- Email: [email protected]
Behaviour of neutron irradiated beryllium during temperature excursions up to and beyond its melting temperature
Journal of Nuclear Materials 465 (2015) 293e300
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Behaviour of neutron irradiated beryllium during temperature excursions up to and beyond its melting temperature rs Zarin¸s Elina Pajuste*, Gunta Kizane, Lıga Avotin¸a, Artu Institute of Chemical Physics, University of Latvia, 4 Kronvalda blvd., Riga, LV-1010, Latvia
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 January 2015 Received in revised form 6 May 2015 Accepted 18 May 2015 Available online 3 June 2015
Beryllium pebble behaviour has been studied regarding the accidental operation conditions of tritium breeding blanket of fusion reactors. Structure evolution, oxidation and thermal properties have been compared for nonirradiated and neutron irradiated beryllium pebbles during thermal treatment in a temperature range from ambient temperature to 1600 K. For neutron irradiated pebbles tritium release process was studied. Methods of temperature programmed tritium desorption (TPD) in combination with thermogravimetry (TG) and temperature differential analysis (TDA), scanning electron microscopy (SEM) in combination with Energy Dispersive X-ray analysis (EDX) have been used. It was found that there are strong relation between tritium desorption spectra and structural evolution of neutron irradiated beryllium. The oxidation rate is also accelerated by the structure damages caused by neutrons. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Beryllium is a light metal with unique properties that make it attractive for the nuclear applications, including future fusion energy power plants. Fusion power reactors will have to breed tritium to provide fuel for the fusion process. Due to its exceptional relevance, the question on the right tritium breeding concept has frequently been discussed. The European Union proposes two concepts of helium cooled tritium breeding blanket modules for testing in the International Thermonuclear Experimental Reactor ITER currently being under construction in France. In one of the concepts e “Helium Cooled Pebble Bed (HCPB) blanket”, lithium ceramic pebbles are used as a tritium breeder and beryllium pebbles as a neutron multiplier [1]. Reference material is chosen to be 1 mm pebbles fabricated by NGK Inc. by Rotating Electrode Process [2]. Beryllium is one of the key components of tritium breeding process as it is a good neutron multiplier; therefore its performance under fusion reactor conditions is an important issue. Beryllium swelling, embrittlement and hardening as a result of the neutron irradiation are important issues regarding its mechanical performance, whereas tritium production and inventory is significant safety issue [3]. At the operational temperature of reactor a large
* Corresponding author. E-mail address: [email protected] (E. Pajuste). http://dx.doi.org/10.1016/j.jnucmat.2015.05.049 0022-3115/© 2015 Elsevier B.V. All rights reserved.
fraction of tritium remains in the bulk of pebbles. In case of accidental temperature excursions this inventory of radioactive gas can be released in a short time and thus putting under the risk employees of the fusion reactor. In case of the accidental exposure to oxygen at high temperature beryllium oxide is formed. Beryllium oxide in the form of the dust can be considered as very toxic. Beryllium pebble behaviour under accidental conditions, such as elevated temperature or/and air exposure, is of great importance since it provides information on the possible safety risks. Tritium release, microstructural changes, phase transitions and oxidation under accidental condition have been studied. 2. Literature review 2.1. Neutron induced effects in microstructure of beryllium High energy neutrons knock out the atoms in the lattice and generate defects such as vacancies and interstitials [4]. There are several possible geometric configurations for clusters of vacancies and self-interstial atoms: planar dislocation loops (faulted or perfect) and three dimensional configurations as stacking fault tetrahedrons and cavities [5]. In beryllium irradiated at low temperatures (below 473 K) dislocation loops have been observed experimentally by several authors [6,7]. Moreover, the accumulation of the dislocations in different lattice planes depends on the type of the dislocation; vacancy loops are located on the basal planes, whereas interstitials on the prismatic ones [8]. Through the
294
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
classical mechanism of vacancy accumulation e voids are formed. Accumulation of these voids on the grain boundaries leads to the embrittlement and swelling of a polycrystalline beryllium [7]. However, displacement damage is not significant factor for limiting beryllium pebble lifetime in fusion reactor since operation temperature will be comparably high. The main problem for beryllium performance under neutron irradiation is the gaseous products of neutron induced transmutations [9]. 2.2. Gaseous products of neutron induced transmutations As a result of neutron induced transmutations of beryllium helium and hydrogen isotope, tritium, are produced in considerable amounts. In the frame of the European Power Plant Conceptual Study, the peak integral gas production in beryllium, at the End-OfLife of HCPB modules (40 000 h operation), has been assessed as 25 700 appm helium and 640 appm tritium, taking account in-pile tritium decay. The global tritium production in the whole of the blanket (390 tons of beryllium) is 23.8 kg [10]. It is assumed that initially helium and tritium form a dynamic solution throughout the entire lattice. Due to low solubility, helium tends to form gas clusters that performs as a nucleation site for further formation of bubbles [11]. In fact, grain boundaries are a strong sink for such defects, thus a large fraction of these clusters could be formed nearby them. Process of the bubble formation occurs via gas precipitation and a vacancy capture and this process is strongly temperature dependent [4]. After irradiation at temperatures below 200 C no bubbles can be observed, whereas at the temperatures above this level e bubbles of few nanometres start to appear [6,12]. Interesting fact is that this tendency remains even at very long exposure time and large helium amount. In the low temperature (70 C) irradiation experiment of beryllium that lasted for 15 years (helium content reached 2.2 at%), no bubbles were observed although significant swelling of the sample was obvious. Bubbles appeared only during post irradiation annealing of the sample [13]. Correlation between irradiation temperature and bubble size is also described in literature. Post irradiation examination of beryllium pebbles showed that bubble size in pebbles irradiated at ~700 C are larger by a factor of 10 and more than those in pebbles irradiated at ~400 C (diameter length range: 40e140 nm and 5e10 nm, respectively) [14]. Bubbles inside the grains were found to have a specific crystallographic form e a shape of hexagonal prisms. It is likely related to beryllium lattice parameters [11]. Swelling of beryllium due to helium accumulation also limits the lifetime of beryllium pebbles [4,15]. Mathematical approach of swelling models has been provided by several authors [4,16,17]. The parameters used in these calculations are neutron flux, initial porosity, amount of impurities, grain size [17]. It has been observed that swelling is less intense if the grain size of beryllium is smaller [12,15,18]. Tritium is assumed not to form its own gas inclusions since its overall concentration is very low if compared with helium. It is believed to accumulate as an interstitial atom or to be trapped by the structural or chemical traps. In contrast to helium tritium has a high chemical reactivity and it can form chemical bonds with impurities, mostly with beryllium oxide forming the hydroxide. It has been found that besides oxide layer on the surface BeO inclusions have a tendency to accumulate on the grain boundaries [6], therefore it can be expected that chemically bonded tritium can be found either on grain boundaries or in the surface layer. However, it is assumed that most of the tritium resides mostly in the helium gas inclusions [19,20]. This assumption is also in good agreement with experimental data on tritium chemical forms beryllium showing that in neutron irradiated tritium is accumulated mainly as a
molecules [21]. To understand tritium behaviour in beryllium it is necessary to know such parameters as solubility, trapping energies and diffusivity. Some properties might be extrapolated from available data about protium or deuterium, however, isotopic effects must be taken into account [22]. In fact, it should be mentioned that in one of the first successful attempts to measure hydrogen solubility in beryllium tritium was used due to its simple detection based on its radioactivity [23]. Comprehensive overview of the experimentally obtained data on hydrogen solubility, diffusivity and permeation has been provided by R. A. Causey in 2002 [24]. Diffusion coefficient of neutron produced tritium in beryllium has been measured by several authors [25e27]. It is clear that overall tritium transport in neutron irradiated beryllium does not take place by a single mechanism of diffusion of tritium atoms dissolved in beryllium lattice. Trapping phenomena and influence of oxide layer have a crucial role. Theoretical model of tritium trapping on the imperfections of the lattice, such as vacancies, grain boundaries, etc. has been described [28]. Possible transport mechanisms of tritium in neutron irradiated beryllium has been studied by a number of authors, and a special computer codes or models have been developed [10,16,27,29e34]. 2.3. Thermal desorption of tritium form neutron irradiated beryllium Tritium and helium thermo-desorption spectra had been analysed widely in order to get comprehensive overview of the processes occurring during high temperature treatment of irradiated beryllium [4,10,26,32,35e40]. Earlier it was found that most of tritium releases together with helium, whereas helium can be released at temperatures close to the melting point of beryllium [41]. However, low temperature peaks of tritium release have been observed in a number of experiments. Some authors believe it is tritium located close to the surface and it is escaping through the micro cracks created by the irradiation or open porosity formed as a result of bubble coalescence along grain boundaries [37,42]. Other associates it with the diffusion of the tritium existing as an interstitial [10,35]. Tritium desorption process is affected by a number of factors, such as size and shape of the sample (diffusion length), structure (accumulation in the pores, grain boundaries, transportation along the cracks, etc), concentration of the helium gas inclusions and size of the gas bubbles (tritium accumulation together with helium), oxide content (tritium chemical bonding), etc. Therefore, desorption spectra might significantly differ. For instance, tritium desorption from pebbles of 0.1 mm mm in diameter, produced by Inert Gas Atomization method and irradiated with 2.70 1025 neutrons per m2 starts at ~360 K, whereas from pebbles of 1 mm in diameter, produced by Rotating Electrode Process and irradiated with (3e4) 1025 neutrons per m2 e at 950 K [43]. Neutron induced damage and helium bubbles also leads to tritium retention in the pebbles [31,44]. Tritium desorption from pebbles in which tritium is loaded by thermo-sorption process occurs at much lower temperatures than neutron irradiated pebbles of the same production batch [45,46]. 2.4. Beryllium oxidation At ambient temperature beryllium has a protective oxide layer of about 1 nm in thickness that prevents its further oxidation [47]. However, if the temperature is raised above 1000 K a breakaway oxidation reaction occurs and non e protective oxide layer starts to be formed [48]. This dramatic increase of oxidation rate has been explained by the fracture of oxide layer due to differences of thermal expansions of metallic beryllium and its oxide [49]. Beryllium oxidation below ~1270 K is believed to follow parabolic rate law,
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
295
Table 1 Irradiation conditions of beryllium pebbles at High Flux reactor. Irradiation time Neutron fluence (E>0.1 MeV) Irradiation temperature 4 He content
294 days 3e4 1025 m2 423-823 K 300-600 appm
Fig. 2. Tritium release, DTA signal and mass increase of neutron irradiated beryllium pebble during thermal treatment in Heþ0.1% H2 atmosphere (gradual heating 5 K/min up to 1600 K).
