Effect of 0.25 and 2.0 MeV He-ion irradiation on short-range ordering in model (EFDA) Fe-Cr alloys

The effect of He+ ion irradiation on a distribution of Cr atoms in model (EFDA) Fe(100-x)Cr(x) (x=5.8, 10.75, 15.15) was studied by means of conversion electrons Mossbauer spectroscopy. The alloys were irradiated to the dose of 1.2E16 ions per cm2 (7.5 dpa) with He+ of 0.25 and 2.0 MeV. The distribution of Cr atoms within the first two coordination shells around the probe Fe atoms was expressed in terms of the Warren-Cowley short-range order parameters alpha1 (first-neighbour shell, 1NN), alpha2 (second-neighbour shell, 2NN) and alpha12 (1NN+2NN. In non-irradiated alloys alpha1 was positive and alpha2 was negative for all three samples which indicates an ordering of Cr atoms. Yet, the value of alpha12 were close to zero, i.e. the distribution of Cr atoms averaged over the first two coordination shells around the probe Fe atoms, was random. The effect of the irradiation of the Fe5Cr sample is similar for the two energies of the He+ projectiles viz. increase of number of Cr atoms in the 1NN and decrease in the 2NN shell. Consequently, the degree of the ordering increased. For the other two samples, the effect of the irradiation depends on the composition, and it is stronger for the less energetic ions where, in the case of Fe10Cr, the disordering disappeared and some traces of Cr clustering (alpha12 weakly positive) can be seen. In the Fe15Cr sample the clustering is clear. In the samples irradiated with 2.0 MeV ions the ordering also survived in the samples with x=10.75 and 15.15, although its degree became smaller than in the Fe5Cr sample. Small changes in the magnetic texture were revealed.


Introduction
Fe-Cr based ferritic steels (FS) such as ODS and ferritic-martensitic (F-MS) steels such as EUROFER constitute an important class of structural materials. Their relevance for numerous industrial and technological applications is a consequence of their desirable swelling, high temperature corrosion and creep resistance properties [1,2]. In these circumstances FS as well as F-MS have been recognized as valuable construction materials for applications in new generations of nuclear power facilities including generation IV fission reactors and fusion reactors as well as for other technologically important plants e.g. high power spallation targets [3][4][5]. They have, for example, been used for manufacture of such systems as container of the spallation target, fuel cladding or primary vessel. These devices work in-service under extreme conditions such as elevated temperatures and/or long-term irradiation. Under these conditions, the materials experience irradiation damage that can gravely degrade their mechanical properties. On the lattice scale, the radiation produces lattice defects including vacancies, interstitials and dislocations. A redistribution of Fe/Cr atoms subsequently occurs and can produce several microscopic phenomena such as short-range order (SRO), segregation or phase decomposition into Fe-rich () and Cr-rich (') phases. All of these effects result in enhanced embrittlement which is highly undesirable. In laboratory conditions, different projectiles have been used as irradiation media to study irradiation effects in Fe-Cr alloys and/or FS or F-MS. In particular, the effects of neutrons [7,8], protons [9,10], self-ions (Fe, Cr) [11,12], He ions [13,14], Kr ions [15] and electrons [16] have been recently studied. Irradiations were performed under different conditions of 3 temperature, projectile energy, flux and dose. This makes a comparison of the obtained results more difficult as the effects of irradiation depend not only on the type of irradiation but, for a given type, also on the irradiation flux and temperature. To exemplify the former, no ' precipitates were observed in Fe88Cr12 alloy irradiated at 300 o C with Fe + ions to 0.6 dpa but they were revealed in the same alloy irradiated with neutrons at the same temperature and with the same dose [17]. Regarding the latter, ' precipitates were found in low dose rate neutron-irradiated samples (10 -9 -10 -6 dpa/s) while they were absent if the dose rate was high (10 -4 dpa/s) [18]. Similar effects of the dose rate on irradiation hardening were reported for Fe-Cr alloys irradiated with Fe + ions [19]. These observations illustrate that results obtained for materials irradiated with high dose rates, which is typically the case for in vitro studies, may not be simply related to the effects caused by in vivo irradiations in which very low dose rates are involved (assuming the same dose of irradiation).
Understanding of irradiation-induced effects can be backed by theoretical calculations such as those recently reported on radiation-induced segregation [20] and radiation-accelerated precipitation [18] in Fe-Cr alloys. In any case further experimental and theoretical studies on these key issues are needed in order to enhance our knowledge and understanding of irradiation-induced phenomena in materials with key engineering significance.
Three model (EFDA/EURATOM) Fe100-xCrx alloys (x=5.8, 10.75 and 15.15) were irradiated to different doses with He + ions of 0.25 and 2.0MeV energy. Samples were investigated using Mössbauer spectroscopy (MS). MS has already proved to be a relevant method for the quantitative investigation of Fe-Cr alloys. Notably issues connected to determination of a distribution of Cr atoms in Fe matrix e. g. [21][22][23][24][25][26][27], the solubility limit of chromium [28,29] and the kinetics of ' precipitation [29] can be successfully studied with high precision using MS.
The irradiation of Fe-Cr alloys with He + ions is of interest, as the production of helium occurs during exposure of the various devices produced therefrom to proton and/or neutron irradiation [1]. Its presence deteriorates mechanical properties of steels. In particular, it lowers the critical stress for inter granular structure and also it may bring about a severe decrease of the fracture toughness [30]. Therefore, understanding not only the effects of radiation damage but also the effects of helium on the mechanical properties of FS/F-MS are important topics to be studied in the context of gaining a better understanding of irradiationinduced degradation processes in engineering materials used in various important branches of industry, including nuclear power. 4

