SUMMARY OF EXPERIMENTAL RESULTS FOR CERAMIC BREEDER MATERIALS

N.Roux - CEA/C.E Saclay (France) G.Hollenberg - Battelle, PNL - Richland (USA) C.Johnson - ANL - Argonne (USA) K.Noda - JAERI - Tokai-Mura (Japan) R.Verrall - AECL Chalk-River (Canada)

Abstract

Lithium containing ceramics were quickly recognized as promising tritium breeding materials for fusion reactor blankets owing, in particular, to their safety advantages. Relevant material properties were investigated to further evaluate their suitability. An extensive R&D program complemented a conceptual blanket design activity either for near term machines or for power reactors. All aspects were addressed: fabricability, overall properties (baseline, thermal, mechanical), compatibility with structures and , tritium release characteristics, irradiation behavior, activation, reprocessing and waste disposal issues. As a result of this investigation, the containing ceramics are considered to be excellent tritium breeding materials.

I. Introduction In the development of tritium breeding blankets for fusion reactors the lithium containing ceramics were quickly recognized as promising tritium breeding materials. Their excellent thermal stability and chemical inertness indicate favorable safety characteristics which is a principal attribute of fusion power. Further, no MHD effect is to be feared using these compounds. An extensive international R&D program was established to confirm the suitability of these attractive materials. Issues closely related to the primary functions of the blanket, i.e., tritium breeding, tritium release, and energy conversion were addressed. Properties characteristics necessary for design analysis, for assessment of blanket concepts performance, lifetime, reliability, and safety were investigated. Thus, the research programs focused on fabrication, properties measurements, compatibility with other blanket materials, tritium release behavior, irradiation behavior, activation characteristics, reprocessing, and waste disposal issues. As a result of this international collaboration, lithium and the ternary ceramics (lithium-aluminate, zirconate, silicate and titanate) were identified as promising candidates. The ceramic breeder research complemented conceptual blanket design activities. Two configurations of ceramics were considered, namely, sintered bodies and pebble beds. Following the of nuclear fission industry, blanket concepts using ceramics in pellet form were naturally envisaged. Thus, the EU DEMO BIT blanket concept is featured by rows of breeder modules containing annular LiA102 or Li2ZrO3 pellets.f I]. The pebble bed concept option is attractive as it partially alleviates cracking concerns. Also, loading of pebbles into complex geometries can be easier than with pellets, and intimate mixing of ceramic breeder and neutron multiplier can be facilitated by mixing pebbles of the two materials. For example, in the latest version of the EU DEMO BOT design pebbles of beryllium and Li4SiO4 are used. The larger 2 mm beryllium pebbles fill about 60% of the bed space. Spaces between the pebbles are filled with a mixture of 0.1 - 0.2 mm pebbles of Li4SiO4 and beryllium [2]. Likewise, the Japanese concept for ITER considers a layered pebble bed consisting of Imm Li2O pebbles separated from beryllium pebbles [3]. Blanket designs considered a lithium ceramic as a first option based upon its attractive properties. Complementary materials research developed the necessary data base and optimized the property characteristics as appropriate. Indeed, several properties of ceramics can be adjusted through material tailoring. 2. Experimental Results and Performance Analysis

Experimental results recalled in this section illustrate the breadth of knowledge acquired and highlight the excellent performance of ceramic breeders.

2.1. Fabrication

Fabrication of ceramic materials represents a determining step as it governs their performance. Fabrication includes preparation of powders and shaping to the required forms, i.e., pellets, or plates, and pebble beds. Pellets are porous, typically 75-80% T.D., in order to ease tritium release whereas pebbles are dense and porosity is that of the bed. Bed porosity can be adjusted through the number of pebble sizes and size ratios. The importance of microstructurefgrain size, , pore morphology) and impurities on properties and performance has long been recognized. Thus, there is a general trend for fine materials as early in-situ tritium release experiments indicated their better tritium release performance. In addition, they are expected to exhibit better mechanical behavior. Impurities should be kept at the lowest level as they can have an adverse effect on properties such as compatibility behavior and activation characteristics. Several preparation methods using liquid or solid routes were worked out and yielded powders with satisfactory purity levels [4-17]. Batch scale production C10-I00 Kg)was easily accomplished at KfK, Ceramics Kingston. Pechiney, and Temav. Likewise, fabrication into the required configurations was successfully accomplished using well proven processes in the ceramic industry in general and in the nuclear fission industry in particular. Scalability of the fabrication processes was explored. For example. L1AIO2 pellets prototypical of the first row breeder modules of the EU DEMO BIT concept were produced by cold uniaxial pressing and sintering by Temav/ENEA [18]. The pellets fulfill the micro structural specifications as defined in the material research, i.e.. - 0.5 |im grain size and 80% T.D and the geometrical specifications optimized in the design work, i.e., annular pellets with an aspect ratio (outside/inside diameter) of 1.8. Several hundreds of pellets were fabricated for laboratory tests and irradiation experiments. Likewise, Li2ZrO3 BIT pellets were produced in batches of hundreds by Uranium Pechiney [19,20] by uniaxial pressing of a spray dried Li2ZrO3 powder and sintering. Adjusting the process parameters allowed spanning the 75 - 85% T.D range while maintaining the grain size at -1 |im to preserve tritium release behavior. The process was transposed to the fabrication of LiAlCn pellets too. Fabrication of Li4SiO4. Li2ZrO3, and Li2O pebbles was successfully accomplished and scaled up to batch level [15.21-231. Thus. 10 Kg batch quantities were produced of a) 1.2 mm Li?ZrO3 spheres at AECL by extrusion, tumbling, and sintering, b) 0.4 - 0.6mm and 0.1 - 0.2 mm Li4SiO4 pebbles at Schott Glaswerke Mainz using a melt spraying process, c) 1 mm Li2O spheres at Kawasaki Heavy Industries using a rotating granulation method, and d) Li2TiO3 spheres at AECL. To date, fabrication scaleup has not raised any major problem, however, when industrial quantities become necessary further development will be required and sufficient lead time will have to be provided. Attention must be paid to storage of the finished product as ceramic breeders, especially Li2O, are sensitive to H2O and CO2 contamination. Irreversible alteration of the materials may result in degradation of material performance. In parallel, improvement of material properties was pursued through composition tailoring and micro structural tailoring. For example.'a) improvement of Li4SiO4 spheres mechanical stability was obtained by the addition of SiO2 [24]. b) lowering tritium release temperature of LiAlCo was obtained by Si substitution on the LiAlCn lattice (25], and c) thermal cycling performance of Li2ZrC>3 pellets was enhanced by density adjustment [20]. 2.2 Properties characteristics

