Synthesis and sintering of Li2SnO3

Synthesis and sintering of Li2SnO3

224 Journal of Nuclear Materials 160 (1988) 224-228 North-Holland, Amsterdam SYNTHESIS M. INAGAKI, Materials AND SINTERING OF Li,SnO, S. NAKAI an...

1MB Sizes 15 Downloads 94 Views


Journal of Nuclear Materials 160 (1988) 224-228 North-Holland, Amsterdam



OF Li,SnO,


Science, Toyohashi University of Techdog,


Toyohashi 440, Japan

Received 22 February 1988; accepted 28 July 1988

As one of candidates for solid blanket materials of fusion reactors, lithium stannate Li,SnO, was synthesized by precipitation of a complex, lithium hexahydroxystannate Li,Sn(OH),, and its thermal decomposition. In order to get single-phase precipitates Li,Sn(OH),, aqueous solution SnCl, had to be added slowly to LiOH aqueous solution. Decomposition of Li,Sn(OH), to Li,SnO, proceeded in two steps, abrupt release of about two molecules of water around 200° C and gradual release of remaining water above 260 o C. The grains of Li,SnO, obtained at 800 o C were porous, consisting of small primary particles of about 0.2 pm size. For the pellets prepared from these porous grains under 20 and 100 MPa, the density saturated to a value of 1.8 and 2.4 g/cm3 within two days at lOOO”C, corresponding to the porosity of 64 and 52&, respectively. The porous nature of the original Li,SnO, grains was found to be kept even after 1000 o C treatment for a week. For the pellet prepared by ground Li,Sn03 under 200 MPa, remarkable sintering was observed at 1000 o C; density increased rapidly and became close to theoretical density (7% porosity) after four days. The product Li,SnO, synthesized by the present process has advantages in terms of manufacturing, handling and possibly tritium recovery of the blanket.

1. Introduction Solid blankets of fusion reactors have been designed to consist of ceramics of polycrystalline lithium compounds. The inventory of tritium formed by nuclear reactions between lithium atoms and neutrons has been pointed out to be very important, even to govern the efficiency of the reactor. Tritium recovery from a solid blanket has been discussed from the following unit steps [l]; (1) bulk diffusion in grains of lithium compound, (2) desorption at grain edges, (3) diffusion along grain boundaries and permeation through pores to edges of sintered body, (4) desorption at the edges of sintered bodies, (5) diffusion and percolation into a stream of purging gas, and (6) convective mass transfer to the stream. The processes (l), (2) and (3) are concerned with the size, crystallinity and morphology of grains of blanket ceramics, and the processes (3) and (5) are mostly influenced by the density of the sintered ceramic bodies. So far, various lithium compounds, such as Li,O, LiAlO,, Li,ZrO,, Li,SiO,, etc., have been tested by different authors [2,3]. Some compounds, particularly Li,O, are highly hygroscopic, highly reactive against structural materials and easy to sinter at 800 o C which

is an expected temperature to keep ceramic blankets in the reactor [4]. In order to solve these problems, different double oxides, such as LiAlO,, have been tested, though the concentration of lithium in the blanket reduces. In the course of preparation of these double oxides by solid state reaction between lithium carbonate and respective metal oxide, erosive lithium compounds are often formed and also their sublimation has to be taken into account. In order to avoid these difficulties, different manufacturing processes have been applied for different compounds, such as co-precipitation, hydrolysis of metal-alkoxides, sprey-drying, etc. [5]. Inagaki et al. [6] have reported that the various crystalline stannates MSnO, (M = Mg, Co, Zn, Mn, Ca, Sr, Ba) are obtained at relatively low temperatures by thermal decomposition of MSn(OH), which are precipitated from aqueous solutions, and that most of the starmates obtained are in fine powder and hard to sinter even at so high temperatures as 1000 o C. In the present work, synthesis of lithium stannate Li,SnO, by precipitation and following thermal decomposition of hydroxystannate, and also its sintering behavior at high temperatures were studied by expecting an application to blanket materials for fusion reactors. Formation process of Li,SnO, from Li,Sn(OH), was

0022-3115/88/$03.50 0 Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)

M. Inagaki et al. / Synthesis and sintering of Li,SnO,


compared with the reaction process of the solid state reaction between Li ,CO, and SnO,.

2. Experimental and results 2.1. Synthesis of Li,Sn(OH), Li,Sn(OH), complex was prepared by the addition of 1M SnCl, aqueous solution to 2M LiOH aqueous solution under strong stirring; spontaneous precipitation of Sn(OH), occurred in the beginning, but it dissolved quickly under stirring. After the solution being kept at room temperature, the hydroxide complex Li,Sn(OH), started to precipitate. In order to get single-phase precipitates Li,Sn(OH),, avoiding precipitation of Sn(OH), at low pH value of the solution, the solution of SnCl, had to be added slowly (about 0.4 cc/min, for example) and stopped before the pH value of the solution went down to 11.5. The precipitates obtained were filtered, washed repeatedly with water and then dried at 80°C. All these processes were carried out in air at room temperature and all chemicals used were reagent grade. The molar ratio of tin ions added to lithium ions had to be less than 2. If the ratio became > 2, Sn(OH), precipitated together with Li,Sn(OH),, and consequently the mixing of Li,SnO, and SnO, was resulted after thermal decomposition. The yield of complex hydroxide Li,Sn(OH), to tin ions added was about 60%. In fig. la, a scanning electron micrograph of the precipitates Li,Sn(OH), is shown. The grains of the precipitates are small pillars with an hexagonal crosssection. An X-ray powder pattern of Li,Sn(OH), obtained is shown in fig. 2a, though it is not indexed yet.