according to the first-principe calculations described in literature [54]. 3. Experimental Fig. 1. Experimental set-up for simultaneous Differential Thermal Analysis DTA, Thermal Gravimetric Analysis and Thermal Desorption Analysis TDA of tritium.
whereas at the temperature beyond this e linear rate law [49]. In the hot air beryllium nitride Be3N2 is also formed [50]. 2.5. Thermal properties and phase transition of beryllium Beryllium is in the hexagonal-close-packed (hcp) structure at ambient pressure and temperature. However, at temperature near its melting point beryllium transforms from hcp structure to bodycentred-cubic (bcc) structure. In the early work of A.J. Martin and A. Moor temperature of hcp-bcc or a / b transition has been experimentally demonstrated. Authors performed experiments with several beryllium grades by heating them in a vacuum at the rate of 1e5 K per minute. They observed a/ b transition in the temperature range between 1541 and 1548 C and melting in range between 1559 and 1564 K [51]. In more recent years H. Kleykamp has performed a thorough study of beryllium thermal properties and provided temperatures and enthalpies of both phase transition and melting of beryllium [52]. Comprehensive overview of thermal properties of beryllium can be found in paper of C.B. Alcock et al. [53]. Beryllium phase transition can also occur at high pressures
3.1. Materials In this study, beryllium pebbles of ~1 mm in diameter and produced by Rotating Electrode Process REP (manufacturer NGK Insulators Ltd., Handa City, Japan) were analysed. Main impurities of the samples given by manufacturer are BeO with content of 2300 ppm and Mg e 300 ppm. A portion of the pebbles were irradiated with neutrons in the Pebble Bed Assembly experiment at the High Flux Reactor in Petten, the Netherlands. Irradiation conditions are given in Table 1. Chemical composition and structure of irradiated and nonirradiated pebbles were analysed also prior to the annealing experiments. 3.2. Methods An experimental set-up has been developed to study tritium desorption, thermal properties and oxidation rate of beryllium. Method is based on simultaneous Differential Thermal Analysis DTA, Thermal Gravimetric Analysis and Thermal Desorption Analysis TDA of tritium. Schematic view of the experimental set-up is given in Fig. 1. Samples were heated in the ceramic crucible in furnace of
Table 2 DTA calibration results. Material
Mass, mg
Heating rate, K/min
Melting temp. Tmelt, K
Enthalpy of melting DmH, J/mol [55]
Peak area Q, mV$s
Sensitivity factor S, mV/mW
Au Au Au Au
151.5 151.5 151.5 151.5 Tmelt (exp.) Tmelt(teor.) 159.2 159.2 159.2 159.2 Tmelt(exp.) Tmelt(teor.)
5 5 5 5
1339 1339 1340 1339 1339 ± 1 1337.33 [56] 1234 1231 1231 1230 1232 ± 2 1234.93 [56]
12552.0 12552.0 12552.0 12552.0
6549 5446 5853 5439 SAu ¼
0.68 0.56 0.60 0.56 0,60 ± 0,05
11296.8 11296.8 11296.8 11296.8
9969 1047 1110 1251 SAg
0.54 0.63 0.67 0,75 0.66 ± 0.06
Ag Ag Ag Ag
5 5 5 5
Sensitivity factor used for calculation of beryllium phase transition and melting enthalpy e (0.63 ± 0.08) mV/mW.
296
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
Fig. 3. Structure of the neutron irradiated beryllium pebbles before treatment. Cross section of pebbles obtained by polishing off half of it by SiC sandpaper and then by diamond paste (1 and 0.05 mm particle size).
Fig. 4. Structure evolution of neutron irradiated beryllium pebbles after thermal treatment up to 923, 1123, 1223 and 1323 K (heating rate 5 K/min). Amplification factors chosen based on the scale of formed structure defects at the particular temperatures.
Fig. 5. Tritium release and structure evolution of neutron irradiated beryllium pebble during thermal treatment (heating rate 5 K/min).
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
differential thermal analyser TG/DTA (Seiko EXTAR 6300) and the gaseous tritium released from beryllium was measured by tritium monitor TEM 2100A with a proportional gas flow-through detector DDH 32. Heating rate was 2.4, 5, 20 and 40 K per minute and a mixture of He and 0.1% H2 were used as a purge gas with a purging rate of 200 mL/min. Composition of purge gas was chosen based on the planned condition in the future fusion devices. Samples were heated up to 1373, 1423 and 1600 K for tritium desorption and thermal analysis. Oxidation rate was studied by comparing the mass increase (TG signal) of the samples during analogous thermal treatment in air. For studies of structure evolution heating was stopped at lower temperatures: 823, 923, 1123, 1223, 1323 and 1523 K, respectively. Microstructure of samples was evaluated by means of scanning electron microscopy SEM (Hitachi S-4800, 15 kVacc, 15 mA). Chemical composition of the samples was assessed by the means on Energy Dispersive X-ray analysis that are conjugated with the SEM. DTA data was used for analysis of thermal properties. Calibration was performed with gold and silver reference materials. Sensitivity factor S for Au was found to be (0.60 ± 0.05) mV/mW, and for Ag e (0.66 ± 0.06) mV/mW (Table 2).
297
Fig. 7. Tritium desorption as a function of temperature depending on the heating rate.
4. Results and discussion Studies on behaviour of neutron irradiated beryllium during thermal treatment included gradual heating in atmosphere of helium with addition of hydrogen and in air. Example of the data obtained is shown on Fig. 2. First process occurring during thermal treatment is tritium release (red line) which starts at temperatures above 850 K. In temperature range between 1120 and 1370 K burst release was observed when the rest of radioactive gas is released. Mass increase of the pebble (black line) that corresponds to the oxidation process starts during or right after the burst release of tritium and becomes more intense close to the melting temperature of beryllium. Mass increase during thermal treatment in He þ 0.1% H2 atmosphere occurs due to the minor though existing presence of oxygen. DTA signal curve (green line) (in web version) clearly shows beryllium phase transition and melting process. In following chapters each process is described in more details. 4.1. Tritium release and structure evolution of pebbles during thermal treatment
Fig. 8. Beryllium pebbles before thermal treatment (on the left) and after thermal treatment up to 1620 K in air atmosphere.
inventory in pebbles, possibilities to prevent it as well as to the processes occurring in case of accidental temperature excursions when this inventory can be released in a very short period of time. SEM analysis of pebbles before thermal treatment revealed a large void in the bulk of the pebble (Fig. 3). These voids are generated during the cooling phase of the fabrication process and are developed to prevent beryllium swelling and failure when
Studies on tritium desorption process and its relation to microstructural evolution of beryllium pebbles gives a better understanding of issues related both to undesired formation of tritium
Fig. 6. Time of full tritium desorption depending of the heating rate.