A wide spectrum of properties characteristics necessary for blanket design analysis were determined. An overview is given here. Lithium atom density, which has an impact on tritium breeding ratio (TBR) and lithium burnup at end-of-life should be as high as possible. In this respect L12O is the most attractive ceramic. However, this factor is not as critical as previously thought since there is evidence that beryllium neutron multiplier and 6Li enrichment are necessary to achieve a TBR greater than unity. DEMO blanket designs do achieve a TBR > 1 even with ternary ceramics. Stability of the lithium ceramics at high temperature is obviously attractive from a safety perspective. In addition, the capacity for blanket operation at higher temperatures is beneficial to thermal efficiency. Melting points for candidate ceramics are all above 12000C indicative of good thermal stability. High temperature phases are considered so that physical changes cannot occur during operation. Vapor pressures of the compounds and their constituents are sufficiently low at the temperatures of interest so that no chemical change can occur either. Total panial pressures over the ceramic remain below 10"- Pa at anticipated blanket operating temperatures and even under limited upside transient conditions ensuring no significant material transport by the purge gas . Thermal properties, i.e.. thermal conductivity and linear thermal expansion, and their dependencies on temperature and. for the former, on microstructure were determined. Thermal conductivity of monolithic lithium ceramics is relatively low. in the range of 0.8 to 3.5 W/mK at 600°C for materials at 80% T.D. and could be'a drawback were it not for the skill of designers at accounting for this property in the design process. Pebble bed thermal conductivity and heat transfer coefficients between the walls and the pebble bed are important because they determine the operating temperatures and temperature gradients in the blanket. Effective thermal conductivity of pebble beds was determined under several different combinations of pebble size and density, packing fraction, purge gas composition, pressure, and flow rate. Measurements were made on Li4SiO4 pebble beds with 0.4-0.6 mm pebbles [26], on mixed pebble beds with 2 mm beryllium pebbles and 0.1-0.2 mm Li4SiO4 pebbles, and on beds with 2 mm beryllium pebbles and mixed 0.1-0.2 mm Li4SiO4 and beryllium pebbles [27]. The thermal conductivity and heat transfer coefficients for the mixed beds are considerably higher than for Li4SiO4 beds. Experimental results agree with predictions of models. Similarly, thermal conductivity of 1.2 mm Li2ZrO3 pebble beds was measured between 300°C and 1200°C [28]. Results follow closely the model predictions and are consistent with bed temperatures observed in irradiation tests. Independent measurements at 100°C at UCLA are in reasonable agreement. Finally, the thermal conductivity of 1 mm Li2O pebble bed as considered for the Japanese ITER design was measured [29]. Though ceramic breeders have no structural role, their thermomechanical behavior is important from a cracking resistance perspective. Breeder cracking is undesirable since it alters heat transfer characteristics and may cause tritium purge channel plugging due to transport and relocation of breeder fragments by the purge gas. Relevant materials characteristics were determined, e.g., elastic constants (Youngs modulus. Poissons ratio) and fracture strength [6.30-33] as well as their dependence on density, grain size, and temperature. The data base for baseline, thermal, and mechanical properties, for elastic constants and thermal creep of ceramic breeders was reviewed and assessed as part of the ITER/CDA activity [34]. Properties correlations were developed, they are reported in [35]. The thermomechanical behavior of both pellets and pebble beds was checked under thermal cycling 120.27.28,36.37]. The thermomechanical behavior of pellets for the EU DEMO BIT concept was checked out-of-reactor under conditions of heat generation, temperature, thermal gradient, cycling temperature range, and coolant flow rate, pressure and temperature prototypical of those encountered in the BIT blanket design. Thus, a 1 m long stack of pellets (-125 pellets) representative of those in the lrst row breeder modules were tested in the facility designed at ENEA. In this facility heat generation in the breeder is simulated by a wire resistor placed in the central channel formed by the annular pellet stack. Thermal cycles of 200 sec burn and 70 sec dwell are applied. Varying the power supplied to the resistor results in varying stresses in the pellets. The hydraulics of a helium purge flowing in the central channel is continuously monitored in order to identify any impact of pellet fracture. The fraction of broken pellets is determined after a number of cycles and the mode of pellet fracture is observed [381. Temav/ENEA LiA102 pellets at'80% T.D were cycled in the temperature range 500- 7000C showing no fracture after 10.000 cycles at 22 MPa equivalent tangential stress. 37c fracture after 20,000 cycles with no further fractures after 10.000 additional cycles at 32 MPa. Early tests of relatively low density Li2ZrO3 BIT pellets produced both at CEA and by Uranium-Pechiney showed poor behavior. However. Li2ZrO3 pellets with higher , i.e., 76.8% T.D (laboratory scale) and 83.5% T.D (industrial), cycled in the 330 - 550°C range performed well. No fracture was observed for the 76.8% T.D pellets after 10.000 cycles at 30 MPa. 14% fracture after 10.000 additional cycles at 40 MPa. No fracture was observed for the 83.5% T.D pellets up to 30.000 cycles at 47 MPa and 39% fracture after 10.000 additional cycles at 57 MPa indicating that the higher the density, the better the behavior [20]. Fig.l displays fracture probability versus fracture strength for pellets with different densities cycled during 100 cycles at increasing stress. These tests demonstrated the good behavior of both LiAlO2 and Li2ZrO3 pellets and emphasized the importance of fabrication routes and related materials characteristics on thermomechanical performance. Fracture occurred neatly along a diameter, suggesting that if pellets do fracture in a breeder rod there should be no detrimental consequence on the purge flow. This condition was indeed verified in the test itself as no change in the hydraulics of the purge flow was observed. The thermal cycling behavior of 1 mm pebbles was investigated for 92% T.D, 42 Hm grain size Li2O, 80% T.D and 40 |im grain size Li7ZrO3, and 93% TD. 40 [im grain size Li4SiO4 [36] under ITER representative conditions, i.e.. 400-8000C temperature range, heating/cooling rates of 20°C/s up to 2000 cycles.Whereas Li?O pebbles performed extremely well, Li2ZrO3 and Li4SiO4 pebbles fractured significantly. In contrast, tests of 0.4-0.6 mm Li4SiO4 pebbles containing 2.2% SiCb produced by Schott Glaswerke indicated good performance [24]. Tests of 0.4 - 0.6 mm Li4SiO4 pebbles cycled for 50 cycles. 2 mm beryllium pebbles mixed with 0.1 - 0.2 mm Li4SiO4 pebbles cycled for 1000 cycles, and 2 mm beryllium pebbles mixed with 0.1 - 0.2 mm Li4SiO4 and beryllium pebbles cycled for 500 cycles showed a good behavior too [27]. The latter indicated a critical temperature gradient of 38 - 42°C/s which is significantly higher that the peak value expected in the DEMO blanket (~ 20°C/s). Further, no fragment came out of the bed under purge gas velocities considerably higher than that anticipated in the concept option. Li2ZrO3 pebbles sintered under various conditions and. hence, exhibiting different densities and fracture strength were cycled from 2O0C to 10000C [37]. Results point out the influence of micro structural characteristics on pebble performance.

2.3. Compatibility behavior

Compatibility of cemmic breeders with other blanket materials is an important consideration with respect to safety and operating limitations. Compatibility behavior is sensitive to impurities. Numerous studies were made of compatibility with structural materials, particularly with austenitic and ferritic steels, both in vacuum and flowing atmospheres, with and without addition [39-43]. Tests with impurity free ceramics indicated a reasonably low interaction at temperatures of interest, i.e.. below 700°C. Penetration depth in steels were established for all ceramic breeders as a function of temperature [34]. Reaction rates increase in the presence of moisture. Effect of irradiation was explored in COMPLIMENT experiment (maximum lithium burnup of 1.4 % [33]. No irradiation induced processes were observed. Comparison of results with those in annealing tests under 1 Pa H2O pressure shows that the latter offer a conservative simulation condition [40], Compatibility of ceramics and beryllium was investigated as some blanket designs place the beryllium and ceramics in intimate contact. In laboratory tests the interaction of beryllium with the ternaries was found to be negligible up to 65O0C [41,44-461, it is larger for Li2O [36.46]. This is contrary to thermodynamic expectations because there is a large free energy driving force for oxidation of beryllium. However it is assumed that a thin Tayer of forms which protects the metal from further oxidation. The impact that neutron irradiation may have on the oxidation kinetics of beryllium was studied in SIBELIUS [47]. In this experiment the interaction of beryllium and lithium ceramics was studied in the mixed spectrum Siloe reactor core during 2000 hours at 55O0C (average materials interface temperature) and He +0.1% H2 purge. No irradiation effect was detected under the conditions explored and at ^Li bum-up of 20-40% for natural ^Li isotopic content [48.49].

