Materials Transactions, Vol. 49, No. 10 (2008) pp. 2259 to 2264 #2008 The Japan Institute of Metals
Precipitation of Solid Transmutation Elements in Irradiated Tungsten Alloys
Takashi Tanno1;*1, Akira Hasegawa1, Mitsuhiro Fujiwara1, Jian-Chao He1;*1, Shuhei Nogami1, Manabu Satou1, Toetsu Shishido2 and Katsunori Abe1;*2
1Department of Quantum Science and Energy Engineering, Graduate School of Engineering, Tohoku University, Sendai 980-8579, Japan 2Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan
Tungsten-based model alloys were fabricated to simulate compositional changes by neutron irradiation, performed in the JOYO fast test reactor. The irradiation damage range was 0.17–1.54 dpa and irradiation temperatures were 400, 500 and 750 C. After irradiation, microstructural observations and electrical resistivity measurements were carried out. A number of precipitates were observed after 1.54 dpa irradiation. Rhenium and osmium were precipitated by irradiation, which suppressed the formation of dislocation loops and voids. Structures induced by irradiation were not observed so much after 0.17 dpa irradiation. Electrical resistivity measurements showed that the effects of osmium on the electrical resistivity, related to impurity solution content, were larger than that of rhenium. Measurements of electrical resistivity of ternary alloys showed that the precipitation behavior was similar to that in binary alloys. [doi:10.2320/matertrans.MAW200821]
(Received April 23, 2008; Accepted August 8, 2008; Published September 18, 2008) Keywords: tungsten, solid transmutation element, precipitation, neutron irradiation, composition change, electrical resistivity
1. Introduction elements from tungsten, there are very few reports concern- ing irradiation effects on W-xRe and W-xRe-yOs alloys. Tungsten (W) is one of the candidate materials for fusion For materials used in fusion reactors, mechanical proper- reactors because of its high melting point, high sputtering ties such as hardness and ductility are important when large resistivity and high temperature strength. During fusion electromagnetic force and thermal stresses are induced in the reactor operation, it is anticipated that due to 14 MeV neutron materials by instability of plasma. We have previously irradiation, in addition to irradiation damage, solid trans- reported differences in the irradiation hardening of W-xRe mutation elements, such as rhenium (Re) and osmium (Os), and W-yOs alloys, formed primarily by irradiation precip- will be induced in pure tungsten. Consequently, with itation.9) Thermal conductivity is also important for cooling increasing neutron fluence, the solid transmutation elements which is related to electrical conductivity. The electrical will accumulate and the original pure tungsten will change resistivity, which is the inverse of the electrical conductivity, to W-xRe or W-xRe-yOs alloys. For example, calculations depends on the solute content. predict that pure W will change to a W-18Re-3Os alloy after The purpose of this study was to investigate precipitation 50 dpa irradiation, corresponding to a neutron fluence of behavior of rhenium and osmium in neutron-irradiated 10 MWy/m2.1) tungsten alloys systematically as a function of the alloy It is well known that tungsten shows brittle behavior at composition and irradiation dose in order to understand room temperature though it has excellent high temperature irradiation-induced property changes of tungsten under properties, and the low temperature embrittlement can be fusion reactor conditions. improved by Re addition.2,3) It was reported that the creep strength of tungsten irradiated with thermal neutrons is 2. Experimental Procedure higher than that of unirradiated tungsten, and substitutional solid solution rhenium transmuted from tungsten seems to be In order to simulate composition changes due to trans- the reason why the creep strength increased.4) Rhenium and mutation, tungsten based model alloys containing rhenium osmium also change the physical properties of tungsten. For and/or osmium were fabricated. The nominal compositions example, tungsten has high thermal conductivity, however of the examined alloys are shown in Table 1. They were this decreases to half the value for pure tungsten by the selected basing compositional change predicted by calcula- addition of 5 mass% Re.5,6) In terms of microstructural tion of transmutation1) in the solid solution area of W-xRe- evolution, it has been reported that phase precipitates yOs alloys. Model alloy fabrications were carried out using (Re3W) were formed in W-xRe alloys after 0.5–0.7 dpa an argon arc furnace. The raw materials were pure W irradiation at 600–1500 C without void formation, and the (99.96%) and W-26Re (W: 74:0 0:2%, Re: 26:0 0:2%) bulk properties were affected by this precipitation.7) Heavy rods supplied by Plansee Ltd. and Os (99.9%) powder neutron irradiation induced large irradiation hardening in W- 26Re alloy by irradiation-induced precipitation.8) However, even though rhenium and osmium are the main transmutation Table 1 Nominal compositions of the fabricated alloys (mass%).