Fig. 9. Mass increase of neutron irradiated and nonirradiated beryllium pebbles during thermal treatment up to 1620 K with heating rate 5 K/min. Mass increase corresponds to the formation of beryllium oxide.
298
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
Fig. 10. Combined SEM (SE) and EDX mapping images (Be, O) of irradiated pebble before thermal treatment, noniradiated and irradiated pebbles after thermal treatment up to 1600 K (heating rate 5 K/min, cooling rate e uncontrolled) in atmosphere of He þ 0.1%H2 with addition with air.
irradiated with neutrons e helium produced in the nuclear reactions is stored in this void instead of beryllium lattice [57,58]. However, this void also acts as a trap for tritium, therefore preventing its release during operation of the reactor. No other porosity was observed. According to the literature [14] helium inclusions after irradiation at this temperature do not exceed several nanometres; therefore they were not observable by the available equipment. After treatment up to temperature when tritium release has already started (920 K or ~0.6Tmelt), porosity has appeared (Fig. 4). The size of the pores reaches several hundreds of nanometres. By further raise of temperature the porosity increases (both concentration and size of the pores). At temperature above 1100 K formation of large radial cracks and connected pore channels starts (Fig. 4). These cracks and open porosity must be the pathway of the release of gases trapped in the technical void. This observation
hence explains burst release of the tritium near the particular temperature (Fig. 5). Tritium desorption experiments were performed with 4 different heating rates: 2.4, 5, 20 and 40 K/min and time required for full desorption estimated. If temperature increases with the rate 20 K/min entire tritium inventory is released in less than 30 min after temperature has reached ~1000 K (Fig. 6). It is an important issue since such a fast release of tritium inventory can be a significant threat for the operators of the reactor. In Fig. 7 tritium desorption as a function of temperature is shown. In the heating rate range 2.4e20 K/min one tritium desorption pattern remains e tritium desorption and burst release starts at approximately similar temperatures. Data from experiments with heating rate 40 K/min are not shown in the graph, since time for tritium to travel from the sample to the volume of detector were ~1e2 min (corresponding temperature increase is in range
Fig. 11. DTA signal (opposite values) of phase transition and melting of nonirradiated and irradiated beryllium samples (red marks indicate peak integration area). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
299
Table 3 Phase transition and melting enthalpies of irradiated and nonirradiated beryllium. Irradiation Mass, mg
Heating rate, K/ min
Phase transition peak area Q, mV$s
Enthalpy of phase transition DmH, J/ mol
mV$s
Enthalpy of melting DmH, J/ mol
Irradiated 0.96 Irradiated 1.01 Irradiated 1.34
5 5 5
196.6 201.3 263.0
2964 3014 2841
235.2 265.9 360.4
3547 3811 2841
Nonirr. Nonirr. Nonirr.
5 5 5
220.2 212.0 243.9
3663 3230 3239
248.1 260.0 249.0
4128 3962 3307
Melting peak area Q,
2900 ± 200 0.87 0.95 1.09
3800 ± 200
3400 ± 300
3800 ± 400
Melting enthalpy do not differ for neutron irradiated and nonirradiated beryllium, whereas phase transition enthalpy is slightly lower for neutron irradiated beryllium.
Fig. 12. Temperatures (a) and enthalpies (b) of phase transition and melting: comparison with the literature data (DTA method only) [51,52,60,61].
from 40 to 80 K) and therefore peak position cannot be estimated correctly. In more detail kinetics of tritium release behaviour during thermal annealing of beryllium pebbles at different temperature increase rates are described in an earlier publication [59]. 4.2. Oxidation of pebbles For oxidation studies beryllium pebbles were heated (5 K/min) up to 1600 K in air atmosphere and the mass increase corresponding to the growth of oxide layer was measured. At the stage when beryllium pebble is fully oxidized expected mass increase is ~180%. As a result of oxidation a brittle, flaking beryllium oxide pebble is formed (Fig. 8). For both irradiated and nonirradiated pebbles oxidation starts at temperature region between 1250 and 1350 K (Fig. 9). However, oxidation rate is higher for irradiated samples and most of the pebble volume is oxidized before beryllium melting whereas considerable oxidation of nonirradiated pebble appears only beyond melting temperature. Higher oxidation intensity of neutron irradiated pebbles can be explained by the structural changes that lead to increase of free surface area. It is demonstrated in Fig. 10 where SEM images of beryllium pebbles after thermal treatment in atmosphere of Heþ0.1%H2 with addition of air (<5%) combined with EDX elemental mapping data are shown. Oxidized areas can be identified by the increased oxygen presence (in web version) (blue colour in EDX mapping image).
Oxidation at operational temperature (~920 K for Be pebbles [1]) is not a critical issue e intense oxidation occurs at the temperature above 1200 K. However, it must be emphasized that in case of accidental temperature increase in combination of air exposure formation of toxic dust of beryllium oxide must be considered as an additional hazard besides tritium. 4.3. Phase transition and melting point In this study, effects of neutron induced lattice damage on the phase transition and melting points were attempted to find. DTA curves reveal two clearly visible peaks that correspond to phase transition and melting processes of the beryllium. In Fig. 11 the opposite values of DTA signal (with subtracted background) is shown for irradiated and nonirradiated pebbles. However, since the mass of samples was very small and the atmosphere of experiment was not fully inert, results of differential thermal analysis DTA give comparable instead of absolute value of enthalpy. Moreover, there is additional peak between phase transition and melting peaks that makes the integration of studied peaks complicated. This peak can be related to beryllium oxide or beryllium nitride since these compounds are already formed at this temperature as a result of beryllium interaction with air (Table 3). It was observed that both aeb transition and melting occurs at similar temperatures for neutron irradiated and nonirradiated beryllium. For nonirradiated beryllium in the experimental conditions transition occur at (1546 ± 2)K and melting (1562 ± 1)K,
300
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