2.4. Tritium release and recovery behavior

In an operating fusion reactor, the tritium breeding blanket will reach a condition in which the tritium release rate equals the production rate. The tritium release rate must be fast enough that the tritium inventory in the blanket does not become excessive. Slow tritium release will result in a large tritium inventory, which is unacceptable from both economic and safety viewpoints. As a consequence, considerable effort has been devoted to understanding the tritium release mechanism from ceramic breeders and beryllium neutron multiplier through theoretical, laboratory, and in-reactor studies. This information is being applied to the development of models for predicting tritium release for various blanket operating conditions. 2.4.1. Theoretical studies

The solid state defect structure of the lithium ceramic (lithium vacancy, defects, traps, etc. ) can strongly influence the tritium transport and release process. The origin of the lithium vacancy can arise from : (Ï) the ^Li (n.a)^T reaction, which generates many defects in transforming ^Li into ^T and "^He atoms: (2) defects created by displacement damage, i.e.. fast neutron scattering and recoil of energetic ^T and ^He atoms: (3) the extrinsic impurity-induced defects that control lithium diffusion, and (4) the intrinsic defects due to thermal equilibrium Tritium release is found to be controlled primarily by two processes: bulk diffusion in the grains and desorption from the grain surface. Also, in the helium purge gas is known to promote the release of tritium from candidate ceramic breeder materials. To better understand the role of hydrogen in the release process, calculations have been initiated to directly simulate the processes through which hydrogen can interact with lithium oxide surfaces. The methods that have been employed include a combination of ab-initio techniques in both crystalline and cluster environments. The method uses no adjustable parameters and may more accurately be described as computer experiments rather than calculations. Initially, attention is being given to examination of the interaction of the hydrogen molecule with the lithium oxide surface [50.511. The nature of the surface sites that may participate in the adsorption process are described in terms of Terrace-Ledge-Kink (TLK) terminology. Terrace sites are associated with the regular planar locations on the flat atomic surfaces, and kinks with the comers. The terrace sites have been examined and found to be energetically unfavorable with respect to hydrogen adsorption because of their relatively high coordination and. therefore, the small number of unsaturated bonds surrounding the sites. Early simulations suggest that hydrogen undergoes dissociative ehemisorption to low coordination sites. The concepts outlined above have been developed into computer models [52-58] that include the phenomena of bulk diffusion and desorption from the grain surface [52], bulk and grain boundary diffusion, desorption and solubility [53-57], and tritium transport through open porosity [58]. Surface heterogeneity was modeled using surface sites with different activation energies for tritium desorption [59]. Some success has been achieved in modeling tritium transport and release by assuming only diffusion and desorption processes [52,59]. The desorption activation energy may change with surface coverage because of the existence of multiple sites for adsorption [59]. At high surface coverage, both low and high energy sites will be occupied. However, it is not necessary to have different sites of adsorption for the desorption activation energy to be surface coverage dependent. The measured desorption activation energy may be due to interactions between adsorbed molecules on the surface. The interaction between the adsorbed hydrogen species (OH" or H") will affect the binding energy to the surface and, therefore, the desorption activation energy. While the computer model still needs further refinement, it can predict tritium inventory for end-of-life experiments. However, as yet. it cannot accurately predict tritium release curves from in-pile experiments over the length of an experiment. 2.4.2 . Laboratory Studies

Tritium release experiments were performed in various laboratories for Li20 [60.61), LiAlOo [62], U4S1O4 [63], Li?ZrO3 [64.65], and U2T1O3 [66.67]. These experiments have focused on determining the tritium extraction parameters, identifying the chemical form of the released tritium, and characterizing the rate-limiting process. For some materials, the tritium release rates have shown significant variance under test conditions. The reason for these variances is not fully understood but may involve poorly controlled experimental conditions, different sample characteristics, or unknown mechanistic effects. What has been understood is that surfaces play an extremely important role in the tritium release process. Studies [61] have demonstrated that limiting mechanisms are very dependent upon grain size in that desorption is limiting for small grain materials(< 200 ^m dia) and diffusion is limiting for large grain materials (>2000 \im dia). Desorption has been determined to be the rate-limitine step in several cases [62,68]. Adsorption/desorption experiments of water or hydrogen on lithium ceramics have examined the surface release mechanisms. Experimental data on water desorption from LiA102 and Li4SiO4 have been analyzed [69] and it was concluded that the process involved multiple desorption sites exhibiting different activation energies. The desorption of water from UA102 has been studied and sites with different activation energies have been identified [70.71]. Studies [72-74] using FTIR spectroscopy have reported several different OH (OD) species on a lithium oxide surface and identified surface exchange reactions using hydrogen isotopes. HT and HTO were the chemical forms of the desorbed tritium. The reaction rate for HT release was proportional to the residual tritium concentration and to the square root of the hydrogen concentration. This implies that the dissociation of hydrogen occurs on the ceramic surface, in agreement with the above theoretical calculations. 2.4.3. In-reactor Experiments

In-reactor experiments combine the effects of neutrons, temperature, and purge cas chemistry in a single test. To date, tritium release experiments have been conducted for the candidate materials : U2O [75,76], LiAlO2 [75.77]. Li2ZrO3 [75.77-79]. and Li4SiO4 [77.80]. One of the capsules in the BEATRDC II experiment [76] contained an Li2O ceramic pellet stack experiencing a large temperature gradient and provided some engineering performance data. In the in-situ tritium release experiments, tritium residence times of one day were found in helium with 0. \% H2 purge gas at ~ 300-3200C for L12O and Li2ZrO3. 3900C for Li4SiO4. and - 45O0C for LiAlC^fsee Fig.2. This is extremely satisfactory. The chemical form of the released tritium was both HT and HTO, the fraction of each depending on the potential of the system. Tritium diffusivities in the grain were measured for some of the ceramic breeders. Tritium diffusivity in the grain boundary was also estimated [81,82]. In most experiments, tritium release was not fully explained by bulk diffusion. Recent experiments [83.84] have shown that desorption is a first-order reaction, and the activation energy for desorption depends on surface coverage. The surface reaction was also studied by varying the H2 pressure. A one-half power dependence of the surface desorption rate constant for single-crystal Li2O and unity power dependence for sintered Li2O pellets has been reported [85]. Tritium inventories were measured at the end-of-life for several experiments and found to be well within current safety design criteria. Current ceramic breeder blanket designs include beryllium for neutron multiplication. Small quantities of tritium are generated in the beryllium via a ^Be neutron reaction producing tritium and helium. Because the buildup of the tritium could become a safety issue, there is strong interest in tritium removal. Initial tritium release studies on beryllium have been on material that has served as a reflector in a nuclear reactor. Generally, these reflector materials have remained in place for several years at low temperature, e.g.. < 100°C. When these reflector materials have been examined under isothermal anneal conditions, they have exhibited burst release of tritium at high temperatures. Experiments on lower density beryllium show significant tritium release a: lower temperatures. Anneals of high density (~ 100% T.D) beryllium discs from the SIBELIUS experiment [48.49.861 have shown that the bulk of the generated tritium was retained in the beryllium, and when the discs were heated to 65O0C and above, the tritium was readily released. The results indicate that tritium release from the beryllium did not exhibit burst release behavior, as previously reported, but rather an orderly release dependent solely upon temperature. Generally. - 99% of the tritium was released by 85O0C. In comparison with literature information on tritium release from ~ 100% dense beryllium, the SIBELIUS data for tritium release vs. temperature show a steeper slope. Also, the literature information shows that for lower density beryllium (80.9% T.D). tritium release is considerably enhanced. Tritium release from the ceramic discs in the SIBELIUS experiment was quite similar to the behavior shown in other dynamic tritium release experiments on lithium ceramics. 2.5. Irradiation behavior