W bal. bal. bal. bal. bal. bal. bal. bal. bal. bal. bal. *1 Graduate Student, Tohoku University Re — 5 10 26 — — 5 10 18 25 5 *2Present address: Hachinohe Institute of Technology, Hachinohe 031- Os————3533335 8501, Japan 2260 T. Tanno et al. supplied by Kojundo Chemical Laboratory Co., Ltd. Inter- 0.635 mm and load P of 0.98 N. Electrical resistivity was stitial impurity levels of the fabricated alloys were in the calculated using eq. (1), where V is the measured voltage range of 40–200 wppm for carbon (C), 20–40 wppm for drop, t is the specimen thickness, C:F: is the conversion oxygen (O) and <12 wppm for nitrogen (N). Disk-shaped factor fixed by the specimen shape (circle, rectangle, etc.) specimens with a diameter of 3 mm and a thickness of and the measurement conditions. In this studym C:F: was 0.3 mm were cut from the ingots using an electro discharge 3.5242.14) machine, mechanically ground and polished to 0.2 mm ¼ðV t=IÞ C:F: ð1Þ thickness, and finally annealed at 1400 C for 1 h in vacuum (<10 5 Pa).10) Neutron irradiation was carried out in the JOYO fast test 3. Results reactor in Japan Atomic Energy Agency (JAEA). The irradiation conditions used are shown in Table 2. Displace- 3.1 Microstructural observations ment damage (dpa) was calculated using the NPRIM-1.3 Figure 1 shows the microstructures of irradiated W-5Re, code11) with a displacement threshold energy of 90 eV.12) W-3Os and W-5Re-3Os alloys. Voids and dislocation loops, Greenwood et al. calculated the compositional change due seen as black dots, were observed in W-5Re after 0.17 dpa to transmutation during irradiation in the fast reactor.13) irradiation at 400 C. Plate or needle-like precipitates on the According to their results, transmutation of tungsten is {110} plane were observed instead of dislocation loops for a negligibly small even after 1.54 dpa irradiation in JOYO. higher dose and irradiation temperature (1.54 dpa irradiation After the irradiation, microstructural observations were at 750 C). Though the mean diameter of the voids grew to carried out with a transmission electron microscope (TEM) twice as large, the number density of voids decreased to one- operating at 200 kV. Electrical resistivity measurements were tenth that of the lower dose and temperature. Precipitates carried out using the four-probe method at 20 C with a load were also observed in other W-xRe alloys after 0.40 dpa and current I of 100 mA, distance between the probes S of above irradiation at 750 C. These precipitates were identified 8) as the phase (Re3W) from the electron diffraction pattern. A special structure due to irradiation was not observed Table 2 Neutron irradiation conditions. in W-3Os after 0.17 dpa irradiation at 400 C. Needle-like precipitates in the {110} plane were observed after 1.54 dpa Irradiation Fluence temperature (En > 0:1 MeV) dpa irradiation at 750 C. A few black dots were observed but ( C) (1025 n/m2) voids were not seen. The widths of the precipitates were 400 1.3 0.17 smaller than those in W-5Re alloys. The precipitates could 500 2.9 0.37 be not identified from the diffraction patterns, but they are 740 3.1 0.40 likely the phase (OsxW1 x, x ¼ 0:20{0:35) because this is 750 12 1.54 the only intermetallic compound, according to the W-yOs phase diagram.15)
W-5Re W-3Os W-5Re-3Os C ° 0.17dpa/400 C ° 1.54dpa/750
Fig. 1 TEM bright field images of W-5Re, W-3Os and W-5Re-3Os alloys after 0.17 dpa irradiation at 400 C and 1.54 dpa at 750 C. The electron incidence directions were Z = [111], only the direction of W-5Re-3Os at 750 C was Z = [001]. There were few visible structures in alloys irradiated at lower dose conditions. Plate or needle-like precipitates in the {110} plane were observed. Precipitation of Solid Transmutation Elements in Irradiated Tungsten Alloys 2261
Table 3 Summary of microstructural observations of alloys after 1.54 dpa irradiation at 750 C.