40 20 , CuKm

Fig. 2. X-ray powder patterns. (a) Li,Sn(OH),

Fig. 1. Scanning electron micrographs. (a) Li,Sn(OH), cipitated, (b) Li,SnO, prepared from Li,Sn(OH), treatment at 800 o C for 4 h.

as-preby heat

2.2. Thermal decomposition to Li,SnO, DTA and TG curves of the precipitates Li,Sn(OH), obtained are shown in fig. 3a. Step-wise weight de-

60 as-precipitated, (b) Li,SnO, for 4 h.

prepared from Li,Sn(OH),

by heat treatment at 800 o C

M. Inagaki et al. / Synthesis and sintering of Li,SnO,










Lb 0






Temperature (‘C )

Temperature ( ‘C 1 Fig. 3. DTA and TG curves. (a) Li,Sn(OH), as-precipitated,(b) a mixture of Li,CO, and GO,.

creases are observed; up to 120 o C, over a range from 120 to 260°C, and over a wide range of 260-700°C, respectively. Two sharp endothermic peaks corresponding to the first two weightlosses are also detected. The first weightloss and endothermic peak seem to be due to the release of absorbed water. The percentage of weightloss and shape of endothermic peak, which looked like combining successive two peaks, strongly depended on the sample precipitates. The second and third weightlosses of about 15 and 7.78, respectively, are due to the release of water molecules by the decomposition of Li,Sn(OH),. Thus the thermal decomposition of hydroxide proceeds probably in two steps; the first step at around 200°C is the release of about two water molecules (theoretical weightloss of 15 wtW) and the second step in a wide temperature range of 260 to 700 o C is that of one molecule by slow diffusion through the first-step product, supposedly Li2SnOz(OH), (theoretical weightloss of 7.7 wt%). The product obtained after thermal analysis above 800 o C was always single-phase of Li, SnO,, but that heated up to 240 o C showed only a very broad band in X-ray powder pattern. After the heat treatment at 240°C for one week, powder pattern of the product roughly coincided with that of Li,SnO,, though all the peaks were rather broad. By the heat treatment at 400 o C for 4 days, this temperature being at the middle of the second step of decomposition of Li,Sn(OH), as shown in fig. 3a, well-crystallized single-phase Li,SnO, was obtained. Therefore, the water release in the second step of decomposition of Li,Sn(OH), is supposed to be governed by slow diffusion of water molecules through disordered structure of the decomposition products of the first step, Li,SnO,(OH),. Theoretically, Li,SnO, can be obtained from hydroxide Li,Sn(OH), at a temperature as low as 240” C

after long time heating, but this temperature is not enough to reconstruct the structure to crystalline state. Practically, therefore, the temperatures between 400 and 650°C were found to be effective to prepare the wellcrystallized Li,SnO, powder. These temperatures are much lower than those at which the solid state reaction between Li,O and SnO, occurs, as will be shown below. By using the present process, we can avoid the erosion of the container by reactive lithium compounds because of direct transformation of double hydroxide to oxide. In fig. 3b, DTA and TG curves on a mixture of Li,CO, and SnO, are shown for comparision. The first weightloss and corresponding broad endothermic peak at about 70-220 o C is probably due to the release of absorbed water from Li,CO,. Above 220 o C gradual weight decrease is observed up to 850° C and rather sharp endothermic peak around 700 o C. These changes seem to be due to the decomposition of Li,CO, to Li,O and its successive reaction with SnO, to form Li,SnO,. The product after thermal analysis up to 10CO°C was Li,SnO, with small amounts of SnO,. Appreciable erosion of platinum crucible by reactive lithium compounds was observed. In this process, therefore, loss of lithium component either by sublimation and reaction with platinum could not be avoided. 2.3. Morphology and sintering of Li,SnO, X-ray powder pattern of Li,SnO, obtained after 800° C-treatment for 4 h is shown in fig 2b. This pattern coincides completely with the published one in JCPDS files [7], which is indexed on the basis of a monoclinic unit cell with the following lattice parameters; a = 5.308 A, b = 9.185 A, c = 10.022 A, p = 100.28”.

h4. Inagaki et al. / Synthesis and sintering ofU,SnO,

Scanning electron micrograph of LizSnO,



(fig. lb) shows that the oxide grains are aggregates of


. 01234567 Heat Treatment


Fig. 4. Changes of bulk density with heating time at 1000 o C. (a) pellet compressed under 20 MPa, (b) pellet compressed under 100 MPa, (c) pellet prepared from ground powder under 200 MPa.