The components of an operating fusion blanket must tolerate high levels of neutron exposure. Breeder blanket lithium ceramics have the added burden of remaining functional while 10 to 25% of the lithium atoms are transformed into helium and tritium atoms. With no significant amount of irradiation data available, worst case scenarios for the impact of irradiation damage in lithium ceramics were generated in order to provide a theoretical envelope for designers to work. For example, the temperature limits for lithium oxide were considered very restrictive at one rime [871. Uncertainty in irradiation performance and volatility of lithium were considered limiting issues. In addition, it was considered possible that irradiation effects in lithium ceramics could significantly increase tritium inventory at high burnups. Today irradiation test results indicate that candidate lithium ceramics readily survive the rigors of neutron irradiation and achieve significant burnups. 2.5.1. Materials performance

Numerous tritium release experiments have been conducted at low lithium burnup levels in order to identify the tritium release kinetics. The functional requirement of the blanket is to achieve very low tritium inventory during operation and to minimize releases during an accident scenario. Results from the BEATRIX II experiments have provided in-situ tritium recovery data at both low and extended burnups (5%) for lithium oxide and lithium zirconate. The data of Fig.3 show a low and stable tritium inventory at high burnup. Although the shape of the curve changed somewhat during testing the inventory changes (i.e.. 0.14. 0.06. 0.09 Ci) confirm that no increase in tritium inventory resulted from irradiation damage [88]. For ternary ceramics, i.e.. Li2ZrO3. LiAlO2, Li4SiO4, the effects of irradiation to 3% burnup in a thermal neutron environment produced very little change in the tritium release behavior during thermal reactor testing [77]. The ongoing irradiation of lithium ceramics to 10% burnup should contribute to our confidence in maintaining low tritium inventory at operating temperatures. As part of the investigation of lithium ceramics behavior in a neutron environment short-term irradiations were conducted on the ternary ceramics in thermal neutron reactors. They are featured by low lithium burnups (1 to 3%) and low dpa (~ 2). temperature levels of 400. 600, and 8000C. and high heating rates inducing large thermal gradients. Post irradiation examination focused on dimensional stability, physical integrity, and microstructural and mechanical testing [33.89.90]. Fragmentation, which wasassigned to partly result from large thermal gradients prevailing in the specimens and to thermal shocks due to reactor trips, was observed in some cases. Decreases in Young's modulus and fracture strength values could be noted. Thermal conductivity, as estimated from records of thermocouples inserted in the center and edge of the specimens, was found to be very stable. As a whole no significant changes were observed under the conditions explored. Data from fast neutron reactor experiments are more relevant as they correspond to higher bumups and displacement damage. The ternary ceramics were found to posses relatively small swelling rates while Li2O exhibited high swelling at modest bumup levels (3%). However, radiographs of a recent fast reactor experiment [91], with burnups in the range of 8 to 10% lithium, demonstrated the stability of the materials under very severe temperature gradients due to the large pellet diameter and high heat generation rates. For example. LiAlCb was exposed, for almost one year, to a 11250C centerline temperature while Li2ZrO3 was subjected to a 1225°C centerline temperature with an edge temperature of 6000C. From the radiographs it was noted that the fine grained LiAlO2 pellets actually contracted while the larger grain LiAlCn pellets expanded slightly. Considering the potential for sintering of fine grained ceramics in this extreme environment these results are consistent. The Li2ZrO3 pellets appeared to expand slightly in both the axial and radial directions, perhaps more from crack opening displacements than from microscopic swelling. The Li4SiO4 and Li2O pellets appeared to expand until the gap around them was closed. Both of these materials are considered to be relatively plastic under these conditions, hence, no deformation of the cladding was observed. Smaller samples under smaller temperature gradients exhibited similar behavior. The swelling in the L12O pellets appears to be the same as that observed at lower burnup levels. A detailed examination of the radiographs reveals that both axial and radial cracks existed as the result of the extreme thermal stresses. The cracking may be the source of the observed swelling in the ternary ceramics. However, in the case of the fine grained lithium aluminate column, the thin disks at the top of the column did not exhibit significant cracking. The cracking in the BEATRIX I test pellets due to the large thermal gradients can be alleviated by appropriate designing for example, thin walled rings or small diameter spheres [1,2]. Ir. the BEATRIX I experiment cracking of spheres was not observed and cracking was not prevalent in the smaller diameter high burnup ternary ceramics which possessed less than 1000C temperature difference across the pellet. Lithium ceramic cracking may be tolerated in certain blanket designs. For Li?O in particular, the transport of lithium was considered to be a design limiting phenomena. A design requirement is that the breeder material not be relocated such that neutron streaming or decreases in the designed tritium breeding ratio might occur. In the BEATRIX II experiment, the operating temperatures in the centerline of the solid pellet were measured to be 10000C and an edge temperature of 550°C. Quantitatively, downstream transport of lithium was not observed in piping or other components. A central hollow core developed in these pellets, but that was considered to be the result of sintering rather than vapor transport [92]. Recent thermodynamic data [93] supports the experimental finding that little lithium transport is to be expected as the vapor pressure of , the species transporting lithium, is a factor of 30 lower than previously stated. Hence, lithium transport does not appear to be an issue, even for L12O at these extreme conditions. Concern was expressed that neutron irradiation may degrade thermal conductivity of lithium ceramics and therefore limit the lifetimes of these materials. After irradiation to 3% burnup there was a significant reduction in thermal conductivity of Li2O and LiAlOT at low temperatures, < 4000C. At higher temperatures, the thermal conductivity values were either unchanged or even higher than unirradiated values [94]. These results were confirmed by in-situ measurements in the BEATRIX II experiment on Li2O in which a central and edge thermocouple measured the temperature difference generated by the neutron and gamma heating developed in the ceramic [88]. The temperatures during irradiation were equivalent to those predicted from literature data and were seen to be stable. A similar stability was observed during irradiation of Li2ZrO3 [79].