W W-5Re W-10Re W-3Os W-5Re-3Os Void Mean diameter [nm] 4.7 3.3 1.6 — — Density [1022/m3] 12 0.65 3.1 — — Swelling [%] 0.72 0.01 0.01 — — Precipitate Mean length [nm] — 14 9.5 7.3 6.8 Density [1022/m3] — 7.3 42 22 67 Volume [%] — 1.5 3.6 0.80 3.4
50 50 W-xRe-3Os
W-xRe cm cm 40 40 /µΩ /µΩ
ρ ρ 30 W-yOs 30
20 20 ° 1.54dpa/750°C 1.54dpa/750 C ° 0.40dpa/740°C 0.40dpa/740 C 10 0.37dpa/500°C 10 0.37dpa/500°C Electrical resistivity Electrical resistivity 0.17dpa/400°C 0.17dpa/400°C unirradiated unirradiated 0 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Re x or Os y /mass% Re x /mass%
Fig. 2 Electrical resistivity measurement results of W-xRe and W-yOs Fig. 3 Electrical resistivity measurement results of W-5Re-3Os alloys alloys before and after irradiation. before and after irradiation.
No specific microstructure formed by the irradiation was as x increases. The values of the resistivity of the irradiated observed by TEM in W-5Re-3Os after 0.17 dpa irradiation at alloys for x ¼ 5 to 10 tended to be smaller as the dose 400 C. The microstructure after 1.54 dpa irradiation at 750 C increased though changes in resistivity due to irradiation was similar to that of W-5Re irradiated under the same were slight. In contrast, the change in the resistivity of W- conditions. However, the observed plate- or needle-like 26Re by irradiation showed the opposite tendency. precipitates on {110} were finer than those in W-5Re, and In the case of the W-yOs alloys, the resistivity increased voids were not observed in the W-5Re-3Os alloys. It was linearly as the osmium content y increased, and this linear also confirmed that same precipitates were formed in W- relationship was maintained after irradiation. The reduction 25Re-3Os alloys irradiated to above 0.40 dpa at 750 C. in resistivity by irradiation was much larger than those of The microstructural features of W, W-5Re, W-10Re, W- W-xRe alloys. In the case of 1.54 dpa irradiation at 750 C, 3Os and W-5Re-3Os alloys irradiated at 1.54 dpa at 750 C the reduction can be explained using eq. (2), wherein are shown in Table 3. Rhenium and osmium addition over a is the reduction in resistivity by 1.54 dpa irradiation, few mass% suppressed void formation and swelling. The unirra. is the resistivity of unirradiated alloy, 1:54 dpa is the number density of precipitates increased with increasing resistivity of alloy after 1.54 dpa irradiation. content of the transmutation elements. ¼ unirr.ðyÞ 1:54 dpaðyÞ 3.2 Electrical resistivity 2 ¼ ½ unirr.ðyÞ unirr.ðy ¼ 0Þ ð2Þ Figure 2 shows the measured electrical resistivity of W- 3 xRe and W-yOs binary alloys before and after irradiation. The reduction ratio decreased as the irradiation dose The resistivity changes in pure tungsten due to the irradiation decreased though the reduction was relative to the Os were less than 20%, and were smaller than the changes content y. caused by solution effects due to rhenium or osmium. Figure 3 shows the measured electrical resistivity of W- In the case of the W-xRe alloys, the variation in the xRe-3Os ternary alloys before and after irradiation. The resistivity is not a simple linear relationship as a function of resistivity increase behavior of the unirradiated alloys was the rhenium content x, rather the resistivity tends to saturate similar to that of the W-xRe binary alloys. The resistivity 2262 T. Tanno et al. tended to decrease as the irradiation dose increased however nucleation and growth of voids might be suppressed as seen the reduction became smaller as the rhenium content for the alloys after 1.54 dpa irradiation at 750 C. Osmium increased. The electrical resistivity of W-5Re-yOs alloys could precipitate with a lower dose because its solubility showed similar behavior to that of W-yOs alloys. is lower than that of rhenium. As a result, precipitates might be finer with a smaller mean space in irradiated alloys 4. Discussion containing osmium.
