Vacancy type defect formation in irradiated α-iron investigated by positron beam Doppler broadening technique

Vacancy type defects formations have been investigated in virgin and irradiated a-iron samples using slow positron beam Doppler broadening technique. Mono-vacancies and vacancy clusters are observed in 1.5 MeV 4He ions irradiated Fe samples at varying fluences from 1×1013 to 1×1017 cm−2. In the 1.2 MeV Yttrium ions implantation at low fluence 1×1014 cm−2 vacancy clusters with higher concentration and larger size are formed. In this sample, vacancy defects are detected deeper than predicted by SRIM calculation due to channelling.


Introduction
Iron based alloys, and for example, Oxide Dispersion Strengthened (ODS) ferritic steels, constitute good candidates for structural components in fission reactors or future fusion reactors which are submitted to extreme conditions of irradiation and temperatures. Changes in microstructure and macroscopic properties of these nuclear materials are governed by the kinetics of defects produced by irradiation, such as vacancy-type clusters or vacancy-impurity complex clusters. In ODS steels, the Yttrium, Titanium oxide nanoparticles are traps for He and pin dislocations allowing to limit the swelling and creep. These materials are fabricated by mechanical alloying and their interesting properties are very dependent of the nanoparticles distribution in size and in space. Fundamental studies are needed to improve the control of this distribution. One of the key-point in the understanding of the formation of these particles is the role of vacancy defects induced during milling. As a key component of these materials, the understanding of defect formation in irradiated bcc iron (α-Fe) is essential. Many efforts, both experimental and computational works, have been put into this study. Some are carried out by using positron annihilation spectroscopy as its sensitivity to the size and density of vacancy clusters in the range from mono-vacancies up to cavities containing 50-100 vacancies [1]. Recently T. Iwai et al. [2,3] have studied the damage induced in iron by various ion species (He, C, O, Fe) by using positron beam Doppler broadening. Mono-vacancies and vacancy clusters have been observed but characteristics S/W of positron annihilation in these defects have not been defined which are important to discriminate between the different vacancy types.
In this study, the vacancy-type defects induced by 1.5 MeV 4 He ions at varying fluences in α-Fe are investigated with slow positron beam Doppler broadening (SPBDB) technique at the CEMHTI laboratory. It allowed defining the annihilation characteristics S and W for mono-vacancy and vacancy clusters. Then the defects produced by 1.2 MeV Yttrium implantation in α-Fe have also been studied. This will be the basis for our study on the interaction of vacancy clusters with Y.

Experimental
Samples are 99.99% pure iron 7 mm×7 mm×0.5 mm sheets. After polishing, the samples were annealed at 800°C for one hour in vacuum. The grain size is about 20-100 μm. 1.5 MeV 4 He irradiations were carried out with the 3.5 MeV Van der Graaff accelerator at CEMHTI Orleans. Eight different fluences have been performed in the range from 1×10 13 to 1×10 17 cm -2 . 1.2 MeV Yttrium implantations were done with the 3 MeV Tandetron accelerator at laboratory HZDR Dresden at the fluence of 1×10 14 cm -2 producing a similar order of magnitude of damage dose compared to 1.5 MeV 4 He irradiation at fluence 1×10 17 cm -2 . Both these two types of irradiations were carried out at room temperature.
The momentum distribution of electron-positron pairs has been measured at 300 K by recording the Doppler broadening of the 511 keV annihilation line characterized by the low (S) and the high (W) momentum annihilation fraction in the momentum range (-2.49, 2.49)× 10 -3 and (-24.88, -9.64), (9.64, 24.88) × 10 -3 m 0 c, respectively. To investigate the depth dependence of S and W, the S(E) and W(E) were measured as a function of the positron energy E changed in 500 eV steps in the 0.5 -25 keV range using the slow positron beam. For this energy range, the maximum positron penetration depth in pure Fe is at most ~2 µm. Figure 1 shows the depth profiles of Y and He concentrations and the corresponding damage calculated by using SRIM 2008 [4] for pure Fe irradiated by 1.5 MeV 4 He and 1.2 MeV Y. In these calculations, the Fe displacement threshold energy has been fixed at a value of 40eV [5]. In the same figure, the slow positron implantation profile is also illustrated for different energies [6]. In the track region of 4 He ions located between surface and 2µm which is probed with the slow positron beam, the damage dose is 0.1 dpa (displacement per atom) and the helium concentration is 4 atomic ppm for the highest fluence 10 17 cm -2 . This concentration is quite low, thus the effect of helium on vacancy defects could be ignored in the following. 1.2 MeV Y ions are implanted between surface and 500 nm in Fe. The projected range (Rp) is 265 nm. Thus, the positrons could detect the entire implanted region. At fluence 1×10 14 cm -2 , the damage dose is 0.14 dpa at the surface and increases up to reach 0.24 dpa at the depth of 170 nm and the Y concentration reaches the maximum value of 58.9 appm at the depth of 270 nm. This concentration is lower than yttrium solubility in iron (290 at. ppm in iron at 800°C [7]). The profile of Y concentration has also been calculated using Marlowe code considering that sample is a polycrystal. It suggests that Y ions stop deeper than Rp up to ~1500 nm.