whereas for irradiated (1546 ± 1)K and melting (1561 ± 1)K, respectively. Obtained values were compared to the data provided by other authors (Fig. 12). The range of enthalpy values given in literature is considerably large: 2000e9100 J/mol. Values obtained in this study lay within this range. The discrepancy of the data obtained by different authors can be explained by the conditions of experiment (equipment, quality and grade of beryllium, atmosphere, and presence beryllium oxide) (Fig. 11). 5. Conclusions Tritium desorption from neutron irradiated beryllium pebbles starts at ~1000 K and is completely released at temperature ~1350 K, thus before beryllium melting. All of the tritium inventory can be released in less than 30 min in case of accidental temperature excursion if the temperature increase rate exceeds 20 K/min. High temperature oxidation process of neutron irradiated pebbles is more intense than that for nonirradiated pebbles. However, it is due to helium induced structure destruction that leads to an increased free surface Thermal properties of beryllium are not considerably affected by neutron irradiation with fluence (3e4) 1025 m2 (E > 0.1 MeV). Acknowledgements This work was supported by the European Communities within the framework of EFDA. The views and opinions expressed herein do not necessarily reflect those of the European Commission. References [1] L.M. Giancarli, M. Abdou, D.J. Campbell, V.A. Chuyanov, M.Y. Ahn, M. Enoeda, C. Pan, Y. Poitevin, E. Rajendra Kumar, I. Ricapito, Y. Strebkov, S. Suzuki, P.C. Wong, M. Zmitko, Fusion Eng. Des. 87 (5e6) (2012) 395e402. € slang, [2] M. Zmitko, Y. Poitevin, L. Boccaccini, J.F. Salavy, R. Knitter, A. Mo €sser, J. Nucl. Mater. 417 (1e3) (2011) A.J. Magielsen, J.B.J. Hegeman, R. La 678e683. [3] D. Carloni, L.V. Boccaccini, F. Franza, S. Kecskes, Fusion Eng. Des. 89 (7e8) (2014) 1341e1345. [4] F. Scaffidi-Argentina, M. Dalle Donne, C. Ferrero, C. Ronchi, Fusion Eng. Des. 27 (1995) 275e282. [5] S.J. Zinkle, 1.03-radiation-induced effects on microstructure, in: R.J.M. Konings (Ed.), Comprehensive Nuclear Materials, Elsevier, Oxford, 2012, pp. 65e98. [6] V.P. Chakin, A.O. Posevin, A.V. Obukhov, P.P. Silantyev, J. Nucl. Mater. 386e388 (2009) 206e209. [7] L.L. Snead, J. Nucl. Mater. 326 (2e3) (2004) 114e124. [8] V.P. Chakin, V.A. Kazakov, R.R. Melder, Y.D. Goncharenko, I.B. Kupriyanov, J. Nucl. Mater. 307e311 (Part 1) (2002) 647e652. [9] B.S. Hickman, G.T. Stevens, Effect of Neutron Irradiation on Beryllium Metal, Australian Nuclear Science and Technology Organisation, Sydney, 1963. [10] E. Rabaglino, C. Ronchi, A. Cardella, Fusion Eng. Des. 69 (2003) 455e461. [11] W. Van Renterghem, A. Leenaers, S. Van denBerghe, J. Nucl. Mater. 374 (1e2) (2008) 54e60. [12] I.B. Kupriyanov, R.R. Melder, V.A. Gorokhov, Fusion Eng. Des. 51e52 (2000) 135e143. [13] A. Leenaers, G. Verpoucke, A. Pellettieri, L. Sannen, S. Van den Berghe, J. Nucl. Mater. 372 (2e3) (2008) 256e262. [14] M. Klimenkov, V. Chakin, A. Moeslang, R. Rolli, J. Nucl. Mater. 443 (1e3) (2013) 409e416. [15] E. Ishitsuka, H. Kawamura, T. Terai, S. Tanaka, J. Nucl. Mater. 258e263 (1998) 566e570. [16] M. Dalle Donne, F. Scaffidi-Argentina, C. Ferrero, C. Ronchi, J. Nucl. Mater. 212e215 (1994) 954e960. [17] G.A. Sernyaev, A.S. Pokrovskiy, R.M. Bagautdinov, S.A. Fabritsiev, I.V. Mazul,
V.R. Barabash, J. Nucl. Mater. 233e237 (1996) 891e897. [18] I.B. Kupriyanov, G.N. Nikolaev, S.K. Zavjalov, L.A. Kurbatova, N.E. Zabirova, M. Zmitko, Fusion Eng. Des. 87 (7e8) (2012) 1338e1341. [19] J.M. Beeston, G.R. Longhurst, R.S. Wallace, J. Nucl. Mater. 195 (1e2) (1992) 102e108. [20] E. Hayward, C. Deo, J. Phys. Condens. Matter 24 (26) (2012) 265402. [21] E. Pajuste, A. Vitins, G. Kizane, V. Zubkovs, P. Birjukovs, Fusion Eng. Des. 86 (9e11) (2011) 2125e2128. [22] R. Kaur, S. Prakash, J. Phys. F: Metal Phys. 12 (7) (1982) 1383. [23] P.M.S. Jones, R. Gibson, J. Nucl. Mater. 21 (3) (1967) 353e354. [24] R.A. Causey, J. Nucl. Mater. 300 (2e3) (2002) 91e117. [25] I.L. Tazhibaeva, V.P. Shestakov, E.V. Chikhray, in: 18th Symposium on Fusion Technology, Germany, Karlsruhe, 22e26 August; Germany, Karlsruhe, 1994, p. 427. [26] D.L. Baldwin, M.C. Billone, J. Nucl. Mater. 212e215 (1994) 948e953. [27] S. Cho, A.R. Raffray, M.A. Abdou, J. Nucl. Mater. 212e215 (1994) 961e965. [28] M.G. Ganchenkova, V.A. Borodin, R.M. Nieminen, Phys. Rev. B 79 (13) (2009) 134101. [29] Y. Verzilov, S. Sato, K. Ochiai, M. Wada, A. Klix, T. Nishitani, Fusion Eng. Des. 82 (1) (2007) 1e9. [30] E. Ishitsuka, H. Kawamura, T. Terai, S. Tanaka, J. Nucl. Mater. 283e287 (2) (2000) 1401e1404. [31] I.B. Kupriyanov, V.A. Gorokhov, V.V. Vlasov, A.M. Kovalev, V.P. Chakin, J. Nucl. Mater. 329e333 (2004) 809e813. [32] F. Scaffidi-Argentina, M. Dalle Donne, H. Werle, Fusion Eng. Des. 58e59 (2001) 641e645. [33] S. Cho, M.A. Abdou, Fusion Eng. Des. 28 (1995) 265e270. €slang, R.A. Pieritz, E. Boller, C. Ferrero, J. Nucl. Mater. 386e388 (2009) [34] A. Mo 1052e1055. [35] E. Rabaglino, J.P. Hiernaut, C. Ronchi, F. Scaffidi-Argentina, J. Nucl. Mater. 307e311 (Part 2) (2002) 1424e1429. [36] F. Scaffidi-Argentina, M. Dalle Donne, H. Werle, J. Nucl. Mater. 258e263 (1998) 595e600. [37] A.K. Klepikov, I.L. Tazhibaeva, V.P. Shestakov, O.G. Romanenko, Y.V. Chikhray, E.A. Kenzhin, Y.S. Cherepnin, L.N. Tikhomirov, J. Nucl. Mater. 233e237 (2) (1996) 837e840. [38] I.B. Kupriyanov, G.N. Nikolaev, V.V. Vlasov, A.M. Kovalev, V.P. Chakin, J. Nucl. Mater. (2007) 367e370. Part A (0), 511-515. [39] C.E. Johnson, D.L. Baldwin, J.P. Kopasz, J. Nucl. Mater. 212e215 (2) (1994) 966e970. [40] A.V. Fedorov, S. van Til, L.J. Magielsen, M.P. Stijkel, J. Nucl. Mater. 442 (1e3, Supplement 1) (2013) S472eS477. [41] E. Rabaglino, Helium and tritium in neutron-irradiated beryllium, Forschungszentrum Karlsr. Karslruhe (2004) 127. [42] S. van Til, A.V. Fedorov, M.P. Stijkel, H.L. Cobussen, R.K. Mutnuru, P. v.d. Idsert, M. Zmitko, J. Nucl. Mater. 442 (1e3, Supplement 1) (2013) S478eS482. [43] E. Pajuste, Tritium Behaviour in Fusion Reactor Materials, University of Latvia, Riga, 2012. [44] S.L. Kanashenko, A.E. Gorodetsky, A.P. Zakharov, W.R. Wampler, Phys. Scr. 1996 (T64) (1996) 36. [45] V. Chakin, A. Moeslang, P. Kurinskiy, R. Rolli, H.C. Schneider, E. Alves, L.C. Alves, Fusion Eng. Des. 86 (9e11) (2011) 2338e2342. [46] A. Vitins, V. Zubkovs, G. Kizane, E. Pajuste, V. Kinerte, Fusion Sci. Technol. 60 (3) (2011) 1143e1146. € ri, In oxidation of beryllium e a scanning auger [47] C. Tomastik, W. Werner, H. Sto investigation, in: 20th IAEA Fusion Energy Conference, Vilamoura, Portugal, 1e6 November, 2004; Vilamoura, Portugal, 2004, pp. FT/P1eP29. [48] D.W. Aylmore, S.J. Gregg, W.B. Jepson, J. Nucl. Mater. 2 (2) (1960) 169e175. [49] R.O. Adams, J.T. Hurd, J. Less Common Metals 18 (4) (1969) 399e409. [50] N.N. Greenwood, A. Earnshaw, Chemistry of the Elements, second ed., Butterworth-Heinemann, Oxford, 1997, p. 340. [51] A.J. Martin, A. Moore, J. less-common metals 1 (1959) 85e93. [52] H. Kleykamp, Thermochim. Acta 345 (2000) 179e184. [53] C.B. Alcock, M.W. Chase, V.P. Itkin, J. Phys. Chem. Reference Data 22 (1) (1993) 1e85. [54] Z.-C. Guo, F. Luo, Y. Cheng, Comput. Mater. Sci. 84 (0) (2014) 139e144. [55] A.T. Dinsdale, CALPHAD 15 (4) (1991) 317e425. [56] H. Preston-Thomas, Metrologia 27 (1) (1990) 3. [57] Ishitsuka, E.; Kawamura, H.; Sakamoto, N.; Nishida, K. Process for Preparing Metallic Beryllium Pebbles. 5958105, sept.28, 1999, 1999. [58] Nishida, K.; Sakamoto, N. Metallic Beryllium Pebble. 08/16/1994 1994. ne, A. Matıss, E. Pajuste, V. Zubkovs, J. Nucl. Mater. 442 (1e3, [59] A. Vıtin¸s, G. K¸iza Supplement 1) (2013) S490eS493. [60] R.G. Loasby, D. Dearden, J. Less Common Metals 52 (1) (1977) 137e144. [61] A. ABEY, Pressure Phase Lines and Enthalpies for the Ab and B Liquid Transitions in Beryllium, 1984.