2.5.2. Fundamental aspects

In parallel to irradiation performance testing, basic studies were launched. Irradiation induced defects such as F+ centers, colloidal lithium metal, metal, and some decomposition products in L12O, LiA102. Li4SiO4, and Li2ZrO3 have been studied using neutron, heavy-, or electron irradiaiion to gain a fundamental understanding of irradiation behavior[95-100]. Fundamental studies of irradiation effects on lithium ion have been carried out for Li2O and Li4SiO4 by measuring the ionic conductivity of ion irradiated specimens [101.102]. These low temperature (< 6250K) studies may relate to tritium transport in lithium ceramics assuming tritium diffusion in lithium ceramics is closely related to the lithium ion diffusion. In one study, irradiation damage in Li2O from fast neutron irradiation was investigated in a series of experiments in the BEATRIX II program [100]. Electron spin resonance and optical absorption measurements were carried out on Li2O after irradiation in FFTF for 300 effective full power days(fluence of 3.9x10^6 n/m2) at about 65O0K. The electron spin resonance data were interpreted to indicate the presence of colloidal lithium. Since Li2O is a very stable material, a regime of non stoichiometry is needed to accommodate the presence of colloidal lithium at this temperature. This subject needs further study.

2.6. Activation and ^Li management

Evaluation of neutron activation of candidate ceramic breeder materials is important from the perspective of fusion reactor maintenance, waste disposal, and recovery of ^Li. This evaluation depends, to some extent, on blanket design conditions, i.e.. materials proportions, neutron fluxes, and neutron wall loading. Activation calculation were made for candidate ceramic breeders and structural material under comparable conditions and allow for a ranking of the materials. For example, the neutron activation of the materials has been calculated for a helium cooled ceramic breeder blanket design exposed to a total integrated neutron dose of 20 MWA/m^ (5 MW/m^ and 4 years operation) [103]. Activation levels at all times after shutdown, biological dose rates, decay heat values and waste disposal ratings are reported. Likewise calculations were made for an outboard blanket region zone of the EEF reactor with neutron flux 4.18xlO18/m2s. integrated dose of 12.5 MWy/m- (5 MW/m^ and 2.5 years of operation). Specific activity at various cooling times, surface dose rate, ingestion and inhalation dose, decay power, and waste management options were considered [104]. Though conditions in the two studies are not readily comparable, predicted activities agree well within a factor of six. The main conclusions are : - The dominant activity in the ceramics arises from the generated tritium. However, residual tritium is expected to be very low in the ceramics after service, otherwise the burnt material will be accessible to a wide range of processes to remove tritium to the lowest practical level. - In the absence of tritium and of troublesome impurities, the intrinsic activities which are dominated by short lived nuclides rank L12O < Li4SiO4 < LiAlCb < Li2ZrC>3 with that of Li2O being three orders of magnitude lower than that of Li2ZrO3. - In all cases, activation of structural materials is higher than that of the ceramic breeders. Thus, the ceramic breeder material will not be the dominant contributor to the gamma activation leading to the limitation of maintenance. With respect to waste disposal, one must consider the production of long-life radionuclides. For the ceramics, the radioactivity after one year is very low in comparison with that of the structural material. The long lived radionuclides such as -6 Al (half life 7.2xl(P years) and 94 Nb (half life 2xlO4 years) are decay products from LiAICb, Li4SiO4. and Li2ZrO3. Li2O does not present a problem for waste disposal, while careful attention to blanket design for the other ceramics should afford some relief on this issue. Reprocessing of ceramic breeders for recovery of ^Li. was investigated for the oxide, aluminate, orthosilicate and zirconate. Dissolution of these ceramics using water and various acids and lithium recovery processes with precipitation methods using ammonium carbonate or ammonium hydroxide and solvent extraction methods were examined [105]. A high dissolution rate (92-100%) and lithium recovery rate (80 to 100%) were attained. Also, preliminary lithium recovery experiments using neutron irradiated lithium aluminate were carried out by the precipitation method and it was confirmed that lithium can be recovered at high efficiency (- 78%). Thus, reprocessing for recovery of lithium was demonstrated by use of this chemical process. Methods for recovery of ^Li were also investigated in [106]. Solvent extraction and ion exchange were considered as potential routes for separation of lithium from the used breeder. To evaluate the feasibility of reprocessing of ceramic breeder materials one must consider radioactivity and residual tritium. The residual tritium levels for ceramic breeders may not be a problem since the solubility of tritium in the ceramic breeder is very low and most of the tritium can be released from the ceramic by heating before reprocessing. With respect to gamma activation, consideration may not be needed for Li2O and Li4SiO4, while some consideration must be given to LiA102 and Li2ZrO3. However it is shown that all materials can be handled remotely in radiologically controlled enclosures with a moderate degree of gamma-ray shielding within a cooling time of 10 years [ 106], These conclusions support the interest currently placed in the investigation of Li2TiO3 which is a low activation ceramic that exhibits overall attractive properties, especially tritium release characteristics, as promising as those of Li2ZrO3 [66.67]- 3. Conclusion

This summary of experimental results shows that aspects relevant to the utilization of ceramic breeders for fusion reactor blankets have been exhaustively covered. Examination of the properties data base, of out-of-reactor tests, irradiation experiments, and performance analysis results highlights the functionality of ceramic breeders and does not reveal any negative. As a result of continuous progress, current generation ceramic materials are adequate to fulfill the requirements for several attractive ceramic breeder blanket options being developed for both near term machines and for power reactors. Advanced ceramic breeders including other ceramic compounds (litliium titanate) as well as added value products are still expected. Work still remains to be done. Specific areas for further research include : a) fundamental studies to fully understand and control pertinent phenomena and help materials improvement, b) tritium release modeling to predict tritium inventories in the whole range of fabrication, operation, and accident conditions, c) development of advanced materials and industrial fabrication capacity, and most important, d) evaluation of materials behavior at end-of-life, i.e.. 20 - 30 % lithium burnup and high displacement damage. To date, safety characteristics, satisfactory properties data base for candidate materials, successful development of pilot scale fabrication using well proven technology, excellent tritium release behavior, very good irradiation performance to extended bumups, good prospects regarding ^Li recovery, materials reprocessing and waste disposal designate lithium ceramics as excellent, first option tritium breeding materials. 4. References [11 E. Proust. L. Anzidei. N.Roux, T.Flament. V.Levy. S.Casadio, G.Dell'Orco. Experimental program in support of the development of the European ceramic breeder-inside-tube test blanket : present status and future work. Fus.Tech. VoI 21 (1992)p.2089-2098.

[2] M. Dalle Donne. E. Bojarsky, U. Fisher, A. Goraieb. P. Norajitra. G. Reimann. H. Reiser, and G. Sordon. Conceptual design of a helium cooled solid breeder blanket based on the use of a mixed bed of beryllium and Li4SiO4 pebbles. Proceedings 17th SOFT. Rome (1992), p 1326.

[3] H.Takatsu, S. Mori, H. Yoshida, T. Hashimoto. T. Kurasawa, K. Koizumi. K. Ioki. and T. Kanzawa. Layered pebble bed concept for ITER breeding blanket. Proceedings 17th SOFT. Rome (1992) p. 1504

[4] R.M. Arons, R.B. Poeppel, M. Tetenbaum, and CE. Johnson. Preparation, characterization, and chemistry of solid breeder materials, J. Nucl. Mater. 103-104 (1981)555-559.

[5] T.Takahashi and H. Watanabe. Preparation and characterization of lithium oxide. Fusion Engrg. and Des. 8 (1989) 399-405.