Virgin iron samples
All as-received polished and 800°C/1 h/vacuum annealed Fe samples have given SPBDB reproducible results and a typical representation is illustrated for two samples in Ni (99,997 wt.%) annealed at 800°C in vacuum for one hour [10]. All     density and momentum at the trap [11]. This effect of helium is negligible at low fluences when the He concentration is very low. In the following we choose to consider the S and W values measured in the energy range between 20 and 22 keV as the annihilation characteristics in the track region of the ions. In this range S(E) changes very slightly, the diffusion back of positron to surface and the annihilation fraction in cascade region are negligible. In this energy range S increases with the fluence (see arrow in Fig 3.a). It indicates that damage increases when the irradiation fluence increases in the pure Fe samples.

1.5 MeV 4 He irradiated iron samples
The mean values of S and W estimated in this energy range will be called S and W respectively to lighten the text. The evolution of S as a function of W is plotted on Figure 3(b). At low fluence from 1×10 13 to 2×10 14 cm -2 (from 1×10 -5 to 2×10 -4 dpa), the S,W points follow the same straight line D1 which goes through the (S ref , W ref ) point. It indicates that the same nature of vacancy defects has been created for these fluences. In these conditions of irradiation, SRIM predicts that the energy of the Primary Knocked-on Atoms (PKA) is low (91.8eV) so the recoiled atoms collisions should lead to the formation of small cascades with essentially single vacancies as molecular dynamics modelling predicts [12]. Thus, we will propose that this D1 line is more likely characteristic of annihilation in single vacancy in Fe. 1.0x10 13 (dpa 1x10 -5 ) 2.0x10 13 (dpa 2x10 -5 ) 1.0x10 14 (dpa 1x10 -4 ) 2.0x10 14 (dpa 2x10 -4 ) 5.0x10 14 (dpa 5x10 -4 ) 1.0x10 15 (dpa 1x10 -3 ) 1.0x10 16 (dpa 1x10 -2 ) 1.  From the fluence of 5×10 14 cm -2 (5×10 -4 dpa), the S,W points begin to leave the single vacancy line. This indicates that the nature of vacancy defects has changed and larger vacancy-type defects are detected. When fluence increases, the shift from D1 characteristic line increases indicating that the size of vacancy clusters grows with the damage dose. Such clusters formation could be explained by the migration of single vacancies generated by irradiation that could agglomerate leading to the formation of larger clusters when their concentration increases that means with the fluence. Indeed, migration temperature of V in Fe is below the room temperature (-53 to 5°C) as suggested by Takaki et al. [14]. T. Iwai et al. [2] investigate accumulation of vacancy-type defects in iron produced by 1 MeV He irradiations at room temperature by positron beam both in-situ during irradiations and after irradiations. Vacancy clusters have been observed from ~1×10 -2 to 0.1 dpa and the results suggest the S values are not likely to change under and after irradiation. Thus, the vacancy clusters in Fe by irradiation with 4 He 1.5 MeV (this study) could be considered to form during the irradiation by migration and agglomeration. Between 1×10 16 cm -2 and the highest fluence of 1×10 17 cm -2 the S and W values don't change too much indicating that saturation occurs. This saturation could correspond to a maximum size of clusters that can be formed in these irradiation conditions. We will define the straight line going from lattice S,W point and S,W measured in sample implanted at the highest fluence as D2 characteristic of annihilation of positrons trapped in vacancy clusters. Being given the migration energy of the single vacancies it is therefore unlikely that they survive in the irradiated sample at room temperature. They should either disappear by diffusion into sinks or agglomerate with other vacancies into clusters or form pairs with some impurities which stabilizes them. It has to be noted, the carbon concentration in these iron samples is about 56 appm. They could be bound to single vacancies and stabilize them as it has been shown in [13]. It suggests that D1 could be characteristic of the detection of single vacancies as proposed above.

1.2 MeV Yttrium implanted iron samples
As shown in Figure 4   Another observation has attracted our attention. The S(E) curve of 1.2 MeV + Y implantation begins to decrease from 15 keV instead of 10 keV energy from which a fraction of positron should annihilate in the defect-free Fe lattice as suggested by the depth damage profile calculated by SRIM (Figure 1.b). It indicates that positrons detect vacancy clusters deeper than expected by SRIM. A structure of four layers allows to fit the S(E) and W(E) experimental curves with Vepfit. The S and W values of each of these layers are plotted in Figure 4 and indicate that vacancy clusters are