- Email: [email protected]
Microstructure evolution during helium irradiation and post-irradiation annealing in a nanostructured reduced activation steel
Journal of Nuclear Materials 479 (2016) 323e330
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Microstructure evolution during helium irradiation and postirradiation annealing in a nanostructured reduced activation steel W.B. Liu a, c, *, Y.Z. Ji b, P.K. Tan a, C. Zhang c, C.H. He a, Z.G. Yang c a
Department of Nuclear Science and Technology, Xi'an Jiaotong University, Xi'an 710049, China Department of Materials Science and Engineering, The Pennsylvania State University, University Park, PA 16802, USA c Key Laboratory of Advanced Materials of Ministry of Education, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, China b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 8 January 2016 Received in revised form 1 July 2016 Accepted 11 July 2016 Available online 14 July 2016
Severe plastic deformation, intense single-beam He-ion irradiation and post-irradiation annealing were performed on a nanostructured reduced activation ferritic/martensitic (RAFM) steel to investigate the effect of grain boundaries (GBs) on its microstructure evolution during these processes. A surface layer with a depth-dependent nanocrystalline (NC) microstructure was prepared in the RAFM steel using surface mechanical attrition treatment (SMAT). Microstructure evolution after helium (He) irradiation (24.8 dpa) at room temperature and after post-irradiation annealing was investigated using Transmission Electron Microscopy (TEM). Experimental observation shows that GBs play an important role during both the irradiation and the post-irradiation annealing process. He bubbles are preferentially trapped at GBs/ interfaces during irradiation and cavities with large sizes are also preferentially trapped at GBs/interfaces during post-irradiation annealing, but void denuded zones (VDZs) near GBs could not be unambiguously observed. Compared with cavities at GBs and within larger grains, cavities with smaller size and higher density are found in smaller grains. The average size of cavities increases rapidly with the increase of time during post-irradiation annealing at 823 K. Cavities with a large size are observed just after annealing for 5 min, although many of the cavities with small sizes also exist after annealing for 240 min. The potential mechanism of cavity growth behavior during post-irradiation annealing is also discussed. © 2016 Elsevier B.V. All rights reserved.
Keywords: Nanocrystalline Ion irradiation Helium bubble Post-irradiation annealing Reduced activation steel
1. Introduction Helium (He) irradiation behavior in metals has been the subject of extensive research in the past several decades [1,2]. For example, investigation by Lane et al. [3] showed that misfit dislocations at grain boundaries (GBs) are preferred nucleation sites of He bubbles in austenitic steels. Furthermore, Singh et al. [4] confirmed that dislocation intersections, both within and outside of GBs, are preferred bubble nucleation sites in aluminum. Since GBs/interfaces are known to act as potential point defect sinks, and nanocrystalline (NC) material has a large volume fraction of GBs [5], He bubbles can be effectively trapped at GBs in NC materials. Therefore, NC materials are likely to have different radiation damage tolerance from their coarse-grained counterparts [6]. Severe plastic deformation (SPD) technology, such as equal
* Corresponding author. Department of Nuclear Science and Technology, Xi'an Jiaotong University, Xi'an 710049, China. E-mail address: [email protected] (W.B. Liu). http://dx.doi.org/10.1016/j.jnucmat.2016.07.030 0022-3115/© 2016 Elsevier B.V. All rights reserved.
channel angular pressing (ECAP) [7], high-pressure torsion (HPT) [8] and accumulated roll bonding (ARB) [9], has been recognized as an effective way to refine microstructures of various materials. Surface mechanical attrition treatment (SMAT), sometimes called ultrasonic shot peening (USSP), is regarded as one of the most important means of SPD, and it has been employed to obtain various NC materials, including pure metals, alloys and intermetallics [10e12]. Experimental results show that large amounts of GBs/interfaces and dislocations are introduced into the matrix during SMAT [13,14], which are likely to improve the radiation resistance [6]. Due to its excellent properties [15,16], reduced activation ferrite/ martensitic (RAFM) steels (Fe-Cr-W-V-Ta) have potential applications as first wall and blanket structural materials in fusion reactors, and a lot of research has been done to investigate the microstructural evolution during He ion irradiations. Studies of Neions irradiated 9Cr ferrite/martensitic steel at high temperatures showed that growth of cavities at GBs, especially at GB junctions, was faster than that within the grain [17]. Microstructural
324
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
observations of high temperature (550 C) helium irradiated 9Cr1Mo martensitic steel [18] also showed that voids were preferred to form on prior austenite GBs and other interfaces, such as carbidematrix interfaces, sub-GBs, lath boundaries and dislocations inside the lath structures, while no preferential nucleation site was observed when the helium implantation was performed at room temperature (RT) [19]. In addition, a typical void distribution near a coincident site lattice boundary was observed in neutron irradiated Fe-15Cr-15Ni steel neutron-irradiated at 749 K to 18 dpa, although void denuded zones (VDZs) were formed near the random GBs [20]. Preferential He bubble formation in carbides of a tempered F82H ferritic-martensitic steel during low temperature He irradiation was observed recently, and this phenomenon was attributed to the diffusing He being trapped in the carbide due to the strong He-C bond, which can lead to the formation of He bubbles with the increase of He concentration in the carbides [21]. However, there is still a lack of research about the effect of GBs in RAFM steels on irradiation induced microstructural evolution. In the past decade, a number of studies have been devoted to investigating the role of interface structures in the damage resistant properties of NC materials [22e24]. Studies of He ion-irradiations on NC Cu/Nb multilayers show that, in contrast to the copious bubbles contained in the He-implanted pure Cu and Nb, the Cu-Nb interfaces in the Cu/Nb multilayers are responsible for suppressing He bubble formation [22,23]. Nevertheless, when He is preferentially trapped at interfaces, there exists a critical He concentration below which bubbles were not observed [24]. Experimental results of radiation damage in He ion irradiated NC iron (Fe), which was produced by a deposition method using the magnetron sputtering technique, showed that smaller grains lead to lower density of He bubbles, and no voids were observed after He implantation [25]. However, a study about He ion irradiation in NC RAFM steel is still needed to clarify the effect of GBs on irradiation induced microstructural evolution, since the NC microstructure produced by severe plastic deformation is critically different from that produced by magnetron sputtering technique. It is reported that voids and helium-filled bubbles have different effects on He-induced embrittlement: voids are the main reason for swelling and high temperature embrittlement, while the effect of helium-filled bubbles is comparatively small or, in some cases, even beneficial [26]. The “bubble” and “void” can be described as [27,28]: under some conditions, a helium-filled cavity with diameter less than ~ 10 nm, neither grows nor shrinks, but remains stable as a “He bubble”; while under other conditions, it grows without bounds by absorbing vacancies, and finally results in the formation of a “void”. Therefore, accurate experimental studies of microstructural evolution during ion irradiation are necessary, and this is one of the main objectives of the present work. The stability of a void is not dependent on the presence of internal pressurization from a gaseous species such as helium, and bubbles are usually defined as pressurized cavities, while cavity can be used to refer to either bubbles or voids [29]. Hence, it is necessary to study the average size evolution of cavities during irradiation and postirradiation annealing. In the present contribution, He bubble evolution during irradiation and post-irradiation annealing in a quenched fully martensitic NC RAFM steel is systematically studied. Section 2 deals with experimental details about materials, the SMAT process, He ion irradiation, annealing treatment and all the characterization methods. The experimental results, concerning NC structure characterization after SMAT, bubble formation and microstructural evolution under He irradiation, and void formation upon postirradiation annealing, are presented in Section 3. In Section 4, the mechanisms of microstructural evolution during SMAT, He bubble formation during low temperature implantation, especially the
effect of GBs/interfaces during irradiation and post-irradiation annealing, are discussed. 2. Experiments 2.1. Material and SMAT process The material used in the present investigation was a RAFM steel with chemical composition (in wt%): 0.09% C, 0.49% Mn, 8.75% Cr, 0.84% Ta, 1.58% W, 0.21% V and balance Fe. Heat treatment included austenitizing at 1253 K for 45 min, followed by water quenching. It is noted that the steel, which was quenched after austenitization but not tempered, produced a fully martensitic microstructure, and the microstructure without M23C6 type carbides was used to study the effect of GBs on irradiation-induced microstructures evolution. The plate sample (F50 4.0 mm in size) of the quenched steel was submitted to SMAT. The equipment and procedure of SMAT are discussed in detail in Refs. [30,31]. The ball size used for surface peening was 5 mm in diameter. The vibration frequency of the chamber was 20 kHz, and the plate was treated for 30 min. During SMAT, the strain and strain rate reduced rapidly with the increase of depth, and the strain rate at the sample surface was estimated to as high as 102e103 s1 [31]. Fig. 1 shows the distinct microstructures of the surface layer and the matrix obtained from cross-sectional SEM observations. No carbides can be obviously seen both in the SMAT layer and in the matrix from the SEM results. Grains with no deformation and grain boundaries can be clearly seen in the matrix (Fig. 1a), and most of the grains are smaller than 10 mm. With increasing depth from the treated surface, gradient nano-submicron-microstructure and obvious deformation bands (Fig. 1b) are found due to the decreasing strain and strain rate. In the topmost layer regions, grains are too small to distinguish their boundaries. Microbands are clearly observed in the inner area. The thickness of the surface deformation layer in Fig. 1b is not uniform (30e46 mm), which can be attributed to the heterogeneous nature of plastic deformation within and between grains [32]. 2.2. He-ion irradiation The He-ion irradiation experiment was performed on the 320 kV ECR (electron cyclone resonance) experimental platform in the National Laboratory of Heavy-ion Accelerators in Lanzhou, China. He-ions with a kinetic energy of 300 keV were used for irradiation. The irradiation experiment was conducted at room temperature with vacuum pressure about 106 Pa. The scanned beam size was about 17 17 mm2. The RAFM steel sample was irradiated with a fluence of 3.31 1017 ion/cm2, which corresponds to the estimated displacement levels of 24.8 dpa (displacement per atom) at peak damage region. The average beam current was about 3.83 1012 ion/cm2/s during the irradiation experiment. Displacement damage dose regimes for structural materials (Ferrite/Martensitic steel) in fusion energy systems can be larger than 100 dpa [33], and basic research is still needed to improve understanding of high dose irradiation-induced microstructure evolution. Hence, dose of ~25 dpa was selected in the present work to investigate microstructure evolution in NC RAFM steel under helium ion irradiation. Fig. 2 shows the dose and deposited He concentration calculated from the stopping and range of ions in matter 2013 (SRIM-2013) software [34]. Materials with the same chemical composition in the experiment and He2þ ions with energy of 300 keV were used in the calculation. The peak value of He concentration can be as high as 18.20 at%, and the peak region of the deposited He in the irradiated
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
325
Fig. 1. Cross-sectional SEM images after SMAT (a) about the matrix (untreated area) (b) about the surface (SMAT-treated) layer.