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Behaviour of neutron irradiated beryllium during temperature excursions up to and beyond its melting temperature rs Zarin¸s Elina Pajuste*, Gunta Kizane, Lıga Avotin¸a, Artu Institute of Chemical Physics, University of Latvia, 4 Kronvalda blvd., Riga, LV-1010, Latvia
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 January 2015 Received in revised form 6 May 2015 Accepted 18 May 2015 Available online 3 June 2015
Beryllium pebble behaviour has been studied regarding the accidental operation conditions of tritium breeding blanket of fusion reactors. Structure evolution, oxidation and thermal properties have been compared for nonirradiated and neutron irradiated beryllium pebbles during thermal treatment in a temperature range from ambient temperature to 1600 K. For neutron irradiated pebbles tritium release process was studied. Methods of temperature programmed tritium desorption (TPD) in combination with thermogravimetry (TG) and temperature differential analysis (TDA), scanning electron microscopy (SEM) in combination with Energy Dispersive X-ray analysis (EDX) have been used. It was found that there are strong relation between tritium desorption spectra and structural evolution of neutron irradiated beryllium. The oxidation rate is also accelerated by the structure damages caused by neutrons. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Beryllium is a light metal with unique properties that make it attractive for the nuclear applications, including future fusion energy power plants. Fusion power reactors will have to breed tritium to provide fuel for the fusion process. Due to its exceptional relevance, the question on the right tritium breeding concept has frequently been discussed. The European Union proposes two concepts of helium cooled tritium breeding blanket modules for testing in the International Thermonuclear Experimental Reactor ITER currently being under construction in France. In one of the concepts e “Helium Cooled Pebble Bed (HCPB) blanket”, lithium ceramic pebbles are used as a tritium breeder and beryllium pebbles as a neutron multiplier [1]. Reference material is chosen to be 1 mm pebbles fabricated by NGK Inc. by Rotating Electrode Process [2]. Beryllium is one of the key components of tritium breeding process as it is a good neutron multiplier; therefore its performance under fusion reactor conditions is an important issue. Beryllium swelling, embrittlement and hardening as a result of the neutron irradiation are important issues regarding its mechanical performance, whereas tritium production and inventory is significant safety issue [3]. At the operational temperature of reactor a large
* Corresponding author. E-mail address: [email protected] (E. Pajuste). http://dx.doi.org/10.1016/j.jnucmat.2015.05.049 0022-3115/© 2015 Elsevier B.V. All rights reserved.
fraction of tritium remains in the bulk of pebbles. In case of accidental temperature excursions this inventory of radioactive gas can be released in a short time and thus putting under the risk employees of the fusion reactor. In case of the accidental exposure to oxygen at high temperature beryllium oxide is formed. Beryllium oxide in the form of the dust can be considered as very toxic. Beryllium pebble behaviour under accidental conditions, such as elevated temperature or/and air exposure, is of great importance since it provides information on the possible safety risks. Tritium release, microstructural changes, phase transitions and oxidation under accidental condition have been studied. 2. Literature review 2.1. Neutron induced effects in microstructure of beryllium High energy neutrons knock out the atoms in the lattice and generate defects such as vacancies and interstitials [4]. There are several possible geometric configurations for clusters of vacancies and self-interstial atoms: planar dislocation loops (faulted or perfect) and three dimensional configurations as stacking fault tetrahedrons and cavities [5]. In beryllium irradiated at low temperatures (below 473 K) dislocation loops have been observed experimentally by several authors [6,7]. Moreover, the accumulation of the dislocations in different lattice planes depends on the type of the dislocation; vacancy loops are located on the basal planes, whereas interstitials on the prismatic ones [8]. Through the
294
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
classical mechanism of vacancy accumulation e voids are formed. Accumulation of these voids on the grain boundaries leads to the embrittlement and swelling of a polycrystalline beryllium [7]. However, displacement damage is not significant factor for limiting beryllium pebble lifetime in fusion reactor since operation temperature will be comparably high. The main problem for beryllium performance under neutron irradiation is the gaseous products of neutron induced transmutations [9]. 2.2. Gaseous products of neutron induced transmutations As a result of neutron induced transmutations of beryllium helium and hydrogen isotope, tritium, are produced in considerable amounts. In the frame of the European Power Plant Conceptual Study, the peak integral gas production in beryllium, at the End-OfLife of HCPB modules (40 000 h operation), has been assessed as 25 700 appm helium and 640 appm tritium, taking account in-pile tritium decay. The global tritium production in the whole of the blanket (390 tons of beryllium) is 23.8 kg [10]. It is assumed that initially helium and tritium form a dynamic solution throughout the entire lattice. Due to low solubility, helium tends to form gas clusters that performs as a nucleation site for further formation of bubbles [11]. In fact, grain boundaries are a strong sink for such defects, thus a large fraction of these clusters could be formed nearby them. Process of the bubble formation occurs via gas precipitation and a vacancy capture and this process is strongly temperature dependent [4]. After irradiation at temperatures below 200 C no bubbles can be observed, whereas at the temperatures above this level e bubbles of few nanometres start to appear [6,12]. Interesting fact is that this tendency remains even at very long exposure time and large helium amount. In the low temperature (70 C) irradiation experiment of beryllium that lasted for 15 years (helium content reached 2.2 at%), no bubbles were observed although significant swelling of the sample was obvious. Bubbles appeared only during post irradiation annealing of the sample [13]. Correlation between irradiation temperature and bubble size is also described in literature. Post irradiation examination of beryllium pebbles showed that bubble size in pebbles irradiated at ~700 C are larger by a factor of 10 and more than those in pebbles irradiated at ~400 C (diameter length range: 40e140 nm and 5e10 nm, respectively) [14]. Bubbles inside the grains were found to have a specific crystallographic form e a shape of hexagonal prisms. It is likely related to beryllium lattice parameters [11]. Swelling of beryllium due to helium accumulation also limits the lifetime of beryllium pebbles [4,15]. Mathematical approach of swelling models has been provided by several authors [4,16,17]. The parameters used in these calculations are neutron flux, initial porosity, amount of impurities, grain size [17]. It has been observed that swelling is less intense if the grain size of beryllium is smaller [12,15,18]. Tritium is assumed not to form its own gas inclusions since its overall concentration is very low if compared with helium. It is believed to accumulate as an interstitial atom or to be trapped by the structural or chemical traps. In contrast to helium tritium has a high chemical reactivity and it can form chemical bonds with impurities, mostly with beryllium oxide forming the hydroxide. It has been found that besides oxide layer on the surface BeO inclusions have a tendency to accumulate on the grain boundaries [6], therefore it can be expected that chemically bonded tritium can be found either on grain boundaries or in the surface layer. However, it is assumed that most of the tritium resides mostly in the helium gas inclusions [19,20]. This assumption is also in good agreement with experimental data on tritium chemical forms beryllium showing that in neutron irradiated tritium is accumulated mainly as a
molecules [21]. To understand tritium behaviour in beryllium it is necessary to know such parameters as solubility, trapping energies and diffusivity. Some properties might be extrapolated from available data about protium or deuterium, however, isotopic effects must be taken into account [22]. In fact, it should be mentioned that in one of the first successful attempts to measure hydrogen solubility in beryllium tritium was used due to its simple detection based on its radioactivity [23]. Comprehensive overview of the experimentally obtained data on hydrogen solubility, diffusivity and permeation has been provided by R. A. Causey in 2002 [24]. Diffusion coefficient of neutron produced tritium in beryllium has been measured by several authors [25e27]. It is clear that overall tritium transport in neutron irradiated beryllium does not take place by a single mechanism of diffusion of tritium atoms dissolved in beryllium lattice. Trapping phenomena and influence of oxide layer have a crucial role. Theoretical model of tritium trapping on the imperfections of the lattice, such as vacancies, grain boundaries, etc. has been described [28]. Possible transport mechanisms of tritium in neutron irradiated beryllium has been studied by a number of authors, and a special computer codes or models have been developed [10,16,27,29e34]. 2.3. Thermal desorption of tritium form neutron irradiated beryllium Tritium and helium thermo-desorption spectra had been analysed widely in order to get comprehensive overview of the processes occurring during high temperature treatment of irradiated beryllium [4,10,26,32,35e40]. Earlier it was found that most of tritium releases together with helium, whereas helium can be released at temperatures close to the melting point of beryllium [41]. However, low temperature peaks of tritium release have been observed in a number of experiments. Some authors believe it is tritium located close to the surface and it is escaping through the micro cracks created by the irradiation or open porosity formed as a result of bubble coalescence along grain boundaries [37,42]. Other associates it with the diffusion of the tritium existing as an interstitial [10,35]. Tritium desorption process is affected by a number of factors, such as size and shape of the sample (diffusion length), structure (accumulation in the pores, grain boundaries, transportation along the cracks, etc), concentration of the helium gas inclusions and size of the gas bubbles (tritium accumulation together with helium), oxide content (tritium chemical bonding), etc. Therefore, desorption spectra might significantly differ. For instance, tritium desorption from pebbles of 0.1 mm mm in diameter, produced by Inert Gas Atomization method and irradiated with 2.70 1025 neutrons per m2 starts at ~360 K, whereas from pebbles of 1 mm in diameter, produced by Rotating Electrode Process and irradiated with (3e4) 1025 neutrons per m2 e at 950 K [43]. Neutron induced damage and helium bubbles also leads to tritium retention in the pebbles [31,44]. Tritium desorption from pebbles in which tritium is loaded by thermo-sorption process occurs at much lower temperatures than neutron irradiated pebbles of the same production batch [45,46]. 2.4. Beryllium oxidation At ambient temperature beryllium has a protective oxide layer of about 1 nm in thickness that prevents its further oxidation [47]. However, if the temperature is raised above 1000 K a breakaway oxidation reaction occurs and non e protective oxide layer starts to be formed [48]. This dramatic increase of oxidation rate has been explained by the fracture of oxide layer due to differences of thermal expansions of metallic beryllium and its oxide [49]. Beryllium oxidation below ~1270 K is believed to follow parabolic rate law,
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
295
Table 1 Irradiation conditions of beryllium pebbles at High Flux reactor. Irradiation time Neutron fluence (E>0.1 MeV) Irradiation temperature 4 He content
294 days 3e4 1025 m2 423-823 K 300-600 appm
Fig. 2. Tritium release, DTA signal and mass increase of neutron irradiated beryllium pebble during thermal treatment in Heþ0.1% H2 atmosphere (gradual heating 5 K/min up to 1600 K).
