Chemistry for Sustainable Development 12 (2004) 379–383 379

Low-Temperature Synthesis of Highly Disperse Lithium Gamma-Monoaluminate

OLGA A. KHARLAMOVA1,2, RAISA P. MITROANOVA1, KONSTANTIN A. TARASOV1, LYUDMILA E. CHUPAKHINA1, VITALIY P. ISUPOV1, ANATOLIY S. ZYRYANOV3 , KONSTANTIN A. ALEKSANDROV3, NIKOLAY N. BATALOV3 and ZELIA R. KOZLOVA3

1Institute of Solid State Chemistry and Mechanochemistry, Siberian Branch of the Russian Academy of Sciences, Ul. Kutateladze 18, Novosibirsk 630128 (Russia) E-mail: [email protected] 2Novosibirsk State University, Ul. Pirogova 2, Novosibirsk 630090 (Russia) 3Institute of High-Temperature Electrochemistry, Ural Branch of the Russian Academy of Sciences, Ul. S. Kovalevskoy 22, Yekaterinburg 620219 (Russia)

Abstract

A new method for the synthesis of highly disperse lithium gamma-monoaluminate is suggested. Binary lithium-aluminum hydroxide in carbon ate form [LiAl2(OH)6]2CO3 ⋅ 3H2O is used for the precursor. Calcination o of a mixture of this hydroxide with lithium carbonate above 800 C forms γ-LiAlO2. The sequence of chemical transformations leading to γ-LiAlO2 has been studied. The first stage of the synthesis yields α-LiAlO2; when o heated to a temperature above 800 C, the latter is transformed into γ-LiAlO2. The specific surface of the aluminate product is 3.5 m2/g. Test experiments showed that the product may be used as a holder for fuel elements with a melted carbonate electrolyte.

INTRODUCTION and inorganic anions [2–4]. The gel is formed by interaction of aluminum alkoxides with Highly disperse lithium gamma-monoalumi- aqueous lithium chloride or nitrate. The gel nate γ-LiAlO2 has attracted attention as a prom- synthesized in this way is filtered, dried, and ising material for use as a holder in fuel ele- calcinated. The reaction temperature is low- ments with a melted carbonate electrolyte. The ered to 700 oC, but the method employs rather particles of the holder keep the melted elec- expensive organoaluminum compounds and trolyte within the matrix. This application re- forms a lot of liquid organic waste. Therefore, quires that γ-LiAlO2 should have dispersity of developing new methods for the synthesis of the order of 0.1 mm. highly disperse γ-LiAlO2 devoid of the above- Two methods are conventionally used for mentioned disadvantages has become a crucial the preparation of γ-LiAlO2. The first is the problem. ceramic technique based on caking a batch The approach to the synthesis of γ-LiAlO2 mixture of alumina and lithium carbonate [1]. used in the present work is a continuation of High-temperature synthesis and long reaction our previous research [5, 6]; for precursors it times are major disadvantages of this process. employs binary aluminum and lithium hydrox-

Thus for synthesis of γ-LiAlO2 from Al2O3 and ides [LiAl2(OH)6]nX ⋅ mH2O. The structure of o Li2CO3, reaction temperature is above 1000 C binary hydroxides consists of layers formed and reaction time is up to 10 h; the product from hydroxide ions arranged as 2D close pack- needs to be intermittently ground in the course ing with aluminum and lithium cations in the of the reaction. The second is the gel-sol method octahedral voids and with anions and crystalli- by which gamma-aluminate is obtained from zation water in the interlayer space. The struc- a gel containing aluminum and lithium cations ture of γ-LiAlO2 is an infinite 3D net formed 380 OLGA A. KHARLAMOVA et al.

by distorted LiO4 and AlO4 tetrahedra. In the to the batch mixture. The batch mixture was net, one can isolate chains of vertex-sharing thoroughly stirred, ground in an agate mor-