30
20
25
15
Dose (dpa)
20 10
15 10
5 5 0 0
200
400
600
800
0 1000
Deposited He concentration (at%)
Dose Deposited He concentration
Depth (nm) Fig. 2. Dose and deposited He concentration calculated from SRIM software.
buried layer is at a depth of 700e740 nm, which is much smaller than the thickness of nanocrystalline layer produced by SMAT. 2.3. Post-irradiation annealing It is reported that the highest operating temperature (based on thermal creep and radiation damage considerations) for the RAFM steel is 773 Ke873 K [35,36], and the nanocrystalline RAFM steel has excellent thermal stability up to that temperature range [37,38]. In order to investigate the thermal stability of the NC structures and helium bubbles after irradiation, SMAT-ed samples subjected to irradiation were annealed at 823 K for 5 min and 240 min. To avoid oxidation, the samples used for annealing were sealed in glass tubes under 0.3 atmosphere of argon. 2.4. Microstructural examination X-ray diffraction (XRD) analysis of the surface layer was carried out to measure the phase and grain sizes. Step size of 0.02 and range of 2q from 20 to 100 were taken to measure the XRD intensity before and after SMAT. Average grain size in the SMAT surface layer were derived from the full width at half maximum intensities of XRD peaks using Scherrer-Wilson equation [39], and
the measured (200) and (211) peaks were selected. Step size and holding time per step in the step scanning mode were 0.01 and 2 s, respectively. Cross-sectional observation of the treated sample was performed on a scanning electron microscope (SEM) of type JEOL JSM4500. The samples were ground and mechanically polished. The polished samples were etched in a particular solution (12 ml alcohol þ 3 ml hydrochloric acid þ 1 g ferric trichloride) to obtain the microstructure and morphology of the matrix and treated surface layer. Cross-sectional observation of the NC RAFM steel after irradiation was performed on transmission electron microscopes (TEM), JEOL JEM-2011 and Tecnai F20. The cross-sectional TEM samples were obtained by means of focused ion beam (FIB), which is an advanced analytical tool of sample preparation for studying microstructure and crystal structure on nanometer scales [40]. In the present work, an electrical field (30 kV) was used to accelerate the Ga ions; the corresponding current was about 1 nA. Firstly, a platinum (Pt) strip was deposited on the surface where the foil is to be cut. Then two trenches, one in front of the foil and the other behind the foil, were sputtered. At last, a TEM foil with dimensions of 5 10 0.15 mm was lifted out. The final thickness of the TEM foil was approximately 40 nm after being milled and polished under smaller accelerate electrical field (5 kV) and current (100 pA). 3. Results 3.1. Microstructures after SMAT A TEM bright field image obtained from the topmost surface of the SMAT sample and the corresponding dark field image taken using the BCC (110) diffraction spot are shown in Fig. 3. The TEM sample was obtained by means of cutting, grinding and thinning from the un-SMAT treated side at low temperatures. Selected area electron diffraction (SAED) patterns, which are composed of continuous rings, show that these grains are almost nano-grained bcc martensite (Fig. 3a). Grains of the original sample have been broken down to equiaxed nano-sized grains and the grains are random in crystallographic orientation (Fig. 3b). The average grain size calculated by statistical analysis is 6.20 nm. Most of the grains are smaller than 6 nm, although the largest grains are about 10 nm in diameter. However, the average grain size in the SMAT surface layer, which is derived from the full width at half maximum intensity of XRD peaks, is 11.5 nm, larger than the result from TEM
326
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
Fig. 3. TEM images taken from the top most surface after SMAT (a) bright field image (b) dark field image taken from the diffractions of bcc (110). Inset in (a) shows a SAED pattern.
(6.2 nm). The reason for this lies in the different measurement depths of XRD and TEM. With regard to the Cu Ka wavelength (0.154 nm) and its extinction depth in Fe, XRD patterns can offer the structure information of the surface layer of about 5 mm thick [13], while the TEM sample is only less than 0.5 mm thick. In this sense, the TEM result is smaller because the observation region is nearer to the surface.
3.2. Microstructures after irradiation The cross-sectional TEM image in Fig. 4 shows the morphology of the NC RAFM steel after irradiation at a low magnification. The boundaries of the deposited He bubble region, which are parallel to the previous SMAT-ed surface, are marked by the two dashed lines according to the locally magnified morphology of TEM images. The peak region of the deposited He in the irradiated buried layer is at a depth of 720e760 nm, which is slightly larger than the SRIM calculation results in Fig. 2 (700e740 nm). In the topmost layer regions (with width about 200 nm), equiaxed nano-sized grains are extremely small. With increasing depth from the treated surface, grains become larger grains with rectangular or elliptical shape,
Fig. 4. Cross-sectional TEM image at a low magnification about the morphology after irradiation.
which implies that the original martensite laths were deformed or even broken up during SMAT. Nearly all the long axis directions of the elliptical grains are parallel to the treated surface, which is in accordance with the results of the cross-sectional SEM image (Fig. 1b). Fig. 5 shows the typical TEM morphologies of the helium bubbles in the irradiated buried layer at a depth of 720e760 nm. The SAED pattern reveals that grains are still nanocrystalline with the BCC structure. Small dispersed helium bubbles are found in the matrix; the mean diameter of the bubbles is 0.85 nm. As marked by the filled arrows in Fig. 5, bubbles with a mean diameter of about 1.08 nm form along GBs or lath boundaries. It is worth noting that many of the bubbles lie in straight lines along GBs or lath boundaries. As marked by dotted line in Fig. 5, there are fewer bubbles in the smaller grains, which can be obviously observed.
3.3. Post-irradiation annealing Fig. 6 shows the cross-sectional TEM images of the NC RAFM steel after post-irradiation annealing at 823 K for different times. It can be seen that the width of the visible cavity region under TEM
Fig. 5. TEM morphologies of the helium bubbles both inside grains and along grain boundaries after irradiation. Inset in the figure shows a SAED pattern.