Samples and irradiation
Samples investigated in this study were prepared from model EFDA/ EURATOM master Fe-Cr alloys fabricated in 2007. They were delivered in the form of bars 10.9 mm in diameter, in a re-crystallized state after cold reduction of 70% and then heat-treated for 1h under flow of pure Ar at the following temperatures: 750 o C for Fe94.2Cr5.8, 800 o C for Fe89.25Cr10.75 and 850 o C for Fe84.85Cr15.15, followed by air cooling. For the present study, a slice 1 mm thick was cut from each bar using a diamond saw, and was subsequently cold-rolled (CR) down to a final thickness of 30 m. The samples in form of 25 mm rectangles were irradiated at the JANNUS multi-ion beam irradiation platform at CEA, Saclay, France with 0.25 and 2.0 MeV He + ions to a dose of 1.210 17 4 He + cm -2 , which is equivalent to radiation damage of 7. The pre surface zone accessible to the CEMS measurements is marked by a vertical stripe.

Spectral measurements and analysis
The Mössbauer spectra were measured at room temperature (RT, 290K) by recording conversion electrons (CEMS mode) in a backscattering geometry using a conventional constant acceleration spectrometer and a 57 Co(Rh) source of 14.4 keV gamma-rays with a nominal activity of 3.7 GBq. The measured spectra contain information from a surface/pre surface zone whose thickness is less than 0.3 m. The spectra were recorded both on irradiated (IR) as well as on non-irradiated (NIR) surfaces of the samples which were built in a proportional gas flow counter. A He/methane mixture (90:10) was used as the counting gas.
The recorded spectra are displayed in Fig. 3. Spectra were analyzed using a two-shell model, . However, the probabilities of the atomic configurations, P(n1,n2), were treated as free parameters (their starting values were those calculated from the binomial distribution) in spectral analysis. All spectral parameters such as X(0,0), Xi, line widths of individual sextets G1, G2 and G3 and their relative intensities (Clebsch-Gordan coefficients) C2 and C3, were also treated as free parameters (C1=1). Very good fits (in terms of a  2 -test) were obtained with the spectral parameters displayed in Table 1. Their values are in close agreement with the corresponding values reported previously [20][21][22][23][24][25][26][27][28].
Knowledge of the atomic configurations, (n1,n2), and their probabilities, P(n1,n2), enabled determination of the average number of Cr atoms in 1NN,   