primary particles with very fine homogeneous size (about 0.2 pm) incorporating small pores (also about 0.2 pm) among primary particles and that the outline of oxide grains seems to keep the appearance and size of the original hydroxide hexagonal pillars (fig. la). The crystalline powder of Li,SnO, was compressed under the pressures of 20, 100 and 200 MPa into pellets with a diameter of 13 mm and the thickness of about 1.5 mm. The pellets thus prepared were sintered at 1000 o C in air. Fig. 4 shows the changes of bulk density of these pellets with heating time at 1000°C. For the pellets compressed under 20 and 100 MPa, the bulk density tends to saturate to the values of 1.8 and 2.4 g/cm3 (36 and 48% of theoretical density, respectively) within two days. These pellets do not show any further densification even after seven days at 1000°C. For the pellet prepared by compressing the ground Li,SnO, powder under 200 MPa, remarkable sintering is ob-

Fig. 5. Sxnning electron micrographs of fractured surfaces of the pellets heat-treated at 1000° C for 7 days. (a) and (b) pellet compressed under 20 MPa, (c) and (d) pellet prepared from ground powder under the compression of 200 MPa.


M. Inagaki et al. / Synthesis and sintering

served; the bulk density increases rapidly with increase in heating time and becomes close to theoretical density (7% porosity) after four days. Scanning electron micrographs of fractured surfaces of these pellets heat-treated at 1000 o C for seven days are shown in fig. 5. For the pellet compressed under 20 MPa, the size of primary particles increases, but their surfaces are round and small pores remain between particles, the characteristic appearance of the starting grains (fig. la) being kept, even after 1000 QC for seven days (figs. 5a and b). This observation agrees with low bulk density of this pellet (the curve a in fig. 4). For the pellet compressed under 200 MPa, on the other hand, well-defined crystal faces are observed on each particles, though their size is not different from those in the pellet under 20 MPa, and very few pores are observed between particles (figs. 5c and d). This observation also coincides well with rapid sintering and high’bulk density of this pellet (the curve c in fig. 4).

3. Discussion Lithium stannate ceramics prepared by the present process seems to have the following advantages in terms of manufacturing and handling; easy preparation of raw complex hydroxides, formation at a temperature as low as 400 o C, no formation of erosive lithium oxides during synthesis, (4 very fine particles of about 0.2 pm, (e) porous grains, (0 not very much hygroscopic, (9) possibility to control of bulk density, C-4no proceeding of sintering at high temperatures. The points (a)-(c) can be advantageous for preparation of Li,SnO, powders. Formation of LizSnO, at low temperatures can avoid sublimation of lithium oxide and consequently avoid segregation of second phases, such as SnO,. Easy preparation of Li,Sn(OH), and no formation of erosive substances during its decomposition can simplify the procedure and equipment for manufacturing. The point (f) is also advantageous for handling the powders and for preparation of blanket body. The points (d), (e), (g) and (h) can be advantageous for tritium recovery from the blanket. In relation to unit processes for tritium recovery, as mentioned in the first section, fine particles and porous grains of

of Li,SnO,

Li,SnO, make bulk diffusion in grains, desorption at their edges and diffusion along grain boundaries easy. Sintering of blanket materials under irradiation has been pointed out to be troublesome for tritium recovery. Generally speaking, lithium compounds are easy to sinter, but stannates have been known to be hard to sinter. The present Li,SnO, was found to have low sinterability even at 1000 o C, higher than the expecting temperature for the blanket of fusion reactors. Sintering at the beginning of heating, however, was found to be controlled by grinding time and compressing pressure of raw powders. Therefore, the sintered body with desired bulk density is able to be prepared and it does not sinter any more during kept at a high temperature for long period. This point might be advantageous for design and construction of the blanket. On the material Li,SnO,, unfortunately, fundamental properties, such as thermal conductivity, thermal expansion coefficient, and also behavior under irradiation, have not been understood well. In spite of this lack of fundamental science on Li,SnO,, it is worth while to study this material as one of candidates for blanket materials of fusion reactors.

Acknowledgments This work is supported by Grant-in-Aid for Fusion Research (No. 62055015) from the Ministry of Education, Science and Culture of Japan. The authors would like to express their thanks to Prof. Y. Takahashi of Tokyo University for his encouragement.

References VI S. Tanaka, Kakuyugou Kenkyu 56 (1986) 318. 121 T. Terai, S. Tanaka and Y. Takahashi, Fusion Technol. 8 (1985) 2143. 131 G.W. Hollenberg and D.L. Baldwin, J. Nucl. Mater. 133 t 134 (1985) 242. 141 G.W. Hollenberg, J. Nucl. Mater. 122 & 123 (1984) 8962. 151 For examples: D. Vollath, H. Wedemeyer and E. Guenther, J. Nucl. Mater. 133 & 134 (1985) 221; S. Hirano, T. Hayashi and T. Kageyama, J. Am. Ceram. Sot. 70 (1987) 171. 161 M. Inagaki, T. Kuroishi, Y. Yamashita and M. Urata, Z. Anorg. Allg. Chem. 527 (1985) 193. 171JCPDS files No. 31-761.