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
327
Fig. 6. Cross-sectional TEM images after post-irradiation annealing for (a) 5 min (b) 240 min. Inset in (b) shows a SAED pattern.
observation after annealing for 240 min (420 nm) is wider than that after annealing for 5 min (330 nm). Since the short-axis grain width is important for the microstructure evolution during the SMAT process [14], the short-axis grain width of the grains are considered as the grain sizes in the present work. Grains with mean short-axis length about 72 nm are found in the region about 3 mm deep from the topmost surface even after annealing for 240 min. The SAED patterns (Fig. 6b) are composed of continuous rings, but the obvious bright points imply that the grains became larger after post-irradiation annealing. Compared with the average grain size of the coarse-grained counterparts, however, the equiaxed grains (with diameters about 50 nm) are still very small in the helium deposited regions (with a width of about 400 nm) after annealing for 240 min, which shows that NC RAFM steel after irradiation also has excellent thermal stability at 823 K. 4. Discussion 4.1. Microstructure evolution during SMAT Since the initial NC microstructure produced by SMAT not only is very important for the subsequent microstructure evolution processes during He irradiation and post-irradiation annealing, but also provides a benchmark for the analysis of the subsequent microstructure, the formation of the microstructures during SMAT is discussed in this subsection. The SMAT process undergoes two stages, one is the elastic-plastic deformation stage and the other is the dynamic restoring stage [41]. The strain and strain rate gradually decrease with the increase of depth, and this is correlated to the graded microstructure from the treated surface to the strainfree matrix. Higher strain and strain rate are more effective in forming nano-grains [31]. However, there seems to be a limitation of the strain and/or strain rate for the increase of the nano-layer depth due to the balance between the elastic-plastic deformation stage and the dynamic restoring stage. Microstructure evolution and grain refinement in pure Cu subjected to SMAT show that there are two different mechanisms for plastic strain-induced grain refinement, which correspond to different levels of strain rate [42]. In the topmost surface layer with a high strain rate, grain refinement includes the formation of twins dividing the original coarse grains into twin-matrix lamellae, the subdivision of the twin-matrix lamellae into equiaxed nano-sized blocks, and the evolution of the blocks into randomly oriented nanosized grains. However, in the subsurface layer of the SMAT Cu
samples with low strain rates, grains are refined via the formation of dislocation cells, the transformation of dislocation cell walls into sub-boundaries with small misorientations, and the evolution of sub-boundaries into highly misoriented grain boundaries. In contrast, in other materials such as nickel [43], aluminum alloy [44] and iron [13], shear bands and twins have not been observed. As for RAFM steel, the mechanisms of grain refinement during SMAT are similar to that of Fe [12,13]. This process includes the formation of dense dislocation walls (DDWs) and dislocation tangles (DTs), the transformation of DDWs and DTs into sub-boundaries with small misorientations, and the evolution of sub-boundaries to highly misoriented grain boundaries. During plastic deformation, dislocation activities are motivated in the original coarse grains. The intersecting DDWs subdivide the original ferrite grains into finer blocks (or dislocation cells) [45]. More and more dislocations are formed and accumulated with increasing strains. Dislocation rearrangement and annihilation occur in dislocation walls (DWs) at a certain strain level. More dislocations are generated and annihilated in the sub-GB with further increase of strains. When the balance between dislocation multiplication rate and annihilation rate is reached, the increase of strains or duration could no longer reduce the grain size, and a stabilized grain size is obtained. As a result, equiaxed nano-sized grains are formed in the topmost surface layer. The SMAT mechanism and thermal stability of a similar quenched NC martensitic RAFM steel were discussed in our previous work [12]. 4.2. Effect of GBs during irradiation Fig. 7 shows the locally magnified TEM image about the peak region of He concentration after irradiation. Compared with the small bubbles in the matrix, bubbles with larger size and smaller density are found at GBs. However, the VDZs near GBs could not be obviously observed, and this phenomenon could be attributed to the low temperature, high dose, high dose rate and GBs character [46]. Dislocations, precipitate-matrix interfaces and GBs are regarded as strong traps for mobile He atoms [4,47]. Experimental data of He irradiation of aluminum at different temperatures shows that helium bubble nucleation on dislocations, dislocation networks and GBs, and bubbles formed on individual dislocations and at the nodal points of the dislocation networks are, on average, larger than those in the matrix [4]; further studies show that the bubbles are observed to pin migrating boundaries [47]. In the present steel, dislocations and GBs with large volume fractions are
328
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
Fig. 7. Bright field micrographs showing bubble morphologies close to grain boundaries after irradiation, (a) and (b) are underfocus and overfocus micrographs, respectively.
introduced after SMAT, which act as strong traps for mobile He atoms during irradiation. As a result, relative larger bubbles are trapped at the GBs or lath boundaries (Fig. 5a). This phenomenon is similar to the bubble morphology in irradiated NC Cu [48] or Nibased superalloys [49], and the reasons can be attributed to the
large fluxes of irradiation induced vacancies arriving at the GBs from the neighboring grains [48]. It is generally believed that the mechanisms of helium diffusion, bubble nucleation and coarsening at low temperatures (<0.2 Tm, Tm is the melting temperature) and high temperatures (>0.5 Tm) are
Fig. 8. TEM images about the cavities after post-irradiation annealing (a) (b) voids morphologies after annealing for 5 min (c) (d) voids with different sizes distributed along GBs and within grains after annealing for 240 min. Inset in (d) shows a void sizes distribution corresponding to (d).
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
different. For He diffusion under irradiation, the radiationenhanced “vacancy mechanism”, where a transient di-vacancyHe-complex is formed in which the He atom jumps from one to the other vacancy, can dominate other mechanisms at high temperatures; while the thermal “displacement or cascade mixing mechanism”, where He diffusion is due to direct displacements, is the dominant one at low temperatures [50]. The displacement mechanism may play the dominant role in the present work because of the low temperature during irradiation process. At low temperatures (<0.2 Tm), where He diffuses by the displacement mechanism, bubble nuclei formed do not exceed sizes of atomic scale (He-vacancy clusters containing up to a few tens of He atoms) [50]. In the “di-atomic nucleation” model, where two He atoms are assumed to form a stable bubble nucleus, the maximum nucleation rate is reached when a newly created He atom is as likely to reach an existing nucleus as to meet another He atom [51,52]. At a given temperature, experimental data showed that bubble nucleation has occurred in an early stage and ceased afterwards; while bubble densities and sizes increase and decrease, respectively, with increasing He production rate [53]. Due to the low implant temperature, high dose and high dose rate in the present work, numerous bubbles with small sizes distributed both on the GBs and within grains are observed after irradiation. 