Short-range order parameters
A distribution of atoms in an alloy (here Fe-Cr) can be quantitatively described using Warren-Cowley short-range order (SRO) parameters, k. The techniques applied in the present study, i.e. Mӧssbauer spectroscopy, makes it possible to determine the SRO-parameters for the first, 1, and for the second, 2, nearest-neighbor shells, separately. This enables calculation of the SRO-parameter for both shells, 12. In turn, knowledge of the SRO-parameters makes it possible to qualitatively discuss the distribution of Cr atoms in the Fe matrix. The values of k (k=1,2,12) can be determined using the following equation [23]: where <nk> is the number of Cr atoms in the k-th near-neighbour shell around the probe Fe atoms as found from the analysis of the spectra, while <nok> is the number of Cr atoms in the 8 k-th near-neighbor shell calculated assuming their distribution is random i.e. <n01 >=0.08x, <n02>=0.06x, and <n01+n02>=0.14x.

Effect of He + energy and chromium concentration
In order to study whether the effects of irradiation depend on the energy of He + ions and alloy composition, all three samples were irradiated at room temperature (ca. 290K) to a dose of 1.210 16 He + /cm 2 (7.5 dpa) with 0. 25   It can be seen that the 1-values are positive for all samples, with the highest value for the lowest-Cr sample. This means, in terms of eq. (1), that the average number of Cr atoms situated in the 1NN-shell around the probe Fe-atoms is smaller than expected for a random 9 distribution. On the other hand, the values of 2 are negative, hence the lattice site occupancy by Cr atoms in the 2NN-shell is overpopulated, as compared with the expectation from a binomial distribution. This means, in terms of the pair interaction potential between Fe and Cr atoms, that the effective potential is repulsive if the separation between these atoms is equal to the 1NN-radius; and it is attractive if the two types of atom are separated from each other by the radius of the 2NN-shell. It must be, however, realized that this effective potential has several contributions including electronic, magnetic, configurational and vibrational ones.
It can be also seen that the absolute values of 1 and 2 are significantly larger for the lowest-Cr sample than for the other two samples. This behavior of 1 and 2 indicates an ordering of Cr atoms and, obviously, a degree of the ordering higher for the Fe94.2Cr5.8 sample than for the other two samples. The SRO-parameter averaged over the two shells, 12, is equal to zero within the error limit. This means that the distribution of Cr atoms as measured within the volume of the 1NN-2NN neighborhood is random.
To figure out the effect of the irradiation on the Cr atom distribution a difference between the corresponding k-values, ions [32].
The irradiation-induced change of the SRO parameter 12 presented in this paper and that reported previously [31] compare pretty well with the recent molecular dynamic simulations (MDS) for disordered Fe100-xCrx alloys [33], as far as the alloys with x=10 and 15 are concerned viz. 12 is unaffected by the irradiation for the former and becomes positive for the latter xvalue. For the least Cr-concentrated sample i.e. x=5, the measured 12 remained unchanged for the alloy irradiated with the 2.0 MeV ions, but increased in the case of the 0.25 MeV ions.
For this alloy the MDS predicted negative value for 12. The comparison should be, however, taken with caution, because it was assumed in the MDS that the distribution of Cr atoms was random in the non-irradiated samples i.e. 12 =0. In our case, 12 was also equal to zero -see Fig. 4, but the distributions of Cr atoms over the 1NN and 2NN shells were not random (1 > 0, 2 < 0). It should be also noticed that the MDS were performed for the overlapping of single 5 keV displacement cascade events, therefore its relevance to our study is not straightforward.
In any case, the results reported in this paper are, to our best knowledge, the first that give evidence on clustering of Cr atoms in a Fe84.85Cr15.15 alloy irradiated with He + ions. This 14 observation may contribute to a better understanding of mechanisms underlying irradiationinduced effects by He ions in Fe-Cr alloys and produced therefrom nuclear structural materials. In the available literature there are numerous reports on the issue e. g. [13,14,[34][35][36][37][38][39][40][41]. The reported results have not permitted to obtain the full understanding of the underlying mechanism and the observed effects like enhanced hardening, embrittlement and swelling are still a matter of discussion and even controversy. This situation has many reasons, and first of all a lack of systematic studies in terms of various parameters that can affect mechanical and structural properties of these materials. It is well known that such conditions of irradiation like: temperature, dose, fluence, radiation damage, energy of He ions and composition of materials have to be taken into account. In particular, it was revealed in previous studies that the lower the temperature of the irradiation the higher the degree of hardness [36,38]. The dose and the radiation damage must be high enough to result in the increased hardening [39,41]. The effect of the irradiation was also revealed to depend on the irradiated material [39,40]. The observed mechanical effects of the He-irradiation are most frequently explained in terms of formation of He-bubbles, especially small ones, and also changes in the microstructure induced by the displacement damage [35,36]. However, Ullmaier and Camus concluded their study that the increase in yield stress following the Heirradiation was solely due to displacement induced defects but not to the presence of He itself [34]. It should be noticed here that the above-discussed changes in the microstructure of the He-irradiated materials were mostly investigated using the transmission electron microscopy (TEM) and, rarely, by small angle neutron scattering (SANS) e.g. [36]. These methods do not allow detecting changes in a distribution of Cr atoms that, as we have shown in this study, occur upon He-ion irradiation. In other words, in order to properly understand mechanisms responsible for the irradiation-caused changes in the mechanical properties of nuclear materials one has also to take into account changes in the distribution of atoms in the matrix.
It is well known that clustering of Cr atoms in Fe-Cr alloys and in Fe-Cr based steels results in an enhanced embrittlement, hence the redistribution cannot be neglected. 15 The observed irradiation-induced changes in the distribution of Cr atoms within the 1NN-2NN volume around the probe Fe atoms can be also expressed in terms of underlying changes in the local concentration of Cr, xk, defined as follows [26]:
The xk -values obtained based on eq. (1) are displayed in Table 1. It can be readily noticed that, as a rule, the x1-values are systematically smaller and those of x2 systematically greater than the corresponding average values. To extract the effect of irradiation a difference in xk,  Concerning the less energetic ions, the effect strongly depends on the sample. In the least Cr- The results displayed in Figs. 5 through 9 clearly show that the Fe94.2Cr5.8 sample behaves differently than the other two ones. Namely, the applied irradiation caused in this sample the greatest changes in the distribution of Cr atoms. Interestingly, the greatest increase in the hardness caused by an Fe + ion irradiation was also observed in a Fe94.2Cr5.8 sample (having the same origin as ours) [19]. Theoretical calculations showed that the mixing energy in random Fe100-xCrx alloys is negative for x 10 and has its minimum at 5at% Cr [32], which correlates with our findings.