4.1 Microstructure 4.2 Solution content In the case of 0.17 dpa irradiation at 400 C, dislocation Electrical resistivity can be explained by Matthiessen’s loops were observed only in the W-5Re alloys. Thus, rule with parameters of temperature, impurity concentration rhenium and osmium might interrupt interstitial cluster and deformation rate. In this study, electrical resistivity formation. Void swelling calculated by the size and density measurements were carried out at approximately the same of voids in the W-xRe alloys after 1.54 dpa irradiation at temperature and no deformation was induced. Therefore, any 750 C decreased drastically by the addition of rhenium. This resistivity changes were only due to the impurity levels. The reduction suggests that rhenium interrupts vacancy cluster resistivity change due to impurities is related to atomic formation. The interruption effects of osmium are much content of the impurities. The mass content is nearly equal to stronger than those of rhenium because no voids were the atomic masses in the model alloys, because tungsten, observed, using TEM, in alloys containing osmium. rhenium and osmium are close to each other in the periodic Rhenium and osmium are under-sized atoms compared to table and have approximately the same mass. Thus, the tungsten. Under-sized solutes in alloys cause irradiation electrical resistivity (m cm) can be explained using eq. (3), effects that do not occur in the corresponding pure metal. The wherein 0 is the electrical resistivity of pure tungsten, first such effect is segregation toward a point defect sink, Ci is the solution content (mass%) and Ai is a parameter such as a grain boundary, a dislocation etc. Over-sized which depends on the concentration of impurity elements solutes migrate by exchange with vacancies. The direction of in tungsten. migration for the over-sized solutes is thus opposite to that of X vacancy fluxes when the direction is toward a sink. On the ¼ 0 þ AiðCi=100Þð1 Ci=100Þð3Þ other hand, under-sized solutes readily bond with interstitial i atoms and form dumbbells. The interstitial atoms are thus 0 at 20 C was 5.9 m cm, which was obtained by measuring dragged toward the point defect sinks. As a result, under- pure tungsten before irradiation. Ais of the binary and ternary sized solutes are transported toward the sink with interstitial alloys were obtained, as shown in Table 4, by measurements atoms and cause segregation at the sink. Precipitation by and curve fitting for the results of this work. The values the interstitial mechanism has been observed at the surface suggest that the effects of rhenium and osmium on electrical of Mo-Re alloys after irradiation at high temperatures.16) resistivity are independent of each other, and the effect of Second, as a result of trapping of interstitial defects by osmium is five times as large as that of rhenium. under-sized solutes, the interstitial defects cannot combine Equation (3) can also estimate the solution content Ci with the vacancies and the survival rate of vacancies from the results of resistivity measurements. The predicted increases. The surviving vacancies can enhance nucleation solution contents in irradiated binary alloys are shown in and growth of voids. It has been reported that iron, which is Table 5. In the case of W-5, 10Re alloys, CRe tended to an under-sized atom compared to vanadium, enhanced void decrease as the dose increased. CRes of alloys irradiated at swelling of irradiated vanadium-added iron.17) Precipitation localized on the grain boundary was not observed in this work. Therefore void swelling decreased due Table 4 Impurity parameters for electrical resistivity. to the addition of rhenium and osmium. This suggests that the in Binary in Ternary sinks, which are recombination sites of interstitial atoms and vacancies, were in the grains. It has been suggested that ARe 138 108 vacancy clusters or micro-voids could be the nuclei for AOs 602 583 precipitation for under-sized atoms, and the interfaces between the precipitates and the matrix could be the sinks of interstitial atoms and vacancies.18) Although this idea was Table 5 Solution content changes estimated by the resistivity. not verified experimentally, it is consistent with results of this Nominal Estimated solution content Ci (mass%) work. Microstructural observations after 0.17 dpa irradiation content 0.17 dpa 0.37 dpa 0.40 dpa 1.54 dpa at 400 C suggested that rhenium and osmium enhanced unirrad. (mass%) 400 C 500 C 740 C 750 C migration of interstitial atoms toward micro-vacancy clus- ters, and the nucleation of voids was interrupted. As a result, W-xRe rhenium and osmium transported with the interstitial atoms 5 5.3 6.4 6.1 6.4 4.7 segregate and precipitate at the site of the micro-vacancy 10 11.6 14.4 12.0 — 10.6 cluster. The interfaces between the matrix and precipitates of 26 24.7 24.9 26.6 29.0 36.7 rhenium and osmium would be a stronger sink for vacancies W-yOs and interstitial atoms. Thus, precipitation of under-sized 3 4.3 2.1 1.6 1.4 1.3 rhenium and osmium would be enhanced; in addition, 5 6.0 3.4 2.4 2.4 1.7 Precipitation of Solid Transmutation Elements in Irradiated Tungsten Alloys 2263 a lower dose were higher than the nominal contents. The 50 increment by irradiation