4.3. Microstructure evolution during post-irradiation annealing An important observation about cavity evolution during annealing in crystalline solids is that a bubble grows due to a net absorption of vacancies to its surface, and conversely shrinks due to a net absorption of interstitials e.g. annihilation of vacancies in the matrix. When a metal, into which He was introduced at low temperatures, is annealed without further He implantation at generally higher temperatures, existing bubbles or bubbles formed at the beginning of annealing tend to coarsen at constant He content, i.e., their average size increases while their density decreases [53]. During bubble coarsening upon annealing, there are two qualitatively different mechanisms: (1) bubble migration and coalescence (MC) [54,55], and (2) Ostwald ripening (OR) [56,57]. The activation energy of OR is much higher than that for MC [53], since OR is related to thermally activated re-solution from small bubbles and re-absorption of He atoms by large bubbles. MC and OR are expected to be dominant at relatively low and high temperatures, respectively. Fig. 8 shows the TEM images about the cavities in the buried layer at a depth of 720e760 nm after post-irradiation annealing. Voids with large size were formed in the peak damage region even when the annealing time is only 5 min (Fig. 8a), and the cavities were also preferred to nucleate and grow up along the GBs/interfaces (Fig. 8b). As marked by the open arrows in Fig. 8c, large cavities with hexagon shape are found on the GBs. As shown in Fig. 8d, some larger cavities with quadrilateral shape are also found. It can be seen that most of the cavities are smaller than 15 nm; but a few cavities are as large as 45 nm. Dispersed cavities with much smaller sizes are also found within some grains, which are marked as filled arrows in Fig. 8c. Because of the extremely low solubility of inert gases in metals and no replenishment of He atoms during annealing processes, MC would be the main possible mechanism of He bubble growth in the present work [55, 58]. 5. Conclusion
experiment results are summarized as follows: 1. The NC martensitic RAFM steel had exceptional radiation resistance; the NC structures are stable after irradiation with dose of 24.8 dpa. The peak value of He concentration can be as high as 18.20 at% from the calculation result of SRIM. 2. GBs and grain sizes play important roles in the bubble (void) evolution during irradiation and post-irradiation annealing. TEM observations show that He bubbles are preferentially trapped at GBs/interfaces after irradiation, but the VDZs near GBs could not be obviously observed. Voids are also preferentially trapped at GBs during post-irradiation annealing. Compared with bubbles at GBs and within larger grains, bubbles with smaller size and higher density are found in smaller grains. 3. The average size of cavities increases extensively with the increase of time during post-irradiation annealing at 823 K. Cavities with large size are observed just after annealing for 5 min, although many of the cavities with small sizes also exist after annealing for 240 min. Acknowledgements The authors are grateful to Dr. R. Wang for production of FIB samples and Dr L.N. Zhang for TEM observation. We are also indebted to Dr. Z.B. Wang of Shenyang Nation Laboratory for Materials Science for the cooperation in the SMAT experiments. Financial support from the Project Funded by China Postdoctoral Science Foundation (no. 2015M582669) and National Basic Research Programs of China (no. 2015GB118001) is acknowledged. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]
In the present work, NC structure characterization after SMAT, He bubble evolution during irradiation and post-irradiation annealing in NC martensitic RAFM steel are studied. The
329
[26] [27] [28]
K. Krishan, Rad. Eff. 66 (1982) 121e155. H. Ullmaier, Rad. Eff. 78 (1983) 1e10. P.L. Lane, P.J. Goodhew, Philos. Mag. A 48 (1983) 965e986. B.N. Singh, T. Leffers, W.V. Green, M. Victoria, J. Nucl. Mater 125 (1984) 287e297. Z. Chen, F. Liu, H.F. Wang, W. Yang, G.C. Yang, Y.H. Zhou, Acta Mater 57 (2009) 1466e1475. W. Liu, Y. Ji, P. Tan, H. Zang, C. He, D. Yun, C. Zhang, Z. Yang, Materials 9 (2016) 105. Y.I. Son, Y.K. Lee, K.T. Park, C.S. Lee, D.H. Shin, Acta Mater 53 (2005) 3125e3134. Y. Ivanisenko, W. Lojkowski, R.Z. Valiev, H.J. Fecht, Acta Mater 51 (2003) 5555e5570. N. Kamikawa, N. Tsuji, X. Huang, N. Hansen, Acta Mater 54 (2006) 3055e3066. Y.S. Li, N.R. Tao, K. Lu, Acta Mater 56 (2008) 230e241. W. Li, W. Xu, X. Wang, Y. Rong, J. Alloys Compd 474 (2009) 546e550. W.B. Liu, C. Zhang, Z.G. Yang, Z.X. Xia, Mater. Sci. Eng. A 583 (2013) 61e68. N.R. Tao, Z.B. Wang, W.P. Tong, M.L. Sui, J. Lu, K. Lu, Acta Mater 50 (2002) 4603e4616. L. Zhou, G. Liu, X.L. Ma, K. Lu, Acta Mater 56 (2008) 78e87. B. Schaaf, D.S. Gelles, S. Jitsukawa, A. Kimura, R.L. Klueh, A. Moslang, G.R. Odette, J. Nucl. Mater 283 (2000) 52e59. L. Huang, X. Hu, W. Yan, W. Sha, F. Xiao, Y. Shan, K. Yang, Mater. Des. 63 (2014) 333e335. C.H. Zhang, J. Jang, M.C. Kim, H.D. Cho, Y.T. Yang, Y.M. Sun, J. Nucl. Mater 375 (2008) 185e191. J. Henry, M.-H. Mathon, P. Jung, J. Nucl. Mater 318 (2003) 249e259. Z. Jiao, N. Ham, G. Was, J. Nucl. Mater 367 (2007) 440e445. Y. Sekio, S. Yamashita, N. Sakaguchi, H. Takahashi, J. Nucl. Mater 458 (2015) 355e360. B. Mazumder, M.E. Bannister, F.W. Meyer, M.K. Miller, C.M. Parish, P.D. Edmondson, Nucl. Mater. Energy 1 (2015) 8e12. € chbauer, A. Misra, X. Zhang, N. Li, O. Anderoglu, H. Wang, J.G. Swadener, T. Ho R.G. Hoagland, Nucl. Instrum. Method B 261 (2007) 1129e1132. T. Hochbauer, A. Misra, K. Hattar, R.G. Hoagland, J. Appl. Phys. 98 (2005) 123516. M.J. Demkowicz, D. Bhattacharyya, I. Usov, Y.Q. Wang, M. Nastasi, A. Misra, Appl. Phys. Lett. 97 (2010) 161903. K.Y. Yu, Y. Liu, C. Sun, H. Wang, L. Shao, E.G. Fu, X. Zhang, J. Nucl. Mater 425 (2012) 140e146. G.R. Odette, D.T. Hoelzer, JOM 62 (2010) 84e92. L.K. Mansur, W.A. Coghlan, J. Nucl. Mater 119 (1983) 1e25. R.E. Stoller, G.R. Odette, J. Nucl. Mater 131 (1985) 118e125.
330
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
[29] S.J. Zinkle, Radiation-induced effects on microstructure, in: R.J.M. Konings (Ed.), Comprehensive Nuclear Materials, vol. 1, 2012, pp. 65e98. [30] G. Liu, S.C. Wang, X.F. Lou, J. Lu, K. Lu, Scr. Mater 44 (2001) 1791e1795. [31] K. Lu, J. Lu, Mater. Sci. Eng. A 375e377 (2004) 38e45. [32] N. Hansen, Mater. Sci. Technol. 6 (1990) 1039e1047. [33] S.J. Zinkle, J.T. Busby, Structural materials for fission & fusion energy, Mater. Today 12 (11) (2009) 12e19. [34] http://www.srim.org. [35] R.L. Klueh, D.J. Alexander, M.A. Sokolov, J. Nucl. Mater 304 (2002) 139e152. [36] S.J. Zinkle, N.M. Ghoniem, Fusion. Eng. Des. 51 (2000) 55e71. [37] W. Liu, C. Zhang, Z. Yang, Z. Xia, G. Gao, Y. Weng, Acta Metall. Sin. 49 (2013) 707e716. [38] W. Liu, C. Zhang, Z. Yang, Z. Xia, Appl. Surf. Sci. 292 (2014) 556e562. [39] H.P. Klug, L.E. Alexander, X-ray Diffraction Procedures for Polycrystalline and Amorphous Materials, Wiley, New York, 1974, p. 662. [40] R. Wirth, Chem. Geol. 261 (2009) 217e229. [41] H. Du, Y. Wei, H. Zhang, W. Lin, L. Hou, Z. Liu, Int. J. Mod. Phys. B 23 (2009) 1924e1930. [42] K. Wang, N.R. Tao, G. Liu, J. Lu, K. Lu, Acta Mater 54 (2006) 5281e5291. [43] Z.P. Luo, H.W. Zhang, N. Hansen, K. Lu, Acta Mater 60 (2012) 1322e1333.
[44] X. Wu, N. Tao, Y. Hong, B. Xu, J. Lu, K. Lu, Acta Mater 50 (2002) 2075e2084. [45] K. Lu, J. Lu, Mater. Sci. Eng. A 375e377 (2004) 38e45. [46] W. Han, M. Demkowicz, E. Fu, Y. Wang, A. Misra, Acta Mater 60 (2012) 6341e6351. [47] B.N. Singh, T. Leffers, W.V. Green, M. Victoria, J. Nucl. Mater. 122&123 (1984) 703e708. [48] W. Han, E.G. Fu, M.J. Demkowicz, Y. Wang, A. Misra, J. Mater. Res. 28 (2013) 2763e2770. [49] H. Zhang, Z. Yao, M.R. Daymond, M.A. Kirk, J. Appl. Phys. 115 (2014) 103508. [50] H. Trinkaus, B.N. Singh, J. Nucl. Mater 323 (2003) 229e242. [51] H. Trinkaus, J. Nucl. Mater 118 (1983) 39e49. [52] C.H. Zhang, J. Jang, M.C. Kim, H.D. Cho, Y.T. Yang, Y.M. Sun, J. Nucl. Mater 375 (2008) 185e191. [53] B.N. Singh, H. Trinkaus, J. Nucl. Mater 186 (1992) 153e165. [54] E.E. Gruber, J. Appl. Phys. 38 (1967) 243e250. [55] P.J. Goodhew, S.K. Tyler, Proc. Roy. Soc. Lond. A 377 (1981) 151e184. [56] G.W. Greenwood, A. Boltax, J. Nucl. Mater 5 (1962) 234e240. [57] A.J. Markworth, Met. Trans. 4 (1973) 2651e2656. [58] R. Blackburn, Metall. Rev. 11 (1966) 159e176.