[6] B.Rasneur, Determination of mechanical characteristics of LiAlO2 and Li2ZrO'V Pioc. 15th SOFT, Utrecht (1988) p. 1030.

[7] C. Alvani. S. Casadio. L. Lorenzini, and G. Brambilla. The ENEA selected fabrication process of LiAlO? ceramic pellets for tritium breeding experiments in fusion technology. High Tech. Ceramics. (North Holland. 1987) pp. 2941.

[8] P. Kennedy. K.E. Gilchrist. D.E. Walker, and S. Broughton, The preparation, characterization, and properties of lithium oxide and lithium metazirconate specimens irradiated in HFR Petten in the second and third EXOTIC experiments. Proc. 14th SOFT. Avignon (1986) pp. 1013-1017.

[9] C. Alvani. S. Casadio. L. Contursi. and F. Fava. Gamma lithium aluminate fabrication by microwave heating process. Ceramics Today - Tomorrow's Ceramics. Elsevier ( 1991) p. 3069

[ 1Oj D. Vollath and H. Wedemeyer. On the preparation of the lithium silicate series from Li2SiO3 to LisSiO6 in alcoholic media, J. Nucl. Mater. 141-143 (1986) 334-338.

[11] D.J. Suiter. J.W. Davis, and B.A. Kirkpatrick, Preparation of solid tritium breeding compounds for fusion reactors, J. Nucl. Mater. 103-104 (1981) 579-584.

[12] CAV. Turner. B.C. Clatworthy, and A.Y.H. Gin. The preparation of lithium aluminate by the hydrolysis of lithium and aluminum alkoxides. Advances in Ceramics. VoI 25 (American Ceramic Society, Westerville. OH 1989) pp. 141-148.

[13] J.W. Kammeyer and OJ. Whittemore, Reactive lithium zirconates. Advances in Ceramics, VoI 25 (American Ceramic Society, Westerville. OH 1989) pp. 117-122. [14] M. Smaihi. D. Peril. J.P. BoiJoi, P. Bergez, and A. Lecompre, Lithium silicate gels: growth, structure, and thermal evolution. Advances in Ceramics, VoI 27 (American Ceramic Society, Westerville, OH 1989) pp. 23-34. [15] J.D. Sullivan. CL. Brayman. R.A. Verrall. J.M. Miller. PJ. Gierszewski. F. Londry, and A. Slavin. Canadian ceramic breeder sphere-pzc technology: capability and recent results. Fusion Engrg. and Des. 17 (1991 i 79-1*5.

[ 16] D. Vollath, H. Wedemeyer. E. Gunther. and H. Elbel. Semi-industrial production of Li4SiO4 powder and "spheres. Fusion Engrg. and Des. 8 ( 1989) 415-419.

[171 O. Renoult. J.P. Boilot. and M. Boncoeur. Alkoxide-hydroxide route to prepare UA1O2 based ceramics, J. Amsr. Ceram. Soc. 77 ( 1 ) ( 1994) 249-253.

[18] C. Alvani. S. Casadio. L. Contursi, N. Zaghini. M. Zanardi. LiAlCn powder production and characterization as breeder material. Proc. Workshop on Thermomechanical Behavior of Ceramic Breeder Materials for Fusion Technology. Medicina. Italy (1992).

[19] B.Bastide and N.Roux. Lithium zirconate elements fabricated by industrial scale process. Ceramic Transactions, VoI 27 (American Ceramic Society. Westerville. OH 1992) pp. 107-116.

[20] N. Roux. J. Mougin, B. Rasneur, E. Proust, L. Giancarli. J.F. Salavy. Current material and design studies for the utilization of Li2ZrC>3 in the European BIT blanket conctpt, Proc. ICFRM-6 September 27- October 1. 1993. Stresa, Italy.

[21] G. Schumacher, M. DalleDonne. and S. Dorner, Properties of lithium orthosilicate spheres, J. Nucl. Mater. 155-157 (1988) 451-454.

[22] T. Suzuki. O. Murata. and S. Hirata. Development of small spherical lithium ceramics for fusion reactor blanket. Ceramic Transactions, VoI 27 ( American Ceramic Society, Westerville. OH 1992) pp. 37-56.

[23] N. Asami, N. Nagashima. M. Nakagura. A. Akiyama. N. Suemori and A. Ohya. Development of fabrication methods for lithium ceramics. Ceramic Transactions. VoI 27 (American Ceramic Society. Westerville. OH 1992) pp. 17-36.

[24] G.Schumacher, M. Dalle Donne, V. Veiler. R. Huber and J. Schub, Improvement of the mechanical stability of lithium orthosilicate pebbles. Fusion Encrg. and Des. 17(1991)31-36.

[251 O. Renoult. J.P. Boilot. J.P. Korb. and M. Boncoeur, LiA102-based ceramics: sol-gel preparation and out-of-reactor tritium release experiments. Proc. Second International Workshop on Ceramic Breeder Blanket Interactions. Paris. France (1993).

[26] M. Dalle Donne and G. Sordon. Heat transfer in pebble beds for fusion blankets. Fusion Technology 17 (1990) 597.

[27] M. Dalle Donne. A. Goraieb. R. Huber. R. Schmitt. G. Schumacher. G. Sordon. and A. Weisenburger, Heat transfer and technological investigation on mixed beds of beryllium and Li4SiO4 pebbles. Proceedings of ICFRM-6.

[28] P. Gierszewski, H. Kawamura, M. Daile Donne, and M. Tillack. Ceramic pebble bed development for fusion blankets, this Conference.

[29] M. Eneoda, S. Satoh. T. Kurasawa. and H. Takatsu, Measurement of effective thermal conductivity of lithium oxide and beryllium sphere packed bed. Proceeding of the Symposium on Fusion Engineering, 1993. [3O) M. Barra Caracciolo. C. Alvani. and S. Casadio. Mechanical properties of L1AIO2 ceramic breeder pellets. Fusion Technology, 14 (1988) p 360.

[31] C. Alvani. F. Alessandrini. M.R. Mancini, and S. Casadio. Environmental attack effects on Li2ZrO3 and properties evolution as a function of the fabrication route. Proc. International Workshop on Ceramic Breeder Blanket Interactions, Paris. 1993.

[32] B. Rasneur, J. Mougin. and N. Roux. Study of the thermomechanical and tritium release behavior of\i2Zr03 as tritium breeding ceramic for the European BIT blanket concept. Proc. ICFRM-6.

[331 W- Dienst and H. Zimmerman. Strength and chemical reactivity of ceramic breeder material near operation conditions. Proc. ICFRM-6.

[3 '] K. Noda. M. Billone, W. Dienst. T. Flament. P. Lorenzetto. and N. Roux, ITER Specialists Meeting on Blanket Materials Data Base ( 1990).

[35] M.C. Billone, W. Dienst. P. Lorenzetto. K. Noda, and N. Roux, Solid breeder and beryllium materials data base, this conference.

[36] M. Enoeda. H. Yoshida. H. Takatsu, T. Kurasawa. M. Nagakura. S Kobayashi, M. Nishida. and Y. Ozawa. Fabrication and compatibility measurement of lithium ceramics. Proceedings second International Workshop on Ceramic Breeder Blanket Interactions. Tokyo. Japan 1992.

[37] R. Verrall, AECL experience with Li2ZrO3 and Li2TiO3, Proceedings second International Workshop on Ceramic Breeder Blanket Interactions. Paris. France 1993.