seemed to be a result of the W-xRe-3Os resistivity increment by irradiation-induced defects. How- ever, there is not matter for relative comparison because the 40 increment of resistivity was independent of dose in this cm work. Thus, the amount of precipitation of rhenium was up ρ /µΩ to 25% of nominal content at the highest irradiation dose 30 condition. This value agrees with the calculated one from the results of microstructural observations, however, the CRe of W-26Re increased as the dose increased. If the volume 20 1.54dpa/750°C of precipitates was sufficiently high, the resistivity of the ° precipitates affected the bulk resistivity related to the 0.17dpa/400 C volume fraction. However, precipitation occurred in other unirradiated 10 1.54dpa(calc.) alloys, therefore the opposite tendency could be not Electrical resistivity 0.17dpa(calc.) explained only by precipitates. Osmium transmuted from unirrad. (calc.) rhenium might be another source. In this study, the 0 predicted production of rhenium from pure tungsten was 0.7% at most, however, osmium production from W-26Re 0 5 10 15 20 25 30 reached about 1% because the cross-section of transmuta- Re x /mass% tion from rhenium to osmium might be larger than that from tungsten to rhenium according to the calculation results.13) Fig. 4 Electrical resistivity measurement results and calculated results of W-xRe-3Os alloys before and after irradiation. The effect on resistivity of osmium is very large as described above, those can explain the irregular CRe increment in W-26Re with dose increasing. 5. Conclusions The decrease of COs clearly depended on the irradiation dose. The decrease reflects the solute content changes In order to investigate irradiation-induced precipitation because the resistivity change due to the solution osmium behavior of solid transmutation elements of tungsten, the is much larger than that due to other factors. The calculated fabrication of tungsten-based model alloys was carried out, results showed that the solute osmium content decreased to and the alloys were irradiated with fast neutrons. The half of that of the unirradiated one after irradiation of just following were obtained from microstructural observations 0.17 dpa at 400 C. Only 30% of osmium existed as solute and electrical resistivity measurements. atoms in the alloys irradiated to 1.54 dpa at 750 C. On the (1) Rhenium and osmium precipitated in W-5Re and W- other hand, precipitation of osmium estimated by micro- 3Os alloys after 1.54 dpa irradiation at 750 C, though structural observations of W-3Os was less than 10% of the the contents were lower than the solubility limit. Void nominal content. In addition, no precipitates were observed formation was suppressed by the presence of rhenium in alloys after 0.17 dpa irradiation at 400 C. Thus, the and osmium. observed precipitates were a part of the precipitated or (2) Few visible structures could be observed at lower segregated osmium, and most of them would exist as irradiation dose conditions. Only grown precipitates invisible micro-clusters. were observed at higher irradiation dose conditions. The resistivity changes of ternary alloys irradiated at the The interface between the matrix and precipitates could lowest and highest dose conditions are shown in Fig. 4. In be strong sink for interstitial atoms and vacancies. this figure, the symbols represent the measured values and (3) Electrical resistivity could be used to estimate the the dotted lines are calculated values from the results of decrease in the solute contents of the transmutation binary alloys obtained using eq. (3). Each dotted line is well elements in the alloys. The calculated decreases of correlated with the measured values. Thus, it is considered rhenium and osmium were 20% and 70% of the original that the behavior of CRe and COs due to irradiation in ternary contents, respectively, after irradiation to 1.54 dpa at alloys were similar to that in binary alloys irradiated at the 750 C. Most osmium would exist as invisible micro- same conditions. clusters after irradiation. Based on the results obtained in this work, precipitation (4) In ternary alloys the behaviors of solute content change behavior in tungsten under fusion operation conditions are by irradiation for rhenium and osmium were similar to predicted, as follows. First, voids are formed by irradiation, the behaviors for rhenium and osmium in binary alloys. and nucleation and growth of the voids are suppressed by the Rhenium and osmium would not affect each other in presence of rhenium as transmutation from tungsten to the behavior at any irradiation condition. rhenium progresses. When the rhenium content reaches a few percent, precipitation starts and the precipitates increase as Acknowledgements the rhenium content increases. Osmium transmuted from rhenium would form clusters rapidly, and finer precipitates This work was supported by the Institute for Materials would form with a high number density. As a result, most Research (IMR), Tohoku University. Post-irradiation experi- of the osmium existing as precipitates or micro-clusters could ments (PIE) were carried out at the International Research be observed by TEM. Center for Nuclear Materials Science of IMR and the 2264 T. Tanno et al.