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Microstructure evolution during helium irradiation and postirradiation annealing in a nanostructured reduced activation steel W.B. Liu a, c, *, Y.Z. Ji b, P.K. Tan a, C. Zhang c, C.H. He a, Z.G. Yang c a
Department of Nuclear Science and Technology, Xi'an Jiaotong University, Xi'an 710049, China Department of Materials Science and Engineering, The Pennsylvania State University, University Park, PA 16802, USA c Key Laboratory of Advanced Materials of Ministry of Education, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, China b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 8 January 2016 Received in revised form 1 July 2016 Accepted 11 July 2016 Available online 14 July 2016
Severe plastic deformation, intense single-beam He-ion irradiation and post-irradiation annealing were performed on a nanostructured reduced activation ferritic/martensitic (RAFM) steel to investigate the effect of grain boundaries (GBs) on its microstructure evolution during these processes. A surface layer with a depth-dependent nanocrystalline (NC) microstructure was prepared in the RAFM steel using surface mechanical attrition treatment (SMAT). Microstructure evolution after helium (He) irradiation (24.8 dpa) at room temperature and after post-irradiation annealing was investigated using Transmission Electron Microscopy (TEM). Experimental observation shows that GBs play an important role during both the irradiation and the post-irradiation annealing process. He bubbles are preferentially trapped at GBs/ interfaces during irradiation and cavities with large sizes are also preferentially trapped at GBs/interfaces during post-irradiation annealing, but void denuded zones (VDZs) near GBs could not be unambiguously observed. Compared with cavities at GBs and within larger grains, cavities with smaller size and higher density are found in smaller grains. The average size of cavities increases rapidly with the increase of time during post-irradiation annealing at 823 K. Cavities with a large size are observed just after annealing for 5 min, although many of the cavities with small sizes also exist after annealing for 240 min. The potential mechanism of cavity growth behavior during post-irradiation annealing is also discussed. © 2016 Elsevier B.V. All rights reserved.
Keywords: Nanocrystalline Ion irradiation Helium bubble Post-irradiation annealing Reduced activation steel
1. Introduction Helium (He) irradiation behavior in metals has been the subject of extensive research in the past several decades [1,2]. For example, investigation by Lane et al. [3] showed that misfit dislocations at grain boundaries (GBs) are preferred nucleation sites of He bubbles in austenitic steels. Furthermore, Singh et al. [4] confirmed that dislocation intersections, both within and outside of GBs, are preferred bubble nucleation sites in aluminum. Since GBs/interfaces are known to act as potential point defect sinks, and nanocrystalline (NC) material has a large volume fraction of GBs [5], He bubbles can be effectively trapped at GBs in NC materials. Therefore, NC materials are likely to have different radiation damage tolerance from their coarse-grained counterparts [6]. Severe plastic deformation (SPD) technology, such as equal
* Corresponding author. Department of Nuclear Science and Technology, Xi'an Jiaotong University, Xi'an 710049, China. E-mail address: [email protected] (W.B. Liu). http://dx.doi.org/10.1016/j.jnucmat.2016.07.030 0022-3115/© 2016 Elsevier B.V. All rights reserved.
channel angular pressing (ECAP) [7], high-pressure torsion (HPT) [8] and accumulated roll bonding (ARB) [9], has been recognized as an effective way to refine microstructures of various materials. Surface mechanical attrition treatment (SMAT), sometimes called ultrasonic shot peening (USSP), is regarded as one of the most important means of SPD, and it has been employed to obtain various NC materials, including pure metals, alloys and intermetallics [10e12]. Experimental results show that large amounts of GBs/interfaces and dislocations are introduced into the matrix during SMAT [13,14], which are likely to improve the radiation resistance [6]. Due to its excellent properties [15,16], reduced activation ferrite/ martensitic (RAFM) steels (Fe-Cr-W-V-Ta) have potential applications as first wall and blanket structural materials in fusion reactors, and a lot of research has been done to investigate the microstructural evolution during He ion irradiations. Studies of Neions irradiated 9Cr ferrite/martensitic steel at high temperatures showed that growth of cavities at GBs, especially at GB junctions, was faster than that within the grain [17]. Microstructural
324
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
observations of high temperature (550 C) helium irradiated 9Cr1Mo martensitic steel [18] also showed that voids were preferred to form on prior austenite GBs and other interfaces, such as carbidematrix interfaces, sub-GBs, lath boundaries and dislocations inside the lath structures, while no preferential nucleation site was observed when the helium implantation was performed at room temperature (RT) [19]. In addition, a typical void distribution near a coincident site lattice boundary was observed in neutron irradiated Fe-15Cr-15Ni steel neutron-irradiated at 749 K to 18 dpa, although void denuded zones (VDZs) were formed near the random GBs [20]. Preferential He bubble formation in carbides of a tempered F82H ferritic-martensitic steel during low temperature He irradiation was observed recently, and this phenomenon was attributed to the diffusing He being trapped in the carbide due to the strong He-C bond, which can lead to the formation of He bubbles with the increase of He concentration in the carbides [21]. However, there is still a lack of research about the effect of GBs in RAFM steels on irradiation induced microstructural evolution. In the past decade, a number of studies have been devoted to investigating the role of interface structures in the damage resistant properties of NC materials [22e24]. Studies of He ion-irradiations on NC Cu/Nb multilayers show that, in contrast to the copious bubbles contained in the He-implanted pure Cu and Nb, the Cu-Nb interfaces in the Cu/Nb multilayers are responsible for suppressing He bubble formation [22,23]. Nevertheless, when He is preferentially trapped at interfaces, there exists a critical He concentration below which bubbles were not observed [24]. Experimental results of radiation damage in He ion irradiated NC iron (Fe), which was produced by a deposition method using the magnetron sputtering technique, showed that smaller grains lead to lower density of He bubbles, and no voids were observed after He implantation [25]. However, a study about He ion irradiation in NC RAFM steel is still needed to clarify the effect of GBs on irradiation induced microstructural evolution, since the NC microstructure produced by severe plastic deformation is critically different from that produced by magnetron sputtering technique. It is reported that voids and helium-filled bubbles have different effects on He-induced embrittlement: voids are the main reason for swelling and high temperature embrittlement, while the effect of helium-filled bubbles is comparatively small or, in some cases, even beneficial [26]. The “bubble” and “void” can be described as [27,28]: under some conditions, a helium-filled cavity with diameter less than ~ 10 nm, neither grows nor shrinks, but remains stable as a “He bubble”; while under other conditions, it grows without bounds by absorbing vacancies, and finally results in the formation of a “void”. Therefore, accurate experimental studies of microstructural evolution during ion irradiation are necessary, and this is one of the main objectives of the present work. The stability of a void is not dependent on the presence of internal pressurization from a gaseous species such as helium, and bubbles are usually defined as pressurized cavities, while cavity can be used to refer to either bubbles or voids [29]. Hence, it is necessary to study the average size evolution of cavities during irradiation and postirradiation annealing. In the present contribution, He bubble evolution during irradiation and post-irradiation annealing in a quenched fully martensitic NC RAFM steel is systematically studied. Section 2 deals with experimental details about materials, the SMAT process, He ion irradiation, annealing treatment and all the characterization methods. The experimental results, concerning NC structure characterization after SMAT, bubble formation and microstructural evolution under He irradiation, and void formation upon postirradiation annealing, are presented in Section 3. In Section 4, the mechanisms of microstructural evolution during SMAT, He bubble formation during low temperature implantation, especially the
effect of GBs/interfaces during irradiation and post-irradiation annealing, are discussed. 2. Experiments 2.1. Material and SMAT process The material used in the present investigation was a RAFM steel with chemical composition (in wt%): 0.09% C, 0.49% Mn, 8.75% Cr, 0.84% Ta, 1.58% W, 0.21% V and balance Fe. Heat treatment included austenitizing at 1253 K for 45 min, followed by water quenching. It is noted that the steel, which was quenched after austenitization but not tempered, produced a fully martensitic microstructure, and the microstructure without M23C6 type carbides was used to study the effect of GBs on irradiation-induced microstructures evolution. The plate sample (F50 4.0 mm in size) of the quenched steel was submitted to SMAT. The equipment and procedure of SMAT are discussed in detail in Refs. [30,31]. The ball size used for surface peening was 5 mm in diameter. The vibration frequency of the chamber was 20 kHz, and the plate was treated for 30 min. During SMAT, the strain and strain rate reduced rapidly with the increase of depth, and the strain rate at the sample surface was estimated to as high as 102e103 s1 [31]. Fig. 1 shows the distinct microstructures of the surface layer and the matrix obtained from cross-sectional SEM observations. No carbides can be obviously seen both in the SMAT layer and in the matrix from the SEM results. Grains with no deformation and grain boundaries can be clearly seen in the matrix (Fig. 1a), and most of the grains are smaller than 10 mm. With increasing depth from the treated surface, gradient nano-submicron-microstructure and obvious deformation bands (Fig. 1b) are found due to the decreasing strain and strain rate. In the topmost layer regions, grains are too small to distinguish their boundaries. Microbands are clearly observed in the inner area. The thickness of the surface deformation layer in Fig. 1b is not uniform (30e46 mm), which can be attributed to the heterogeneous nature of plastic deformation within and between grains [32]. 2.2. He-ion irradiation The He-ion irradiation experiment was performed on the 320 kV ECR (electron cyclone resonance) experimental platform in the National Laboratory of Heavy-ion Accelerators in Lanzhou, China. He-ions with a kinetic energy of 300 keV were used for irradiation. The irradiation experiment was conducted at room temperature with vacuum pressure about 106 Pa. The scanned beam size was about 17 17 mm2. The RAFM steel sample was irradiated with a fluence of 3.31 1017 ion/cm2, which corresponds to the estimated displacement levels of 24.8 dpa (displacement per atom) at peak damage region. The average beam current was about 3.83 1012 ion/cm2/s during the irradiation experiment. Displacement damage dose regimes for structural materials (Ferrite/Martensitic steel) in fusion energy systems can be larger than 100 dpa [33], and basic research is still needed to improve understanding of high dose irradiation-induced microstructure evolution. Hence, dose of ~25 dpa was selected in the present work to investigate microstructure evolution in NC RAFM steel under helium ion irradiation. Fig. 2 shows the dose and deposited He concentration calculated from the stopping and range of ions in matter 2013 (SRIM-2013) software [34]. Materials with the same chemical composition in the experiment and He2þ ions with energy of 300 keV were used in the calculation. The peak value of He concentration can be as high as 18.20 at%, and the peak region of the deposited He in the irradiated
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
325
Fig. 1. Cross-sectional SEM images after SMAT (a) about the matrix (untreated area) (b) about the surface (SMAT-treated) layer.
