the deterioration of mechanical properties [4,[6][7][8][9]. In current study, the irradiation effect caused by 5 wt% Re addition in W is investigated compared to pure W, which is considered as a possible guideline for further research on the application of W-based alloys in fusion field.
High resolution transmission electron microscopy (TEM) is preferred by many researchers to investigate the structure of neutron/ion irradiation-induced micro-defects [5,10]. However, it is difficult to detect the defect with size smaller than 1 nm by TEM. The distribution and dynamic of vacancy-type defects in the implantation depth become more significant in structural material characterization, which makes it necessary to utilize the positron annihilation spectroscopy (PAS). PAS is an effective and convenient technique to study the irradiation-induced vacancy-type defects, such as vacancy, vacancy cluster, vacancy-type dislocation loops as well as vacancy-impurity complex [11][12][13][14]. Positron is exclusively sensitive to vacancies, providing information of the size, density and depth distribution of vacancy-type defects [3,15,16]. Sato et al. [17] took the PAS technique to study the change of the hydrogen decorated vacancies in Ogorodnikova et al. [18] applied the PAS to investigate W exposed to 21.6 MeV protons and found that proton irradiation mainly induced the mono-vacancies. Mono-vacancies started to migrate at ~ 473 K in He-irradiated W [13]. High temperature promoted the vacancy clustering along the three-dimensional direction and the formation of micro-voids. The collapse of some micro-voids leads to the production of vacancy-type dislocation loops which begin to anneal at ~ 1023 K. These evaluation progresses resulted in a change of the positron annihilation lifetime [13,19]. Based on the mechanism of positron capture and annihilation, the types of defects and the chemical information of elemental changes could be explored, especially the basic information of the element-defects complex structures.

Experimental
Polycrystalline W (99.95 wt%) and W5Re alloys (W-5 wt%Re) were supplied by Advanced Technology & Materials Co., Ltd. (AT&M). After mechanical and electrochemical polishing, all samples were thermally annealed at 1373 K with the duration of 2 h in vacuum (~ 10 -5 Pa).
He-ion irradiation at room temperature (RT) was performed on a 200 kV ion implanter. More details of He-ion irradiation are shown in Table 1. The sequential He-ion irradiation  of 16 keV, 70 keV and 200 keV on W-3 and W5Re-3 was determined to simulate the neutron average damage distribution in bulk layer of 0-500 nm. He-ion flux ranged from (0.42-1.72) × 10 13 cm −2 ·s −1 , which made the displacement damage up to 0.83 dpa (displacement per atom) for W-1 and W5Re-1, 1.62 dpa for W-2 and W5Re-2, and ~ 1.45 dpa for W-3 and W5Re-3. Isochronal (1 h, ~ 5.5 × 10 -5 Pa) thermal annealing was performed on all irradiated samples at 373 K, 573 K, 673 K, 773 K, 873 K, 1073 K and 1273 K. Following each annealing process, the samples were naturally cooled down and characterized by positron annihilation lifetime spectroscopy (PALS) and Doppler broadening spectroscopy (DBS) at RT.

Characterization methods
As a sensitive probe to vacancies, positron can be effectively captured by vacancy-type defects. In the process of annihilation with electrons, γ rays carrying the information of microstructure and chemical environment at the position annihilation site are produced [19][20][21][22][23]. The time interval between the production of the 0.511 MeV-γ-ray and the emission of the 1.28 MeV-γ-ray is regarded as the positron annihilation lifetime, which is closely related to the electron density where the annihilation occurs. The radioactivity of the 22 Na source used in PALS measurement is ~ 0.185 MBq and the time resolution of associated BaF 2 detector is ~ 195 ps. The source is encapsulated by Kapton films with thickness of ~ 7 μm, which is sandwiched between two samples to be detected.