Change in the magnetic texture
The knowledge of a relative intensity of the 2 nd /5 th line, C2/C3, can be used to determine an average angle between the direction of the -rays (in this case perpendicular to samples surface) and that of the magnetization vector, . For this purpose the following equation can be used: The -values obtained from equ. (3) and the C2/C3 data displayed in Table 1  It is clear that the -values are characteristic of the sample. In the non-irradiated samples  linearly decreases with x. However, this behavior does not necessarily reflect the effect of composition, but it may be due to a various degree of deformation caused by cold rolling of the samples [26]. In order to extract the effect of the irradiation we calculated the difference in , =(IR)-(NIR), and the result is displayed in Fig. 11.

Conclusions
The following conclusions can be drawn based on the results obtained in this study: (1) The population of Cr atoms in the first-neighbour shell (1NN) around the probe Fe atoms in the non-irradiated alloys is lower (1>0) and in the second-neighbour shell (2NN) is higher (2<0) than expected from the binomial distribution.
(2) Upon He + ion irradiation, the degree of the ordering increased, i.e. 1 became more negative and 2 more positive in the least Cr-concentrated sample whereas it remained quasi unchanged in the other two alloys. 20 (3) Irradiation with the 0.25 MeV ions proved to be more effective as far as the redistribution of Cr atoms is concerned, and, in particular, it caused a clustering of Cr atoms in the most Crconcentrated sample.
(4) Clustering of Cr atoms was not found in the samples irradiated with the 2.0 MeV ions, but the ordering observed in the non-irradiated samples was enhanced. The degree of the enhancement was decreasing with the Cr content.  The pre surface zone accessible to the CEMS measurements is marked by a vertical stripe.