30
20
25
15
Dose (dpa)
20 10
15 10
5 5 0 0
200
400
600
800
0 1000
Deposited He concentration (at%)
Dose Deposited He concentration
Depth (nm) Fig. 2. Dose and deposited He concentration calculated from SRIM software.
buried layer is at a depth of 700e740 nm, which is much smaller than the thickness of nanocrystalline layer produced by SMAT. 2.3. Post-irradiation annealing It is reported that the highest operating temperature (based on thermal creep and radiation damage considerations) for the RAFM steel is 773 Ke873 K [35,36], and the nanocrystalline RAFM steel has excellent thermal stability up to that temperature range [37,38]. In order to investigate the thermal stability of the NC structures and helium bubbles after irradiation, SMAT-ed samples subjected to irradiation were annealed at 823 K for 5 min and 240 min. To avoid oxidation, the samples used for annealing were sealed in glass tubes under 0.3 atmosphere of argon. 2.4. Microstructural examination X-ray diffraction (XRD) analysis of the surface layer was carried out to measure the phase and grain sizes. Step size of 0.02 and range of 2q from 20 to 100 were taken to measure the XRD intensity before and after SMAT. Average grain size in the SMAT surface layer were derived from the full width at half maximum intensities of XRD peaks using Scherrer-Wilson equation [39], and
the measured (200) and (211) peaks were selected. Step size and holding time per step in the step scanning mode were 0.01 and 2 s, respectively. Cross-sectional observation of the treated sample was performed on a scanning electron microscope (SEM) of type JEOL JSM4500. The samples were ground and mechanically polished. The polished samples were etched in a particular solution (12 ml alcohol þ 3 ml hydrochloric acid þ 1 g ferric trichloride) to obtain the microstructure and morphology of the matrix and treated surface layer. Cross-sectional observation of the NC RAFM steel after irradiation was performed on transmission electron microscopes (TEM), JEOL JEM-2011 and Tecnai F20. The cross-sectional TEM samples were obtained by means of focused ion beam (FIB), which is an advanced analytical tool of sample preparation for studying microstructure and crystal structure on nanometer scales [40]. In the present work, an electrical field (30 kV) was used to accelerate the Ga ions; the corresponding current was about 1 nA. Firstly, a platinum (Pt) strip was deposited on the surface where the foil is to be cut. Then two trenches, one in front of the foil and the other behind the foil, were sputtered. At last, a TEM foil with dimensions of 5 10 0.15 mm was lifted out. The final thickness of the TEM foil was approximately 40 nm after being milled and polished under smaller accelerate electrical field (5 kV) and current (100 pA). 3. Results 3.1. Microstructures after SMAT A TEM bright field image obtained from the topmost surface of the SMAT sample and the corresponding dark field image taken using the BCC (110) diffraction spot are shown in Fig. 3. The TEM sample was obtained by means of cutting, grinding and thinning from the un-SMAT treated side at low temperatures. Selected area electron diffraction (SAED) patterns, which are composed of continuous rings, show that these grains are almost nano-grained bcc martensite (Fig. 3a). Grains of the original sample have been broken down to equiaxed nano-sized grains and the grains are random in crystallographic orientation (Fig. 3b). The average grain size calculated by statistical analysis is 6.20 nm. Most of the grains are smaller than 6 nm, although the largest grains are about 10 nm in diameter. However, the average grain size in the SMAT surface layer, which is derived from the full width at half maximum intensity of XRD peaks, is 11.5 nm, larger than the result from TEM
326
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
Fig. 3. TEM images taken from the top most surface after SMAT (a) bright field image (b) dark field image taken from the diffractions of bcc (110). Inset in (a) shows a SAED pattern.
(6.2 nm). The reason for this lies in the different measurement depths of XRD and TEM. With regard to the Cu Ka wavelength (0.154 nm) and its extinction depth in Fe, XRD patterns can offer the structure information of the surface layer of about 5 mm thick [13], while the TEM sample is only less than 0.5 mm thick. In this sense, the TEM result is smaller because the observation region is nearer to the surface.
3.2. Microstructures after irradiation The cross-sectional TEM image in Fig. 4 shows the morphology of the NC RAFM steel after irradiation at a low magnification. The boundaries of the deposited He bubble region, which are parallel to the previous SMAT-ed surface, are marked by the two dashed lines according to the locally magnified morphology of TEM images. The peak region of the deposited He in the irradiated buried layer is at a depth of 720e760 nm, which is slightly larger than the SRIM calculation results in Fig. 2 (700e740 nm). In the topmost layer regions (with width about 200 nm), equiaxed nano-sized grains are extremely small. With increasing depth from the treated surface, grains become larger grains with rectangular or elliptical shape,
Fig. 4. Cross-sectional TEM image at a low magnification about the morphology after irradiation.
which implies that the original martensite laths were deformed or even broken up during SMAT. Nearly all the long axis directions of the elliptical grains are parallel to the treated surface, which is in accordance with the results of the cross-sectional SEM image (Fig. 1b). Fig. 5 shows the typical TEM morphologies of the helium bubbles in the irradiated buried layer at a depth of 720e760 nm. The SAED pattern reveals that grains are still nanocrystalline with the BCC structure. Small dispersed helium bubbles are found in the matrix; the mean diameter of the bubbles is 0.85 nm. As marked by the filled arrows in Fig. 5, bubbles with a mean diameter of about 1.08 nm form along GBs or lath boundaries. It is worth noting that many of the bubbles lie in straight lines along GBs or lath boundaries. As marked by dotted line in Fig. 5, there are fewer bubbles in the smaller grains, which can be obviously observed.
3.3. Post-irradiation annealing Fig. 6 shows the cross-sectional TEM images of the NC RAFM steel after post-irradiation annealing at 823 K for different times. It can be seen that the width of the visible cavity region under TEM
Fig. 5. TEM morphologies of the helium bubbles both inside grains and along grain boundaries after irradiation. Inset in the figure shows a SAED pattern.