The DBS based on low positron beam with a positron source of 22 Na (45 mCi in 2016) were carried on all samples. The fast positrons from source are thermalized by the moderator of multi-layer W nets, and then thermally emit from moderator's surface with a certain kinetic energy (~ 2.1 eV). A negative high voltage is applied to the sample target to realize the continuous adjustment of positron energy in the range of 0.18-20.18 keV. The mean incident depth Z (nm) of low-energy mono-energetic positrons in target follows the empirical equation where E is the positron energy (keV), is the material density (g·cm -3 ). The positron-electron pair annihilation producing γ rays especially those with energy from 505.1-508.4 keV are recorded by a high-purity germanium detector with an energy resolution of 1.2 keV at 511 keV. The number of photons recorded in intermediate-energy window (510.2-511.8 keV) in Doppler spectrum is the annihilation contribution between positron and low-momentum valence electron in target, which determines S parameter. The number of photons recorded in the both two energy windows (503.3-507.2 keV and 514.8-518.7 keV) in Doppler spectrum comes from the annihilation contribution of positron and high-momentum core electron in target, which determines the W parameter [24,25]. The S parameter is sensitive to the open volume defects and W parameter is most probably related to the chemical surrounding at the annihilation sites. The depth dependence of the positron annihilation parameters is denoted by S-E or W-E profiles.

SRIM simulation
Using the stopping and range of ions in matter (SRIM) computer program [26], the He-ion irradiation damage (dpa) and He atom concentration (at.%) as a function of positron penetration depth (nm) were predicted in both W and W5Re, as shown in Fig. 1. "Full Cascade" mode was selected as the calculation type. The displacement threshold energy was 90 eV and the lattice binding energy was 3 eV for both W and W5Re alloy [27]. W5Re plots are not exhibited in Fig. 1

Evolution of the type and associated intensities of the vacancy-type defects
In order to study the thermal evolution of irradiation-induced vacancy-type defect, PALS measurements were carried out on W-1, 2, 3 and W5Re-1, 2, 3. The PALS spectra with a total count of 2 × 10 6 were analyzed by LT-9.0 program [28]. It should be noted that a small fraction (~ 4.57%) of positrons from the 22 Na source (average energy of the emitted positrons is ~260 keV) annihilated in the depth range within 500 nm. The advantage of positron self-searching for vacancy-type defects makes more positrons annihilate in irradiation-induced region. Besides, W has a large (absolute value) positron negative work function (−2.1 eV). Therefore, the experimental results show that some qualitative assessment can be still performed as the vacancytype defects detected by positrons which are expected to be exclusively from the implanted region. The intensity of the defect component is, however, meaningless without a precise assessment of the fraction of signal coming from unirradiated bulk. So the relative variation of intensity would be discussed.
The PALS characterization data of W and W5Re alloys is listed in Tables 2 and 3, respectively. The source contribution in Kapton films was 20% and the lifetime component was 382 ps. The PALS spectra of He-ion irradiated sample were decomposed by three lifetime components: τ 1 , τ 2 and τ 3 . Here, τ 3 is the positron annihilation lifetime of the orthopositronium in sample surface and even in the sample-film interface [29]. The intensity of τ 3 is extremely small (< 1%) and commonly keep constant in PALS characterization for different metals, and therefore τ 3 is not listed in Tables 2  and 3. Positrons are annihilated in free state with electrons in perfect crystal regions, which are easy to be captured by vacancy-type defects (such as vacancies, dislocation loops, and vacancy clusters.) in ion-irradiated materials. τ 1 is the positron annihilation lifetime in the "free" Bloch state which is in general less than the positron annihilation lifetime in perfect lattice. τ 2 is the positron annihilation lifetime at the vacancy-type defect sites, and its variation provides significant information on the kinetics of irradiation-induced defects. I 1 and I 2 are the intensities of defects with τ 1 and τ 2 . τ m is conducted by the equation  [3,30]. As shown in Tables 2 and 3, the free state positron annihilation intensity reached ~ 99% in both well-annealed W and W5Re. τ m is ~ 118 ps for the well-annealed W, which is close to the mean positron annihilation lifetime (~ 120 ps) in defect-free W [31]. The addition of 5 wt% Re element in W prolongs the mean-lifetime to ~ 134 ps in well-annealed W5Re alloys. Figure 2 shows the positron annihilation lifetimes (τ 1 , τ 2 and τ m ) and the intensities (I 2 ) as functions of annealing temperature from RT to 1273 K. After irradiation, an obvious   [32]. However, smaller number of τ 2 (~ 140 ps) with larger intensity is observed in irradiated W5Re alloys than that in W. It is reported that the positron annihilation lifetime at dislocation loops is ~ 150 ps in pure W [13]. The dislocation loops with size larger than 2 nm are observed in neutron-irradiated W [19]. Many literatures have reported that the addition of Re decreased the energy barrier which was account for a high yield of dislocation loops [7,9,33,34]. Thus, it can be referred that a large number of dislocation loops are introduced into W5Re alloy under RT-He-ion irradiation. As the annealing temperature increased to 673 K, both τ 1 and τ 2 are elevated to ~ 120 ps and ~ 350 ps for W and W5Re alloys, respectively, as shown in Fig. 2. Nambissan et al. [13] has reported that an increasing number of positrons were trapped by vacancies when the annealing temperature reached 573 K. In this case, the vacancies may come from the displacement cascade damage and the dissociation of the vacancy-helium complexes. The migration and coalescence of mono-vacancies and di-vacancies directly promote the production of small vacancy clusters. After 673 K annealing process, the decomposed second lifetime component τ 2 increases to ~ 350 ps combined with the sharp decrement of the intensity (I 2 ). The strong population reduction of vacancy clusters is presumably consistent with the growth of large vacancy clusters at the expense of small vacancy clusters. There is a little deviation of the positron annihilation lifetime components in the region from 673 to 1073 K. The τ 2 around 350 ps is the positron annihilation lifetime in the three-dimensional vacancy clusters composed of 4-8 mono-vacancies [35]. High temperature heating promotes the continuous growth of vacancy clusters, leading to the possibly formation of micro-voids with a decreasing intensity which are still the effective trapping site of positron. It was reported that the number of the dislocation loops decreased about one order of magnitude when the annealing temperature increased from 773 to 1073 K [19]. At 1273 K annealing, the τ 2 decreases significantly to ~ 240 ps, which is due to the significant decrease of the dislocation loops in the samples [36]. At this temperature, the number density of the line dislocations is much lower than that of the dislocation loops. In addition, the grain boundaries act as the sinks for vacancies to release themselves below 1073 K. It directly reduces the nucleation points of the micro-voids at 1273 K. Similar conclusions were drawn in Ref. [13] that micro-voids in polycrystalline W tend to be annealed by the grain boundary effect between 1023 and 1323 K.  Fig. 3b and c. This is due to the trapping of He atom at vacancy-type defects and the formation of vacancyhelium complexes [3,10]. A turning point between 1073 and 1273 K is observed in all figures, which will be discussed later. Because of the epithermal positron emissions at the surface, the non-elastically scattered positron and the local changes of the surface dipole, the surface S parameters of W and W5Re are relatively high in both single-or sequentialion irradiation [23,37]. From Fig. 4, the S (versus W) parameter in the matrix of well-annealed W5Re is ~ 0.43 (0.08), which is slightly higher than that of W matrix in Fig. 3, which is due to the different atomic size of Re and W atoms. The atomic volume Fig. 2 Positron annihilation lifetimes (the short-lifetime τ 1 , the longlifetime τ 2 and the mean-lifetime τ m ) and the intensities (I 2 ) as function of annealing temperature from RT to 1273 K of Re is smaller than that of W, leading to a slight lattice distortion (contraction) in W5Re alloys during the irradiation (the atomic volumes are 15.85 × 10 -27 nm 3 for W and 14.70 × 10 -27 nm 3 for Re) [25,38]. The thermal evaluation of vacancy-type defects in W5Re alloys is similar to that in W within the region of 0-27 nm. However, from 27 to 249 nm, the defects in W5Re alloys are more susceptible to the annealing temperature. It is reported that the microstructural development of neutron irradiated W-Re alloys in high flux isotope reactor (HFIR) is suppressed by the accumulation of transmutation solid Re [6]. What's more, irradiation induced vacancies, dislocations or vacancy clusters are likely to act as the strong trapping sites for Re atoms [9,34,39]. Therefore, due to the pinning effect of Re clusters, the Re atom would suppress the growth and migration of vacancies to a certain extent. The increasing thermal evolution temperature may promote the formation of these vacancy clusters-Re complexes, reducing the positron annihilation fraction at the open volume sites.