according to the first-principe calculations described in literature [54]. 3. Experimental Fig. 1. Experimental set-up for simultaneous Differential Thermal Analysis DTA, Thermal Gravimetric Analysis and Thermal Desorption Analysis TDA of tritium.
whereas at the temperature beyond this e linear rate law [49]. In the hot air beryllium nitride Be3N2 is also formed [50]. 2.5. Thermal properties and phase transition of beryllium Beryllium is in the hexagonal-close-packed (hcp) structure at ambient pressure and temperature. However, at temperature near its melting point beryllium transforms from hcp structure to bodycentred-cubic (bcc) structure. In the early work of A.J. Martin and A. Moor temperature of hcp-bcc or a / b transition has been experimentally demonstrated. Authors performed experiments with several beryllium grades by heating them in a vacuum at the rate of 1e5 K per minute. They observed a/ b transition in the temperature range between 1541 and 1548 C and melting in range between 1559 and 1564 K [51]. In more recent years H. Kleykamp has performed a thorough study of beryllium thermal properties and provided temperatures and enthalpies of both phase transition and melting of beryllium [52]. Comprehensive overview of thermal properties of beryllium can be found in paper of C.B. Alcock et al. [53]. Beryllium phase transition can also occur at high pressures
3.1. Materials In this study, beryllium pebbles of ~1 mm in diameter and produced by Rotating Electrode Process REP (manufacturer NGK Insulators Ltd., Handa City, Japan) were analysed. Main impurities of the samples given by manufacturer are BeO with content of 2300 ppm and Mg e 300 ppm. A portion of the pebbles were irradiated with neutrons in the Pebble Bed Assembly experiment at the High Flux Reactor in Petten, the Netherlands. Irradiation conditions are given in Table 1. Chemical composition and structure of irradiated and nonirradiated pebbles were analysed also prior to the annealing experiments. 3.2. Methods An experimental set-up has been developed to study tritium desorption, thermal properties and oxidation rate of beryllium. Method is based on simultaneous Differential Thermal Analysis DTA, Thermal Gravimetric Analysis and Thermal Desorption Analysis TDA of tritium. Schematic view of the experimental set-up is given in Fig. 1. Samples were heated in the ceramic crucible in furnace of
Table 2 DTA calibration results. Material
Mass, mg
Heating rate, K/min
Melting temp. Tmelt, K
Enthalpy of melting DmH, J/mol [55]
Peak area Q, mV$s
Sensitivity factor S, mV/mW
Au Au Au Au
151.5 151.5 151.5 151.5 Tmelt (exp.) Tmelt(teor.) 159.2 159.2 159.2 159.2 Tmelt(exp.) Tmelt(teor.)
5 5 5 5
1339 1339 1340 1339 1339 ± 1 1337.33 [56] 1234 1231 1231 1230 1232 ± 2 1234.93 [56]
12552.0 12552.0 12552.0 12552.0
6549 5446 5853 5439 SAu ¼
0.68 0.56 0.60 0.56 0,60 ± 0,05
11296.8 11296.8 11296.8 11296.8
9969 1047 1110 1251 SAg
0.54 0.63 0.67 0,75 0.66 ± 0.06
Ag Ag Ag Ag
5 5 5 5
Sensitivity factor used for calculation of beryllium phase transition and melting enthalpy e (0.63 ± 0.08) mV/mW.
296
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
Fig. 3. Structure of the neutron irradiated beryllium pebbles before treatment. Cross section of pebbles obtained by polishing off half of it by SiC sandpaper and then by diamond paste (1 and 0.05 mm particle size).
Fig. 4. Structure evolution of neutron irradiated beryllium pebbles after thermal treatment up to 923, 1123, 1223 and 1323 K (heating rate 5 K/min). Amplification factors chosen based on the scale of formed structure defects at the particular temperatures.
Fig. 5. Tritium release and structure evolution of neutron irradiated beryllium pebble during thermal treatment (heating rate 5 K/min).
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
differential thermal analyser TG/DTA (Seiko EXTAR 6300) and the gaseous tritium released from beryllium was measured by tritium monitor TEM 2100A with a proportional gas flow-through detector DDH 32. Heating rate was 2.4, 5, 20 and 40 K per minute and a mixture of He and 0.1% H2 were used as a purge gas with a purging rate of 200 mL/min. Composition of purge gas was chosen based on the planned condition in the future fusion devices. Samples were heated up to 1373, 1423 and 1600 K for tritium desorption and thermal analysis. Oxidation rate was studied by comparing the mass increase (TG signal) of the samples during analogous thermal treatment in air. For studies of structure evolution heating was stopped at lower temperatures: 823, 923, 1123, 1223, 1323 and 1523 K, respectively. Microstructure of samples was evaluated by means of scanning electron microscopy SEM (Hitachi S-4800, 15 kVacc, 15 mA). Chemical composition of the samples was assessed by the means on Energy Dispersive X-ray analysis that are conjugated with the SEM. DTA data was used for analysis of thermal properties. Calibration was performed with gold and silver reference materials. Sensitivity factor S for Au was found to be (0.60 ± 0.05) mV/mW, and for Ag e (0.66 ± 0.06) mV/mW (Table 2).
297
Fig. 7. Tritium desorption as a function of temperature depending on the heating rate.
4. Results and discussion Studies on behaviour of neutron irradiated beryllium during thermal treatment included gradual heating in atmosphere of helium with addition of hydrogen and in air. Example of the data obtained is shown on Fig. 2. First process occurring during thermal treatment is tritium release (red line) which starts at temperatures above 850 K. In temperature range between 1120 and 1370 K burst release was observed when the rest of radioactive gas is released. Mass increase of the pebble (black line) that corresponds to the oxidation process starts during or right after the burst release of tritium and becomes more intense close to the melting temperature of beryllium. Mass increase during thermal treatment in He þ 0.1% H2 atmosphere occurs due to the minor though existing presence of oxygen. DTA signal curve (green line) (in web version) clearly shows beryllium phase transition and melting process. In following chapters each process is described in more details. 4.1. Tritium release and structure evolution of pebbles during thermal treatment
Fig. 8. Beryllium pebbles before thermal treatment (on the left) and after thermal treatment up to 1620 K in air atmosphere.
inventory in pebbles, possibilities to prevent it as well as to the processes occurring in case of accidental temperature excursions when this inventory can be released in a very short period of time. SEM analysis of pebbles before thermal treatment revealed a large void in the bulk of the pebble (Fig. 3). These voids are generated during the cooling phase of the fabrication process and are developed to prevent beryllium swelling and failure when
Studies on tritium desorption process and its relation to microstructural evolution of beryllium pebbles gives a better understanding of issues related both to undesired formation of tritium
Fig. 6. Time of full tritium desorption depending of the heating rate.