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
327
Fig. 6. Cross-sectional TEM images after post-irradiation annealing for (a) 5 min (b) 240 min. Inset in (b) shows a SAED pattern.
observation after annealing for 240 min (420 nm) is wider than that after annealing for 5 min (330 nm). Since the short-axis grain width is important for the microstructure evolution during the SMAT process [14], the short-axis grain width of the grains are considered as the grain sizes in the present work. Grains with mean short-axis length about 72 nm are found in the region about 3 mm deep from the topmost surface even after annealing for 240 min. The SAED patterns (Fig. 6b) are composed of continuous rings, but the obvious bright points imply that the grains became larger after post-irradiation annealing. Compared with the average grain size of the coarse-grained counterparts, however, the equiaxed grains (with diameters about 50 nm) are still very small in the helium deposited regions (with a width of about 400 nm) after annealing for 240 min, which shows that NC RAFM steel after irradiation also has excellent thermal stability at 823 K. 4. Discussion 4.1. Microstructure evolution during SMAT Since the initial NC microstructure produced by SMAT not only is very important for the subsequent microstructure evolution processes during He irradiation and post-irradiation annealing, but also provides a benchmark for the analysis of the subsequent microstructure, the formation of the microstructures during SMAT is discussed in this subsection. The SMAT process undergoes two stages, one is the elastic-plastic deformation stage and the other is the dynamic restoring stage [41]. The strain and strain rate gradually decrease with the increase of depth, and this is correlated to the graded microstructure from the treated surface to the strainfree matrix. Higher strain and strain rate are more effective in forming nano-grains [31]. However, there seems to be a limitation of the strain and/or strain rate for the increase of the nano-layer depth due to the balance between the elastic-plastic deformation stage and the dynamic restoring stage. Microstructure evolution and grain refinement in pure Cu subjected to SMAT show that there are two different mechanisms for plastic strain-induced grain refinement, which correspond to different levels of strain rate [42]. In the topmost surface layer with a high strain rate, grain refinement includes the formation of twins dividing the original coarse grains into twin-matrix lamellae, the subdivision of the twin-matrix lamellae into equiaxed nano-sized blocks, and the evolution of the blocks into randomly oriented nanosized grains. However, in the subsurface layer of the SMAT Cu
samples with low strain rates, grains are refined via the formation of dislocation cells, the transformation of dislocation cell walls into sub-boundaries with small misorientations, and the evolution of sub-boundaries into highly misoriented grain boundaries. In contrast, in other materials such as nickel [43], aluminum alloy [44] and iron [13], shear bands and twins have not been observed. As for RAFM steel, the mechanisms of grain refinement during SMAT are similar to that of Fe [12,13]. This process includes the formation of dense dislocation walls (DDWs) and dislocation tangles (DTs), the transformation of DDWs and DTs into sub-boundaries with small misorientations, and the evolution of sub-boundaries to highly misoriented grain boundaries. During plastic deformation, dislocation activities are motivated in the original coarse grains. The intersecting DDWs subdivide the original ferrite grains into finer blocks (or dislocation cells) [45]. More and more dislocations are formed and accumulated with increasing strains. Dislocation rearrangement and annihilation occur in dislocation walls (DWs) at a certain strain level. More dislocations are generated and annihilated in the sub-GB with further increase of strains. When the balance between dislocation multiplication rate and annihilation rate is reached, the increase of strains or duration could no longer reduce the grain size, and a stabilized grain size is obtained. As a result, equiaxed nano-sized grains are formed in the topmost surface layer. The SMAT mechanism and thermal stability of a similar quenched NC martensitic RAFM steel were discussed in our previous work [12]. 4.2. Effect of GBs during irradiation Fig. 7 shows the locally magnified TEM image about the peak region of He concentration after irradiation. Compared with the small bubbles in the matrix, bubbles with larger size and smaller density are found at GBs. However, the VDZs near GBs could not be obviously observed, and this phenomenon could be attributed to the low temperature, high dose, high dose rate and GBs character [46]. Dislocations, precipitate-matrix interfaces and GBs are regarded as strong traps for mobile He atoms [4,47]. Experimental data of He irradiation of aluminum at different temperatures shows that helium bubble nucleation on dislocations, dislocation networks and GBs, and bubbles formed on individual dislocations and at the nodal points of the dislocation networks are, on average, larger than those in the matrix [4]; further studies show that the bubbles are observed to pin migrating boundaries [47]. In the present steel, dislocations and GBs with large volume fractions are
328
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
Fig. 7. Bright field micrographs showing bubble morphologies close to grain boundaries after irradiation, (a) and (b) are underfocus and overfocus micrographs, respectively.
introduced after SMAT, which act as strong traps for mobile He atoms during irradiation. As a result, relative larger bubbles are trapped at the GBs or lath boundaries (Fig. 5a). This phenomenon is similar to the bubble morphology in irradiated NC Cu [48] or Nibased superalloys [49], and the reasons can be attributed to the
large fluxes of irradiation induced vacancies arriving at the GBs from the neighboring grains [48]. It is generally believed that the mechanisms of helium diffusion, bubble nucleation and coarsening at low temperatures (<0.2 Tm, Tm is the melting temperature) and high temperatures (>0.5 Tm) are
Fig. 8. TEM images about the cavities after post-irradiation annealing (a) (b) voids morphologies after annealing for 5 min (c) (d) voids with different sizes distributed along GBs and within grains after annealing for 240 min. Inset in (d) shows a void sizes distribution corresponding to (d).
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
different. For He diffusion under irradiation, the radiationenhanced “vacancy mechanism”, where a transient di-vacancyHe-complex is formed in which the He atom jumps from one to the other vacancy, can dominate other mechanisms at high temperatures; while the thermal “displacement or cascade mixing mechanism”, where He diffusion is due to direct displacements, is the dominant one at low temperatures [50]. The displacement mechanism may play the dominant role in the present work because of the low temperature during irradiation process. At low temperatures (<0.2 Tm), where He diffuses by the displacement mechanism, bubble nuclei formed do not exceed sizes of atomic scale (He-vacancy clusters containing up to a few tens of He atoms) [50]. In the “di-atomic nucleation” model, where two He atoms are assumed to form a stable bubble nucleus, the maximum nucleation rate is reached when a newly created He atom is as likely to reach an existing nucleus as to meet another He atom [51,52]. At a given temperature, experimental data showed that bubble nucleation has occurred in an early stage and ceased afterwards; while bubble densities and sizes increase and decrease, respectively, with increasing He production rate [53]. Due to the low implant temperature, high dose and high dose rate in the present work, numerous bubbles with small sizes distributed both on the GBs and within grains are observed after irradiation. 