AlO4 tetrahedra [7]. Thus the structure of lith- tar, and heated in a SNOL muffle furnace in ium-aluminum binary hydroxides, as well as ambient conditions (heating rate ~6 oC/min). After the required temperature had been that of γ-LiAlO2, contains aluminum-oxygen chains with short distances between aluminum reached, the batch mixture was calcined for cations in the chain (8ig. 1). This facilitates struc- 2 h. An X-ray phase analysis (DRON-4 diffrac- tometer, CuK radiation, measurement rate ture formation of γ-LiAlO2 from binary hy- α droxide. Moreover, the precursor already con- 2 oC/min) was undertaken to examine the tains Al3+ and Li+ cations “mixed” at the mo- composition of the reaction products. 8urther- lecular level, which lowers the reaction tem- more, synthesis of the monoaluminate from a perature relative to the traditional ceramic batch mixture of lithium carbonate and binary procedure. hydroxide was studied in situ on a Siemens D5000 diffractometer with a thermostatted accessory, and it was also investigated by thermal analysis. EXPERIMENTAL Thermogravimetric analysis was performed on a Perkin–Elmer instrument (ambient conditions, 8or the precursor we employed the carbon- 15 mg samples, heating rate 6 oC/min). The area ate variety of binary hydroxide Al–Li of the specific surface of the synthesized pow-

[LiAl2(OH)6]2CO3 ⋅ 3H2O synthesized by the an- der was measured by the BET (Brunauer – Em- ion exchange method using the procedure of mett –Teller) method using argon desorption. The [8]. The precursor having an atomic ratio of morphology of the resulting particles of lithium lithium to aluminum of 0.5, for synthesizing monoaluminate was examined with a JSM-T20

γ-LiAlO2 we added lithium in carbonate form scanning electron microscope.

8ig. 1. Structural fragments of the [LiAl2(OH)6]2CO3 ⋅ 3H2O precursor and γ-LiAlO2. LOW-TEMPERATURE SYNTHESIS O HIGHLY DISPERSE LITHIUM GAMMA-MONOALUMINATE 381

RESULTS AND DISCUSSION

8igure 2 presents the thermogram of a mix- ture of binary hydroxide and lithium carbon- ate [LiAl2(OH)6]2CO3 ⋅ 3H2O + Li2CO3. As can be seen from the thermogram, the sample mass starts to change at 50–60 oC because of the loss of adsorbed and interlayer water. Mass loss corresponding to the first intense endothermal peak is ~20 %. As removal of three interlayer water molecules would lead to only 9–10 % mass loss, one can assume that the first peak is due not only to removal of interlayer water molecules, but also to decomposition of other structural fragments, namely, dehydration of metal hydroxide layers and decomposition of carbonate ions in [LiAl2(OH)6]2CO3 ⋅ 3H2O. The 8ig. 3. X-Ray diffractograms of the products of thermal possibility for metal hydroxide layers to be decomposition of a stoichiometric mixture of [LiAl2(OH)6]2CO3 ⋅ 3H2O and Li2CO3 obtained in situ: destroyed simultaneously with carbonate ions 1 – starting mixture; t, oC: 300 (2), 400 (3), 500 (4), during thermal decomposition of lithium-alu- 600 (5), 700 (6), 800 (7). Reflections: ** [LiAl2(OH)6]2CO3 ⋅ minum binary hydroxide in carbonate form 3H2O; * Li2CO3; n α-LiAlO2. was revealed in [6]. The second endothermal peak whose maximum lies at 230 oC may also The suggested scheme of structural transfor- be due to the concurrent removal of water mations is confirmed by analysis of X-ray dif- and decomposition of the carbonate ions and fractograms obtained in situ. metal hydroxide groups. 8urther heating leads When the mixture was heated to 300 oC to a plateau at 500 oC on the TG curve. At this (8ig. 3), the reflections of binary hydroxide point, mass loss is about 40 % and corresponds [LiAl2(OH)6]2CO3 ⋅ 3H2O completely vanished. to an almost complete removal of water and The reflections of lithium carbonate did not o carbonate ions from [LiAl2(OH)6]2CO3 ⋅ 3H2O. change in intensity at 300 and 400 C, but slightly 8urther decrease in mass in the temperature decreased at 500 oC; in the latter case, a weak o range 500–800 C may be caused by decompo- broadened reflection of α-LiAlÎ2 appeared in sition of the carbonate ion in lithium carbon- the region 2θ = 18–19o. When the temperature ate and formation of lithium monoaluminate. was elevated further, the intensity of the re-