Depth distribution of vacancy-type defects in thermal annealed W and W5Re alloys
The S-E curves show obvious and regular thermal evaluation characteristics of S parameter in the shallow layer of 0-27 nm and the intermediate layer of 27-249 nm for both W and W5Re. Through the formula of where the ΔS is the net S parameter, S irra is the irradiation induced S parameter and the S non is the well-annealing induced S parameter, the ΔS which weakens the influence of intrinsic defects can be obtained as a function of positron implantation energy (ΔS-E). Figure 5 shows the temperature dependence of average ΔS (ΔS ave ) parameters in the depth regions of 0-27 nm (SIR: shallow implanted region) and 27-249 nm (IIR: intermediate implanted region) of W and W5Re, respectively.
The ΔS ave parameter in SIR presents an obvious zigzag decline compared with that in IIR from RT (post-irradiated) to 1073 K which is defined as "negative slope stage". From 1073-1273 K, the S parameter increases markedly which is defined as "positive slope stage". The negative slope stage is related to the aggregation and growth of vacancy-type defects which is at a large cost of the population of small vacancy clusters. It can be seen from Fig. 5a that the ΔS ave parameters of all samples in SIR increase first and then decrease with the annealing temperature increasing from 573 to 773 K, which are shown by the red dot arrows in Fig. 5a. In contrast, the turning point at 673 K is not obvious in Fig. 3 a-c S-E profiles and d-f W-E profiles of isochronal thermal annealing W-1, W-2 and W-3 at 373 K, 573 K, 673 K, 773 K, 873 K, 1073 K, and 1273 K for 1 h, respectively. For comparison, the S-E and W-E profiles of the well-annealed W and post-irradiated W were included. The top axis was the corresponding positron implantation depth. The shadow region was from 0-27 nm IIR. It can be predicted that the number density and size of vacancy-type defects change more obviously in near surface layer at 673 K. It is reported that the nature of the vacancytype defects changes at 623 K because the clustering of mono-vacancies already occurred during the migration to the surface when the temperature was lower than 623 K [40]. The growth of the intermediate-or large-size clusters at expense of small size clusters were dominated above 673 K [19]. For most samples, the minimum value of S parameter is obtained at 1073 K in both SIR and IIR. As the extremely  low diffusion barrier of He (~ 0.02 eV) in W promotes the diffuse rapidly to the dislocations to form the He clustersdislocation complexes at high temperature of 798-1198 K [41,42]. Especially, the screw dislocation nodal points were reported to be the most probable nucleation sites of He bubbles [43]. All of these could lead to a decrease of the positron trapping probability at 1073 K annealing, even though large-sized vacancy clusters (~ 360 ps) were determined by PALS in this case. The positive slope stage is related to the recovery of small size vacancy-type defects. The S parameter in both SIR and IIR increases sharply at 1273 K annealing, as shown by the black dot arrows in Fig. 5. At the same temperature, the second lifetime component τ 2 decreases to ~ 240 ps which is still larger than 190 ps (mono-vacancies) in both W and W5Re. The recovery of vacancies and small dislocation loops reduces the nucleation points of large vacancy clusters. On the other hand, the dissociation of the unstable He n V m clusters and collapse of micro-voids into small size clusters with high number density may lead to an increase of S parameters. It should be noted that the thermal evolution of vacancy clusters is a competitive consequence of the shrink of the open volume size and the growth of the open volume density [19,44].