Fig. 9. Mass increase of neutron irradiated and nonirradiated beryllium pebbles during thermal treatment up to 1620 K with heating rate 5 K/min. Mass increase corresponds to the formation of beryllium oxide.
298
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
Fig. 10. Combined SEM (SE) and EDX mapping images (Be, O) of irradiated pebble before thermal treatment, noniradiated and irradiated pebbles after thermal treatment up to 1600 K (heating rate 5 K/min, cooling rate e uncontrolled) in atmosphere of He þ 0.1%H2 with addition with air.
irradiated with neutrons e helium produced in the nuclear reactions is stored in this void instead of beryllium lattice [57,58]. However, this void also acts as a trap for tritium, therefore preventing its release during operation of the reactor. No other porosity was observed. According to the literature [14] helium inclusions after irradiation at this temperature do not exceed several nanometres; therefore they were not observable by the available equipment. After treatment up to temperature when tritium release has already started (920 K or ~0.6Tmelt), porosity has appeared (Fig. 4). The size of the pores reaches several hundreds of nanometres. By further raise of temperature the porosity increases (both concentration and size of the pores). At temperature above 1100 K formation of large radial cracks and connected pore channels starts (Fig. 4). These cracks and open porosity must be the pathway of the release of gases trapped in the technical void. This observation
hence explains burst release of the tritium near the particular temperature (Fig. 5). Tritium desorption experiments were performed with 4 different heating rates: 2.4, 5, 20 and 40 K/min and time required for full desorption estimated. If temperature increases with the rate 20 K/min entire tritium inventory is released in less than 30 min after temperature has reached ~1000 K (Fig. 6). It is an important issue since such a fast release of tritium inventory can be a significant threat for the operators of the reactor. In Fig. 7 tritium desorption as a function of temperature is shown. In the heating rate range 2.4e20 K/min one tritium desorption pattern remains e tritium desorption and burst release starts at approximately similar temperatures. Data from experiments with heating rate 40 K/min are not shown in the graph, since time for tritium to travel from the sample to the volume of detector were ~1e2 min (corresponding temperature increase is in range
Fig. 11. DTA signal (opposite values) of phase transition and melting of nonirradiated and irradiated beryllium samples (red marks indicate peak integration area). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
299
Table 3 Phase transition and melting enthalpies of irradiated and nonirradiated beryllium. Irradiation Mass, mg
Heating rate, K/ min
Phase transition peak area Q, mV$s
Enthalpy of phase transition DmH, J/ mol
mV$s
Enthalpy of melting DmH, J/ mol
Irradiated 0.96 Irradiated 1.01 Irradiated 1.34
5 5 5
196.6 201.3 263.0
2964 3014 2841
235.2 265.9 360.4
3547 3811 2841
Nonirr. Nonirr. Nonirr.
5 5 5
220.2 212.0 243.9
3663 3230 3239
248.1 260.0 249.0
4128 3962 3307
Melting peak area Q,
2900 ± 200 0.87 0.95 1.09
3800 ± 200
3400 ± 300
3800 ± 400
Melting enthalpy do not differ for neutron irradiated and nonirradiated beryllium, whereas phase transition enthalpy is slightly lower for neutron irradiated beryllium.
Fig. 12. Temperatures (a) and enthalpies (b) of phase transition and melting: comparison with the literature data (DTA method only) [51,52,60,61].
from 40 to 80 K) and therefore peak position cannot be estimated correctly. In more detail kinetics of tritium release behaviour during thermal annealing of beryllium pebbles at different temperature increase rates are described in an earlier publication [59]. 4.2. Oxidation of pebbles For oxidation studies beryllium pebbles were heated (5 K/min) up to 1600 K in air atmosphere and the mass increase corresponding to the growth of oxide layer was measured. At the stage when beryllium pebble is fully oxidized expected mass increase is ~180%. As a result of oxidation a brittle, flaking beryllium oxide pebble is formed (Fig. 8). For both irradiated and nonirradiated pebbles oxidation starts at temperature region between 1250 and 1350 K (Fig. 9). However, oxidation rate is higher for irradiated samples and most of the pebble volume is oxidized before beryllium melting whereas considerable oxidation of nonirradiated pebble appears only beyond melting temperature. Higher oxidation intensity of neutron irradiated pebbles can be explained by the structural changes that lead to increase of free surface area. It is demonstrated in Fig. 10 where SEM images of beryllium pebbles after thermal treatment in atmosphere of Heþ0.1%H2 with addition of air (<5%) combined with EDX elemental mapping data are shown. Oxidized areas can be identified by the increased oxygen presence (in web version) (blue colour in EDX mapping image).
Oxidation at operational temperature (~920 K for Be pebbles [1]) is not a critical issue e intense oxidation occurs at the temperature above 1200 K. However, it must be emphasized that in case of accidental temperature increase in combination of air exposure formation of toxic dust of beryllium oxide must be considered as an additional hazard besides tritium. 4.3. Phase transition and melting point In this study, effects of neutron induced lattice damage on the phase transition and melting points were attempted to find. DTA curves reveal two clearly visible peaks that correspond to phase transition and melting processes of the beryllium. In Fig. 11 the opposite values of DTA signal (with subtracted background) is shown for irradiated and nonirradiated pebbles. However, since the mass of samples was very small and the atmosphere of experiment was not fully inert, results of differential thermal analysis DTA give comparable instead of absolute value of enthalpy. Moreover, there is additional peak between phase transition and melting peaks that makes the integration of studied peaks complicated. This peak can be related to beryllium oxide or beryllium nitride since these compounds are already formed at this temperature as a result of beryllium interaction with air (Table 3). It was observed that both aeb transition and melting occurs at similar temperatures for neutron irradiated and nonirradiated beryllium. For nonirradiated beryllium in the experimental conditions transition occur at (1546 ± 2)K and melting (1562 ± 1)K,
300
E. Pajuste et al. / Journal of Nuclear Materials 465 (2015) 293e300