[38] G. Dell'Orco and G.C. Bertacci. Experimental thermomechanical testing apparatus for a DEMO relevant blanket desisn cf a ceramic cladded pin. Fusion Engrg. and Des. 17(1991) 161-164.

[39] P. Hofmann and W. Dienst. Compatibility studies of metallic materials with lithium based . J. Nucl. Mater. 155-157 (1988) 485-490.

[40] P. Hofmann and W. Dienst. Chemical compatibility' of oxide breeder materials with cladding steels. Proc. 17th SOFT. Rome (1992) p. 1374.

[41] P. Hofmann and W. Dienst. Chemical interactions of lithium based oxides with beryllium and stainless steel. Proc. 16th SOFT. London. ( 1990) p. 777.

[42] R.J. Pulham. W.R. Watson, and N.P. Youna. The corrosion of steels by lithium oxide and lithium silicate. Proc. 14th SOFT. Avigon. (1986) p. 963.

[43] H. Yoshida. S. Tanaka. M. Enoeda. T. Kurasawa. N. Asami. K. Nagashima. M. Nakagura. H. Akiyama. A. Ohya, M. Nishikawa. and H. Imaizumi. Integrity and compatibility study of blanket materials. Ceramic Transactions. VoI .17 (American Ceramic Society. Westerville. OH 1992) pp. 175-192

[44j A. Terlain. D. Herpin. P. Perodaud. T. Flament. and J. Sannier. Compatibility problems with bervllium in ceramic blankets. Proc. 15th SoFT. Utrecht. (1988) p. 1179.

[45] M. Broc. T. Flameni. A. Terlain. and J. Sannier, Compatibility problems in tritium breeding blankets. J. Nucl. Mater. 179 (1991) 820. [46] T. Flament. D. Herpin. 1. Feve. and J. Sannier. Compatibility of stainless steels and lithium-based ceramics with beryllium. J. Nucl. Mater. 191-194 (1992) 163.

[47] N. Roux. M. Briec. M. Bruet. T. Flament. C. Johnson. M. Masson, A. Terlain. and F. Toumebize. The SIBELIUS experiment: a study of the irradiation behavior of beryllium/ceramic and beryllium/steel compacts. J. Nucl. Mater. 179 (1991) 827.

[48] N. Roux. JJ. Abassin. M. Briec, D. Cruz. T. Flament and I.Schuster. Compatibility behavior of beryllium with UA102 and Li2ZrO3 ceramics, with 316L and 1.4914 steels in SIBELIÛS.J. Nucl. Mater. 191-194 (1992) 168.

[49] W. Dienst. D. Schild. and H. Werle. Tritium release from Li4SiO4. UnO. and beryllium and chemical compatibility of beryllium with Li4SiO4. U2O. and steel. KfK Report 5109. December 1992.

[50] S.W. Tarn. Argonne National Laboratory. Private Communication ( 1994)

[51] D. Yamaki, S. Tanaka. M. Yamawaki. Luminescence from Li2O under He ion and neutron irradiation, Proceedings 2nd International Workshop on Ceramic Breeder Blanket Interactions, September 22-24,1993, Paris. France.

[52] J.P. Kopasz. S.W. Tarn, CE. Johnson, Modeline of tritium behavior in ceramic breeder materials. J. Nucl. Mater. 155-157 (1988) 500.

[53] K. Noda, T. Kurasawa, and H. Watanabe, Modeling tritium behavior in fusion blanket ceramics. Advances in Ceramics, Vol. 25 (American Ceramic Society. Westerville, OH. 1989) pp. 73-82.

[54] M.C. Billone. H. Attaya. Y. Gohar, and CC. Lin, Tritium retention and release analysis for the US-ITÉR driver blanket. Fusion Technol. 19 (1991) 976-983

[55] S. Tanaka. D. Yamaki. and M. Yamawaki, Modeling of surface reactions in tritium release from solid breeding materials, ibid., p. 1023.

[56] G. Federici. A.R. Raffray, and M.A. Abdou. A comprehensive model for tritium transport in lithium-base ceramics. Part I: theory and description of model capabilities. J. Nucl. Mater. 173 (1990) 185.

[57] G. Federici. A.R. Raffray, and M.A. Abdou. A comprehensive model for tritium transport in lithium base ceramics. Part II: comparison of model predictions with experimental results, ibid., p. 214.

[58] S.W. Tarn, Argonne National Laboratory, Private Communication (1991).

[59] J.P. Kopasz. S.W. Tarn, and R.A. Verrall. Modelinc unusual tritium release behavior from Li2O. Fusion Technol. 15 (1989)1217-1222

[60] J.M. Miller. R.A. Verrall. S.R. Bokwa. and.O.H. Rose. The effect of material characteristics on tritium release from lithium oxide. Advances in Ceramics. Vol. 25 (American Ceramic Society, VVesterville. OH. 1989) pp. 53-60.

[61] J.F. Quanci, PhD Thesis. Princeton University, January 1989.

[62] F. Botter, S. Tistchenko, M. Briec, and J.P. Kopasz. Progress in the knowledge of the mechanisms of tritium release from lithium ceramics. Fusion Engrg. and Des. 17 (1991) 49-54. |63] W. Breitung. H. Elbel. J. Lebkucher. G. Schumacher, and H. Werle. Out-of-pile tritium extraction from lithium silicates, J. Nucl. Mater. 155-157 (1988) 507.

164] K. Okuno and H. Kudo, Tritium diffusivity in lithium-based ceramic breeders irradiated with neutrons. Fusion Engrg. and Des. 8 (1989) 355-358.

[651 J-M. Miller, S.R. Bokwa, D.S. MacDonald, and R.A. Verrall. Tritium recovery from lithium zirconaie spheres. Fusion Technol. 19 (1991) 996-999.

[66] J.P. Kopasz. J.M. Miller, and CE. Johnson. Tritium release from lithium titanate a low activation material. Proceedings ICFRM-6. September 27-October i.1993. Stresa. Italy. [67] J.M. Miller. H.B. Hamilton, and J.D. Sullivan. Testing of lithium titanate as an alternate blanket material. Proceedings ICFRM-6.

[68] J.P. Kopasz. CA. Seils. and CE. Johnson. Spatial tritium transport in single crystal lithium aluminate. Proceedings ICFRM-6. September 27 - October 1.1993, Stresa. Italy.

[69] J.P. Kopasz. A.K. Fischer. CE. Johnson. Multiple activation energies for tritium release from ceramic breeders. Advances in Ceramics. Vol. 27 (American Ceramic Society, Westerville. OH. 1990) pp. 317-328.

[70] A.K. Fischer and CE. Johnson. Desorption of H2O and HT from steel and LiAlO? by temperature programmed desorption. J. Nucl. Mater. 179-181 (1991) 812.

[71] A.K. Fischer. Processes for desorption from LiAlO2 treated with hydrogen as studied by temperature prosrrammed desorption. Fusion Technol. 19 (1991) 1012- 1017.

[72] S. Tanaka.Y.Yamaki and M. Yamawaki Effects of surface heterogenetiy on tritium release. Proceedings 2nd International Workshop on Ceramic Breeder Blanket Interactions. Paris. France. September 22-24. 1993.

[73] K. Munakata and M. Nishikawa. Isotope exchange reactions on lithium oxide. J. Nucl. Mater. 170(1990) 187.