Surface morphology of W and W5Re alloys at varied annealing processes
From the previous analysis, it can be inferred that the irradiation-induced defects present obvious thermal evolution characteristics in the range of RT-673 K, 673-1073 K and 1073-1273 K. Figures 6 and 7 show the surface morphologies of W-1, 2, 3 and W5Re-1, 2, 3 after He-ion irradiation and subsequent isochronal thermal annealing at 773 K, 1073 K and 1273 K, respectively. The surface of post-irradiated W is smooth with clear grain boundary, compared with that of post-irradiated W5Re. Evidently, the grain of W5Re alloy is refined due to the addition of Re in matrix.  [45] also found the obvious bulges and protrusions with size of ~ μm which were proved to be blisters on the surface  [46]. These black spots are speculated to be break of the blisters-like protrusion. Many studies [13,40,45,46] reported that He atoms begin to release from the He clusters and dislocation loops near He n V m complexes below 573 K in W. As the temperature is elevated to 773 K, particular vacancy containing the high ratio of He atoms becomes the main site of He desorption. When reaching 973 K, He were mainly released from the He n V m complexes. The desorption temperature increased up to 140 K when formation of stable He n X m V l (X represents the solid foreign element such as Me (mixed element) [47] and Ta (tantalum) [45]) is uniformly distributed in W. The release of He in W and W5Re is correlated to the mutation such as split and rupture of He  bubbles at 1273 K. In addition, the possible Re clusters in W5Re alloy may serve as the nucleation sites of He bubbles, which accelerates the swelling and protrusion formation of the grain boundary and surface. The thermal evolution of defects in He-irradiated W and W5Re alloys are summarized in Table 4, which is divided into three stages: RT-673 K, 673-1073 K and 1073-1273 K. The irradiation cascade displacement induced mono-vacancies (~ 190 ps) in W, together with a large amount of dislocation loops (~ 150 ps) in W5Re alloys. As the annealing temperature increases to 473 K, the vacancy starts to migrate and aggregate to form the V n clusters [13,40]. The number "n" reaches 11 after 1273 K annealing. The research reported a strong binding energy of ~ 4.5 eV [48] between He atom and vacancy, which is ~ 1 eV between two He atoms at close separation distance in W matrix. Thus, as temperature increases, the vacancy-type defects continuously attract the incident He atoms to form He n V m complexes by increasing the annealing temperature [10]. The growth of vacancy clusters results in large positron annihilation lifetime in these effective and open volume sites. However, the thermal desorption spectrum (TDS) results in Ref. [45] of He-irradiated W and W5Ta showed the desorption of He beginning from 450 K. The desorption rate of He in TDS spectra increased with the annealing temperature until it reached the maximum desorption rate at ~ 997 K [45]. This result is in well agreement with the data in this work. In this progress, the migration and aggregation of vacancies promote the formation of vacancy clusters. The coarsening vacancy clusters transform into micro-voids, in which the positron annihilation lifetime is even about 380 ps [13]. Another small desorption peak is detected in the TDS spectrum at ~ 1280 K in Ref. [45]. In the similar case, the positron lifetime is reduced to ~ 240 ps, implying that the mediumsize vacancy clusters are attributed to the dissociation of large He n V m complexes and the bursting of the He bubbles even the annealing of micro-voids remain in both W and W5Re.

Conclusion
The thermal evolution kinetics of He-irradiation induced defects in W and W5Re alloys of 0.83 dpa, 1.62 dpa and ~ 1.45 dpa were studied by positron annihilation spectroscopy. All irradiated samples were isochronal annealed from 373-1273 K for 1 h. Similar thermal evolution characteristics of defects were shown in both W and W5Re, which could be divided into three stages: RT-673 K, 673-1073 K and 1073-1273 K. The conclusions of this work are summarized as follow.
1. He-ion irradiation introduced a lot of mono-vacancies (positron annihilation lifetime of ~ 190 ps) in W, together with dislocation loops (~ 150 ps) in W5Re alloys.
2. As the annealing temperature increased to 1078 K, large-size vacancy clusters composed of 4-8 monovacancies formation by expensing of small vacancy clusters were detected in both W and W5Re. The vacancy-type defects in W5Re were more susceptible to the annealing temperature in IIR (27-249 nm) due to the formation of vacancy clusters-Re complexes. 3. After 1273 K annealing process, smaller positron annihilation lifetime (~ 240 ps) was detected, which were attributed to the recovery of dislocation loops, the dissociation of large He n V m complexes and the annealing of micro-voids in both W and W5Re. 4. Re clusters in W5Re alloy may serve as the nucleation sites of He bubbles, which promote a great plastic deformation such as swelling and protrusion formation on the surface.