whereas for irradiated (1546 ± 1)K and melting (1561 ± 1)K, respectively. Obtained values were compared to the data provided by other authors (Fig. 12). The range of enthalpy values given in literature is considerably large: 2000e9100 J/mol. Values obtained in this study lay within this range. The discrepancy of the data obtained by different authors can be explained by the conditions of experiment (equipment, quality and grade of beryllium, atmosphere, and presence beryllium oxide) (Fig. 11). 5. Conclusions Tritium desorption from neutron irradiated beryllium pebbles starts at ~1000 K and is completely released at temperature ~1350 K, thus before beryllium melting. All of the tritium inventory can be released in less than 30 min in case of accidental temperature excursion if the temperature increase rate exceeds 20 K/min. High temperature oxidation process of neutron irradiated pebbles is more intense than that for nonirradiated pebbles. However, it is due to helium induced structure destruction that leads to an increased free surface Thermal properties of beryllium are not considerably affected by neutron irradiation with fluence (3e4) 1025 m2 (E > 0.1 MeV). Acknowledgements This work was supported by the European Communities within the framework of EFDA. The views and opinions expressed herein do not necessarily reflect those of the European Commission. References [1] L.M. Giancarli, M. Abdou, D.J. Campbell, V.A. Chuyanov, M.Y. Ahn, M. Enoeda, C. Pan, Y. Poitevin, E. Rajendra Kumar, I. Ricapito, Y. Strebkov, S. Suzuki, P.C. Wong, M. Zmitko, Fusion Eng. Des. 87 (5e6) (2012) 395e402. € slang, [2] M. Zmitko, Y. Poitevin, L. Boccaccini, J.F. Salavy, R. Knitter, A. Mo €sser, J. Nucl. Mater. 417 (1e3) (2011) A.J. Magielsen, J.B.J. Hegeman, R. La 678e683. [3] D. Carloni, L.V. Boccaccini, F. Franza, S. Kecskes, Fusion Eng. Des. 89 (7e8) (2014) 1341e1345. [4] F. Scaffidi-Argentina, M. Dalle Donne, C. Ferrero, C. Ronchi, Fusion Eng. Des. 27 (1995) 275e282. [5] S.J. Zinkle, 1.03-radiation-induced effects on microstructure, in: R.J.M. Konings (Ed.), Comprehensive Nuclear Materials, Elsevier, Oxford, 2012, pp. 65e98. [6] V.P. Chakin, A.O. Posevin, A.V. Obukhov, P.P. Silantyev, J. Nucl. Mater. 386e388 (2009) 206e209. [7] L.L. Snead, J. Nucl. Mater. 326 (2e3) (2004) 114e124. [8] V.P. Chakin, V.A. Kazakov, R.R. Melder, Y.D. Goncharenko, I.B. Kupriyanov, J. Nucl. Mater. 307e311 (Part 1) (2002) 647e652. [9] B.S. Hickman, G.T. Stevens, Effect of Neutron Irradiation on Beryllium Metal, Australian Nuclear Science and Technology Organisation, Sydney, 1963. [10] E. Rabaglino, C. Ronchi, A. Cardella, Fusion Eng. Des. 69 (2003) 455e461. [11] W. Van Renterghem, A. Leenaers, S. Van denBerghe, J. Nucl. Mater. 374 (1e2) (2008) 54e60. [12] I.B. Kupriyanov, R.R. Melder, V.A. Gorokhov, Fusion Eng. Des. 51e52 (2000) 135e143. [13] A. Leenaers, G. Verpoucke, A. Pellettieri, L. Sannen, S. Van den Berghe, J. Nucl. Mater. 372 (2e3) (2008) 256e262. [14] M. Klimenkov, V. Chakin, A. Moeslang, R. Rolli, J. Nucl. Mater. 443 (1e3) (2013) 409e416. [15] E. Ishitsuka, H. Kawamura, T. Terai, S. Tanaka, J. Nucl. Mater. 258e263 (1998) 566e570. [16] M. Dalle Donne, F. Scaffidi-Argentina, C. Ferrero, C. Ronchi, J. Nucl. Mater. 212e215 (1994) 954e960. [17] G.A. Sernyaev, A.S. Pokrovskiy, R.M. Bagautdinov, S.A. Fabritsiev, I.V. Mazul,
V.R. Barabash, J. Nucl. Mater. 233e237 (1996) 891e897. [18] I.B. Kupriyanov, G.N. Nikolaev, S.K. Zavjalov, L.A. Kurbatova, N.E. Zabirova, M. Zmitko, Fusion Eng. Des. 87 (7e8) (2012) 1338e1341. [19] J.M. Beeston, G.R. Longhurst, R.S. Wallace, J. Nucl. Mater. 195 (1e2) (1992) 102e108. [20] E. Hayward, C. Deo, J. Phys. Condens. Matter 24 (26) (2012) 265402. [21] E. Pajuste, A. Vitins, G. Kizane, V. Zubkovs, P. Birjukovs, Fusion Eng. Des. 86 (9e11) (2011) 2125e2128. [22] R. Kaur, S. Prakash, J. Phys. F: Metal Phys. 12 (7) (1982) 1383. [23] P.M.S. Jones, R. Gibson, J. Nucl. Mater. 21 (3) (1967) 353e354. [24] R.A. Causey, J. Nucl. Mater. 300 (2e3) (2002) 91e117. [25] I.L. Tazhibaeva, V.P. Shestakov, E.V. Chikhray, in: 18th Symposium on Fusion Technology, Germany, Karlsruhe, 22e26 August; Germany, Karlsruhe, 1994, p. 427. [26] D.L. Baldwin, M.C. Billone, J. Nucl. Mater. 212e215 (1994) 948e953. [27] S. Cho, A.R. Raffray, M.A. Abdou, J. Nucl. Mater. 212e215 (1994) 961e965. [28] M.G. Ganchenkova, V.A. Borodin, R.M. Nieminen, Phys. Rev. B 79 (13) (2009) 134101. [29] Y. Verzilov, S. Sato, K. Ochiai, M. Wada, A. Klix, T. Nishitani, Fusion Eng. Des. 82 (1) (2007) 1e9. [30] E. Ishitsuka, H. Kawamura, T. Terai, S. Tanaka, J. Nucl. Mater. 283e287 (2) (2000) 1401e1404. [31] I.B. Kupriyanov, V.A. Gorokhov, V.V. Vlasov, A.M. Kovalev, V.P. Chakin, J. Nucl. Mater. 329e333 (2004) 809e813. [32] F. Scaffidi-Argentina, M. Dalle Donne, H. Werle, Fusion Eng. Des. 58e59 (2001) 641e645. [33] S. Cho, M.A. Abdou, Fusion Eng. Des. 28 (1995) 265e270. €slang, R.A. Pieritz, E. Boller, C. Ferrero, J. Nucl. Mater. 386e388 (2009) [34] A. Mo 1052e1055. [35] E. Rabaglino, J.P. Hiernaut, C. Ronchi, F. Scaffidi-Argentina, J. Nucl. Mater. 307e311 (Part 2) (2002) 1424e1429. [36] F. Scaffidi-Argentina, M. Dalle Donne, H. Werle, J. Nucl. Mater. 258e263 (1998) 595e600. [37] A.K. Klepikov, I.L. Tazhibaeva, V.P. Shestakov, O.G. Romanenko, Y.V. Chikhray, E.A. Kenzhin, Y.S. Cherepnin, L.N. Tikhomirov, J. Nucl. Mater. 233e237 (2) (1996) 837e840. [38] I.B. Kupriyanov, G.N. Nikolaev, V.V. Vlasov, A.M. Kovalev, V.P. Chakin, J. Nucl. Mater. (2007) 367e370. Part A (0), 511-515. [39] C.E. Johnson, D.L. Baldwin, J.P. Kopasz, J. Nucl. Mater. 212e215 (2) (1994) 966e970. [40] A.V. Fedorov, S. van Til, L.J. Magielsen, M.P. Stijkel, J. Nucl. Mater. 442 (1e3, Supplement 1) (2013) S472eS477. [41] E. Rabaglino, Helium and tritium in neutron-irradiated beryllium, Forschungszentrum Karlsr. Karslruhe (2004) 127. [42] S. van Til, A.V. Fedorov, M.P. Stijkel, H.L. Cobussen, R.K. Mutnuru, P. v.d. Idsert, M. Zmitko, J. Nucl. Mater. 442 (1e3, Supplement 1) (2013) S478eS482. [43] E. Pajuste, Tritium Behaviour in Fusion Reactor Materials, University of Latvia, Riga, 2012. [44] S.L. Kanashenko, A.E. Gorodetsky, A.P. Zakharov, W.R. Wampler, Phys. Scr. 1996 (T64) (1996) 36. [45] V. Chakin, A. Moeslang, P. Kurinskiy, R. Rolli, H.C. Schneider, E. Alves, L.C. Alves, Fusion Eng. Des. 86 (9e11) (2011) 2338e2342. [46] A. Vitins, V. Zubkovs, G. Kizane, E. Pajuste, V. Kinerte, Fusion Sci. Technol. 60 (3) (2011) 1143e1146. € ri, In oxidation of beryllium e a scanning auger [47] C. Tomastik, W. Werner, H. Sto investigation, in: 20th IAEA Fusion Energy Conference, Vilamoura, Portugal, 1e6 November, 2004; Vilamoura, Portugal, 2004, pp. FT/P1eP29. [48] D.W. Aylmore, S.J. Gregg, W.B. Jepson, J. Nucl. Mater. 2 (2) (1960) 169e175. [49] R.O. Adams, J.T. Hurd, J. Less Common Metals 18 (4) (1969) 399e409. [50] N.N. Greenwood, A. Earnshaw, Chemistry of the Elements, second ed., Butterworth-Heinemann, Oxford, 1997, p. 340. [51] A.J. Martin, A. Moore, J. less-common metals 1 (1959) 85e93. [52] H. Kleykamp, Thermochim. Acta 345 (2000) 179e184. [53] C.B. Alcock, M.W. Chase, V.P. Itkin, J. Phys. Chem. Reference Data 22 (1) (1993) 1e85. [54] Z.-C. Guo, F. Luo, Y. Cheng, Comput. Mater. Sci. 84 (0) (2014) 139e144. [55] A.T. Dinsdale, CALPHAD 15 (4) (1991) 317e425. [56] H. Preston-Thomas, Metrologia 27 (1) (1990) 3. [57] Ishitsuka, E.; Kawamura, H.; Sakamoto, N.; Nishida, K. Process for Preparing Metallic Beryllium Pebbles. 5958105, sept.28, 1999, 1999. [58] Nishida, K.; Sakamoto, N. Metallic Beryllium Pebble. 08/16/1994 1994. ne, A. Matıss, E. Pajuste, V. Zubkovs, J. Nucl. Mater. 442 (1e3, [59] A. Vıtin¸s, G. K¸iza Supplement 1) (2013) S490eS493. [60] R.G. Loasby, D. Dearden, J. Less Common Metals 52 (1) (1977) 137e144. [61] A. ABEY, Pressure Phase Lines and Enthalpies for the Ab and B Liquid Transitions in Beryllium, 1984.