[74] S. Tanaka. S. Suzuki. K. Yamaguchi, and M. Yamawaki. Hydrogen-tritium exchange on the surface of ceramic breeders. Advances in Ceramics, Vol. 27 (American Ceramic Society. Westerville. OH. 1990) pp. 345-360.

[75] M. Briec. J.J.Abassin. CE. Johnson. M. Masson, N. Roux, and H. Watanabe, The MOZART experiment: in-pile tritium extraction from lithium oxide, aluminate. zirconate. Proc. 15th SOFT. Utrecht, (1988) p. 1105.

[76] T. Kurasawa. O.D. Slagle, G.W. Hollenberg. and R.A.Verrall. In-situ tritium recovery from L'PO irradiated in fast neutron flux - BEATRIX II initial results. Fusion Technol. 19(1991)931-937.

[77] H. Kwast. R. Conrad. R. May. S. Casadio. N. Roux, and H. Werle. The behavior of ceramic breeder materials with respect to tritium release and pellet/pebble mechanical integrity, Proc. ICFRM-6 September 27-October 1, 1993. Stresa. Italy.

[78] J.M. Miller and R.A. Verrall. Performance of a Li2ZrO3 sphere-pac assembly in the CRITIC II irradiation experiment, ibid. [79] R.A. Verrall. O.D. Slagle, G.W. Hollenberg. T. Kurasawa, and J.D. Sullivan. Irradiation of lithium zirconate pebble bed in BEATRIX II phase II, ibid.

180] W. Breitung and H. Werle. Experimental evidence for tritium release controlling processes in lithium silicates. J. Nucl. Mater. 179-181 (1991) 847.

[81] T. Terai, Y. Takahashi. and S. Tanaka. In-situ tritium release experiments from solid breeding materials (TTTEx) - Released chemical form of tritium and its diffusivities in Li2O, Fusion Engrg. and Des. 7 (1989) 345-351.

[82] S. Tanaka. A. Kawamoto. M. Yamawaki. T. Terai. Y. Takahashi, H. Kawamura. and M. Saito, In-situ tritium release experiments from solid breeding materials (TTTEx) - Tritium diffusion coefficient and surface reaction on lithium aluminare, Fusion Engrg. and Des. 8 (1989) 155-160.

[83] R.A. Verrall, J.M. Miller, LJ. Hastings, D.S. MacDonald. and D.H. Rose. CRITIC I: tritium release and post irradiation examination of large grain lithium oxide. Advances in Ceramics. VoI 27 (American Ceramic Society, Westerville, OH, 1990) pp. 299-316.

[84] M. Briec, J.P. Kopasz. S. Casadio, and H. Werle. In-pile tritium release from lithium ceramics and influences of surface processes. Proc. 16th SOFT, London, (1990) p. 857.

[85] T. Kurasawa and H. Watanabe. Influence of surface phenomena on in-situ tritium release from lithium oxide.J. Nucl. Mater. 179-181 (1991) 851.

[86] CE. Johnson. D.L. Baldwin, and .' P. Kopasz. Tritium release from beryllium discs and lithium ceramics irradiateu in the SIBELIUS experiment. Proc. ICFRM- 6, September 27 - October 1. '993. Stresa. Italy.

[87] Y. Gohar. H. Attaya. M. Billone et al.U.S breeder solid breeder blanket design for ITER. Fusion Tech. 19 (1991) 1538

[88] G.W. Hollenberg. H. Watanabe. LJ. Hastings. S.E. Berk. BEATRIX II : A multinational solid breeder materials experiment. J.Nucl.Mater. 191-194 (1992) p.23-29

[89] B. Rasneur. G. Thevenot. Y. Bouilloux. Irradiation behaviour of LiAlOT and Li2ZrO3 ceramics in the ALICE 3 experiment. J.Nucl.Mat. 191-194 (1992) p?243

[9Oj C. Alvani. S. Casadio. G. Cosoli. A. De Luca. G. Filacchioni and A.Miliozzi. Irradiation behaviour of LiAlO") breeder pellets. Proc. 17th SOFT .Rome, 1992 p. 1271

[91] G.W. Hollenberg. C. Alvani. B. Rasneur. H. Elbel. H. Wedemeyer, D.E. Walker. T. Takahashi. K. Noda. H. Akiyama. The FUBR-IB experiment in BEATRIX-I design and materials characterization. J.Nucl. Mat. 155-157 (1988) 563-567.

[92[ O.D. Slagie, F.D. Hobbs. D.L. Baldwin, t. Takahashi. K. Noda. and R.A. Verrall. Post-irradiation examination of BEATRIX-II, phase I. Proceedincs ICFRM.6

[93] CE. Johnson, I. Johnson, L.A. Curtis and P.E. Blackburn. Thermodynamic behavior of the Li2O-H2O (g)-LiOH (g) system. Proceedings Int. Workshop on Ceramic breeder blanket interactions Tokyo (1992) p. 48 [94] .L. Ethridge and D.E. Baker. Effects of fast neutron irradiation on thermal conductivity of L12O and LiAlCb. Advances in Ceramics Vol. 25 p. 165

[95] K. Noda, Radiation damage and irradiation effects in solid breeders. J. Nucl. Mater. 179-181(1991)37/

[961 K. Noda. Y. Isgii. H. Matsui. and H. Watanabe. Radiât. Eff. 97 (1986) 297.

[97] Y. Asaoka. H. Moriyama. K Iwasaki. K. Moritani, and Y. Ito. J. Nucl. Mater. 183(1991)174

[98] M.H. Auvray-Gely. A. Perez, and A. Dunlop. Electron paramagnetic resonance and optical absorption studies of irradiated lithium aluminate. Philos. Mag. B57 (1988) 137.

[99] H. Moriyama. T. Nagae. K. Moritani. and Y. Ho. A comparison of production behavior of irradiation defects in lithium ceramics, Proc 17th SOFT, Rome, 1992 p.1435.

[100] N.M. Masaki. K. Noda. H. Watanabe. R.G. Clemmer. and G. W. Hollenberg, J. Nucl. Mater, in press.

[101] K. Noda, Y. Ishii.H. Matsui. H. Ohno. and H. Watanabe, Electrical conductivity of Li2O under and after irradiationJ. Nucl. Mater. 179-181 (1991) 835.

[1021 K. Noda, Y. Ishii. T. Nakawaza. H. Matsui. D. Vollath. and H. Watanabe. Model experiments for tritium behavior in pure and Al-doped lithium orthosilicate by using ion conductivity measurements. Fusion Engrg. and Des. 17 (1991) 55.

[103] E.T. Cheng, K.R. Schultz. and C.P.C. Wong, Activation evaluation of fusion solid breeder materials. Proc. Symp. on Fusion Engineering 1989 pp. 1116-1119.

[104] GJ. Butterworth. Activation characteristics and waste management options for some candidate tritium breeders. J. Nucl. Mater. 184 (1991) 197-211.

[105] K. Tsuchiya. K. Tatenuma. H. Sagawa. M. Kainose. and H. Kawamura. Reprocessing technology development of used ceramic breeders. Proceedings 2nd International Workshop on Ceramic Breeder Blanket Interactions, September 22- 24,1993. Paris. France.

[106] J.H. Miles and GJ. Butterworth. Management options for used lithium ceramic breeder materials. J. Nucl. Mater. 191-194(1992) 199-203.