Laboratory of Alpha-ray Emitters of IMR. The authors 8) Y. Nemoto, A. Hasegawa, M. Satou and K. Abe: J. Nucl. Mater. 283– thank Mr. K. Obara for assistance with the arc melting 287 (2000) 1144–1147. 9) T. Tanno, A. Hasegawa, J. C. He, M. Fujiwara, S. Nogami, M. Satou, processes. The authors also thank Prof. M. Narui, Prof. Y. T. Shishido and K. Abe: Mater. Trans. 48 (2007) 2399–2402. Satoh, Dr. N. Nita (present: Kobelco Research Institute, Inc.) 10) J. C. He, A. Hasegawa, M. Fujiwara, M. Satou, T. Shishido and K. Abe: and the IMR staff for PIE. Mater. Trans. 45 (2004) 2657–2660. 11) S. Shimakawa, N. Sekimura and N. Nojiri: Proc. the 2002 Sympo. on Nuclear Data (2003) pp. 283–288. REFERENCES 12) C. H. M. Broeders and A. Yu. Konobeyev: J. Nucl. Mater. 328 (2004) 197–214. 1) T. Noda, M. Fujita and M. Okada: J. Nucl. Mater. 258–263 (1998) 13) L. R. Greenwood and F. A. Garner: J. Nucl. Mater. 252 (1998) 934–939. 635–639. 2) P. Makarov and K. Povarova: J. Refractory Metal & Hard Materials 14) M. A. Logan: Bell Syst. Tech. J. 40 (1961) 865–919. 20 (2002) 277–285. 15) A. Taylor, B. J. Kagle and N. J. Doyle: J. Less-Common MET. 3 (1961) 3) H. P. Gao and R. H. Zee: J. Mater. Sci. Lett. 20 (2001) 885–887. 333–347. 4) R. C. Rau, J. Moteff and R. L. Ladd: J. Nucl. Mater. 24 (1967) 164–173. 16) R. A. Erck, C. M. Wayman and L. E. Rehn: Radiation-Induced Changes 5) M. Fujitsuka, B. Tsuchiya, I. Mutou, T. Tanabe and T. Shikama: in Microstructure 13th Int. Sympo. (Part I), ASTM STP 955, F. A. ed. J. Nucl. Mater. 283–287 (2000) 1148–1151. by Garner et al., (ASTM, Philadelphia, 1987) pp. 721–729. 6) T. Tanabe, C. Eamchotchawalit, C. Busabok, S. Taweethavorn, M. 17) H. Nakajima, S. Yoshida, Y. Kohno and H. Matsui: J. Nucl. Mater. Fujitsuka and T. Shikama: Mater. Lett. 57 (2003) 2950–2953. 191–194 (1992) 952–955. 7) R. K. Williams, F. W. Wiffen, J. Bentley and J. O. Stiegler: Metall. 18) V. K. Sikka and J. Moteff: J. Appl. Phys. 43 (1972) 4942. Trans. A 14 (1983) 655–666.