4.3. Microstructure evolution during post-irradiation annealing An important observation about cavity evolution during annealing in crystalline solids is that a bubble grows due to a net absorption of vacancies to its surface, and conversely shrinks due to a net absorption of interstitials e.g. annihilation of vacancies in the matrix. When a metal, into which He was introduced at low temperatures, is annealed without further He implantation at generally higher temperatures, existing bubbles or bubbles formed at the beginning of annealing tend to coarsen at constant He content, i.e., their average size increases while their density decreases [53]. During bubble coarsening upon annealing, there are two qualitatively different mechanisms: (1) bubble migration and coalescence (MC) [54,55], and (2) Ostwald ripening (OR) [56,57]. The activation energy of OR is much higher than that for MC [53], since OR is related to thermally activated re-solution from small bubbles and re-absorption of He atoms by large bubbles. MC and OR are expected to be dominant at relatively low and high temperatures, respectively. Fig. 8 shows the TEM images about the cavities in the buried layer at a depth of 720e760 nm after post-irradiation annealing. Voids with large size were formed in the peak damage region even when the annealing time is only 5 min (Fig. 8a), and the cavities were also preferred to nucleate and grow up along the GBs/interfaces (Fig. 8b). As marked by the open arrows in Fig. 8c, large cavities with hexagon shape are found on the GBs. As shown in Fig. 8d, some larger cavities with quadrilateral shape are also found. It can be seen that most of the cavities are smaller than 15 nm; but a few cavities are as large as 45 nm. Dispersed cavities with much smaller sizes are also found within some grains, which are marked as filled arrows in Fig. 8c. Because of the extremely low solubility of inert gases in metals and no replenishment of He atoms during annealing processes, MC would be the main possible mechanism of He bubble growth in the present work [55, 58]. 5. Conclusion
experiment results are summarized as follows: 1. The NC martensitic RAFM steel had exceptional radiation resistance; the NC structures are stable after irradiation with dose of 24.8 dpa. The peak value of He concentration can be as high as 18.20 at% from the calculation result of SRIM. 2. GBs and grain sizes play important roles in the bubble (void) evolution during irradiation and post-irradiation annealing. TEM observations show that He bubbles are preferentially trapped at GBs/interfaces after irradiation, but the VDZs near GBs could not be obviously observed. Voids are also preferentially trapped at GBs during post-irradiation annealing. Compared with bubbles at GBs and within larger grains, bubbles with smaller size and higher density are found in smaller grains. 3. The average size of cavities increases extensively with the increase of time during post-irradiation annealing at 823 K. Cavities with large size are observed just after annealing for 5 min, although many of the cavities with small sizes also exist after annealing for 240 min. Acknowledgements The authors are grateful to Dr. R. Wang for production of FIB samples and Dr L.N. Zhang for TEM observation. We are also indebted to Dr. Z.B. Wang of Shenyang Nation Laboratory for Materials Science for the cooperation in the SMAT experiments. Financial support from the Project Funded by China Postdoctoral Science Foundation (no. 2015M582669) and National Basic Research Programs of China (no. 2015GB118001) is acknowledged. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]
In the present work, NC structure characterization after SMAT, He bubble evolution during irradiation and post-irradiation annealing in NC martensitic RAFM steel are studied. The
329
[26] [27] [28]
K. Krishan, Rad. Eff. 66 (1982) 121e155. H. Ullmaier, Rad. Eff. 78 (1983) 1e10. P.L. Lane, P.J. Goodhew, Philos. Mag. A 48 (1983) 965e986. B.N. Singh, T. Leffers, W.V. Green, M. Victoria, J. Nucl. Mater 125 (1984) 287e297. Z. Chen, F. Liu, H.F. Wang, W. Yang, G.C. Yang, Y.H. Zhou, Acta Mater 57 (2009) 1466e1475. W. Liu, Y. Ji, P. Tan, H. Zang, C. He, D. Yun, C. Zhang, Z. Yang, Materials 9 (2016) 105. Y.I. Son, Y.K. Lee, K.T. Park, C.S. Lee, D.H. Shin, Acta Mater 53 (2005) 3125e3134. Y. Ivanisenko, W. Lojkowski, R.Z. Valiev, H.J. Fecht, Acta Mater 51 (2003) 5555e5570. N. Kamikawa, N. Tsuji, X. Huang, N. Hansen, Acta Mater 54 (2006) 3055e3066. Y.S. Li, N.R. Tao, K. Lu, Acta Mater 56 (2008) 230e241. W. Li, W. Xu, X. Wang, Y. Rong, J. Alloys Compd 474 (2009) 546e550. W.B. Liu, C. Zhang, Z.G. Yang, Z.X. Xia, Mater. Sci. Eng. A 583 (2013) 61e68. N.R. Tao, Z.B. Wang, W.P. Tong, M.L. Sui, J. Lu, K. Lu, Acta Mater 50 (2002) 4603e4616. L. Zhou, G. Liu, X.L. Ma, K. Lu, Acta Mater 56 (2008) 78e87. B. Schaaf, D.S. Gelles, S. Jitsukawa, A. Kimura, R.L. Klueh, A. Moslang, G.R. Odette, J. Nucl. Mater 283 (2000) 52e59. L. Huang, X. Hu, W. Yan, W. Sha, F. Xiao, Y. Shan, K. Yang, Mater. Des. 63 (2014) 333e335. C.H. Zhang, J. Jang, M.C. Kim, H.D. Cho, Y.T. Yang, Y.M. Sun, J. Nucl. Mater 375 (2008) 185e191. J. Henry, M.-H. Mathon, P. Jung, J. Nucl. Mater 318 (2003) 249e259. Z. Jiao, N. Ham, G. Was, J. Nucl. Mater 367 (2007) 440e445. Y. Sekio, S. Yamashita, N. Sakaguchi, H. Takahashi, J. Nucl. Mater 458 (2015) 355e360. B. Mazumder, M.E. Bannister, F.W. Meyer, M.K. Miller, C.M. Parish, P.D. Edmondson, Nucl. Mater. Energy 1 (2015) 8e12. € chbauer, A. Misra, X. Zhang, N. Li, O. Anderoglu, H. Wang, J.G. Swadener, T. Ho R.G. Hoagland, Nucl. Instrum. Method B 261 (2007) 1129e1132. T. Hochbauer, A. Misra, K. Hattar, R.G. Hoagland, J. Appl. Phys. 98 (2005) 123516. M.J. Demkowicz, D. Bhattacharyya, I. Usov, Y.Q. Wang, M. Nastasi, A. Misra, Appl. Phys. Lett. 97 (2010) 161903. K.Y. Yu, Y. Liu, C. Sun, H. Wang, L. Shao, E.G. Fu, X. Zhang, J. Nucl. Mater 425 (2012) 140e146. G.R. Odette, D.T. Hoelzer, JOM 62 (2010) 84e92. L.K. Mansur, W.A. Coghlan, J. Nucl. Mater 119 (1983) 1e25. R.E. Stoller, G.R. Odette, J. Nucl. Mater 131 (1985) 118e125.
330
W.B. Liu et al. / Journal of Nuclear Materials 479 (2016) 323e330
[29] S.J. Zinkle, Radiation-induced effects on microstructure, in: R.J.M. Konings (Ed.), Comprehensive Nuclear Materials, vol. 1, 2012, pp. 65e98. [30] G. Liu, S.C. Wang, X.F. Lou, J. Lu, K. Lu, Scr. Mater 44 (2001) 1791e1795. [31] K. Lu, J. Lu, Mater. Sci. Eng. A 375e377 (2004) 38e45. [32] N. Hansen, Mater. Sci. Technol. 6 (1990) 1039e1047. [33] S.J. Zinkle, J.T. Busby, Structural materials for fission & fusion energy, Mater. Today 12 (11) (2009) 12e19. [34] http://www.srim.org. [35] R.L. Klueh, D.J. Alexander, M.A. Sokolov, J. Nucl. Mater 304 (2002) 139e152. [36] S.J. Zinkle, N.M. Ghoniem, Fusion. Eng. Des. 51 (2000) 55e71. [37] W. Liu, C. Zhang, Z. Yang, Z. Xia, G. Gao, Y. Weng, Acta Metall. Sin. 49 (2013) 707e716. [38] W. Liu, C. Zhang, Z. Yang, Z. Xia, Appl. Surf. Sci. 292 (2014) 556e562. [39] H.P. Klug, L.E. Alexander, X-ray Diffraction Procedures for Polycrystalline and Amorphous Materials, Wiley, New York, 1974, p. 662. [40] R. Wirth, Chem. Geol. 261 (2009) 217e229. [41] H. Du, Y. Wei, H. Zhang, W. Lin, L. Hou, Z. Liu, Int. J. Mod. Phys. B 23 (2009) 1924e1930. [42] K. Wang, N.R. Tao, G. Liu, J. Lu, K. Lu, Acta Mater 54 (2006) 5281e5291. [43] Z.P. Luo, H.W. Zhang, N. Hansen, K. Lu, Acta Mater 60 (2012) 1322e1333.
[44] X. Wu, N. Tao, Y. Hong, B. Xu, J. Lu, K. Lu, Acta Mater 50 (2002) 2075e2084. [45] K. Lu, J. Lu, Mater. Sci. Eng. A 375e377 (2004) 38e45. [46] W. Han, M. Demkowicz, E. Fu, Y. Wang, A. Misra, Acta Mater 60 (2012) 6341e6351. [47] B.N. Singh, T. Leffers, W.V. Green, M. Victoria, J. Nucl. Mater. 122&123 (1984) 703e708. [48] W. Han, E.G. Fu, M.J. Demkowicz, Y. Wang, A. Misra, J. Mater. Res. 28 (2013) 2763e2770. [49] H. Zhang, Z. Yao, M.R. Daymond, M.A. Kirk, J. Appl. Phys. 115 (2014) 103508. [50] H. Trinkaus, B.N. Singh, J. Nucl. Mater 323 (2003) 229e242. [51] H. Trinkaus, J. Nucl. Mater 118 (1983) 39e49. [52] C.H. Zhang, J. Jang, M.C. Kim, H.D. Cho, Y.T. Yang, Y.M. Sun, J. Nucl. Mater 375 (2008) 185e191. [53] B.N. Singh, H. Trinkaus, J. Nucl. Mater 186 (1992) 153e165. [54] E.E. Gruber, J. Appl. Phys. 38 (1967) 243e250. [55] P.J. Goodhew, S.K. Tyler, Proc. Roy. Soc. Lond. A 377 (1981) 151e184. [56] G.W. Greenwood, A. Boltax, J. Nucl. Mater 5 (1962) 234e240. [57] A.J. Markworth, Met. Trans. 4 (1973) 2651e2656. [58] R. Blackburn, Metall. Rev. 11 (1966) 159e176.