flections of Li2CO3 decreased, while that of α-LiAlÎ2 increased. As can be seen from 8ig. 3, o γ-LiAlO2 did not form below 800 C because of the short calcination times of the batch mix- ture in situ. This compound formed when the calcination time was 2 h. Calcination of the batch mixture at 400 oC led to destruction of the binary hydroxide and formation of a phase practically amorphous to X-ray diffraction (8ig. 4). When the tempera- ture was raised to 700 oC, broadened reflec-

tions of α-LiAlO2 appeared. The reflections of o Li2ÑO3 did not change in intensity. At 800 C, the reflections of Li2ÑO3 vanished, and almost 8ig. 2. Thermogram of the stoichiometric mixture pure α-LiAlO2 formed. 8urther growth of tem- o of [LiAl2(OH)6]2CO3 ⋅ 3H2O and Li2CO3. Heating in air, perature to 900 C gave rise to weak reflec- heating rate 6 oC/min: 1 – TG, 2 – DTA. tions of γ-LiAlO2. 382 OLGA A. KHARLAMOVA et al.

8ig. 5. Measurements of volt-ampere characteristics of the fuel element at different operating times, h: 0 (1), 33.5 (2), 55.3 (3), 86.9 (4). 8ig. 4. X-Ray diffractograms of the products of thermal decomposition of a stoichiometric mixture of odes were preliminarily annealed and pre- [LiAl2(OH)6]2CO3⋅ 3H2O and Li2CO3 at different temperatures, °C: 400 (1), 700 (2), 800 (3), 900 (4), 950 (5). pared; then they were placed in metal separa- Reflections: * Li2CO3; • α-LiAlO2; ♦ γ-LiAlO2. tors – construction elements for dividing the gas flow. The matrix plate together with a plate 8inally, pure γ-monoaluminate (without an of a lithium-potassium eutectic mixture served α-form impurity) formed at 950 oC (see 8ig. 4). as a sealant between them. At an operating The reaction leading to γ-LiAlO may be 2 temperature (650 oC), the carbonate melted and schematically represented as follows: impregnated the porous matrix plate. The re- sulting electrolyte-matrix plate was hermeti- [LiAl2(OH)6]2CO3 ⋅ 3H2O + Li2CO3 cally sealed and gas-proof. → (Li O ⋅ 2Al O ) + Li CO <400 °C 2 2 3 r/àm 2 3 Service trials of fuel elements based on the synthesized matrix material indicated that the → 4α-LiAlO → 4γ-LiAlO 500–800 °C 2 >800 °C 2 latter may be used in electrolyte-matrix plates. According to electron microscopy data, the Thus at a voltage of 0.7 V, current density 2 particle size of the product was ~0.2–0.3 mm. was up to 150 mA/cm , and the characteristics The particles formed platelike aggregates sized remained at the same level for 200 h (8ig. 5). 5–15 mm, easily destroyed upon grinding to When the load was off, the emf also remained particles with dimensions of 2–3 mm. The constant throughout the whole period of trials. specific surface of this powder was 3.5 m2/g (BET data). CONCLUSION The synthesized lithium aluminate was used to prepare matrix plates by rolling. 8or this, Thus the given procedure is applicable to the aluminate powder was mixed with a solu- the synthesis of highly disperse lithium gam- tion of an organic binder (polyvinylbutyral) in ma-monoaluminate, which proved useful as a ethanol with a plasticizer addition. The result- material for fuel elements. The temperature of ing mixture was rolled on roll mills. Electrolyte thermal treatment of a batch mixture is lower matrix plates were prepared directly in fuel and the specific surface of lithium aluminate is elements after the binder had been burnt off higher than the corresponding characteristics of in the course of element triggering. the ceramic method. The procedure does not The units of the fuel element were tested require expensive organoaluminum reagents, as in two-element cells – batteries with parallel was the case with the sol-gel method. internal commutation of gases. Wetted hydro- gen was used for the anode gas, and a mixture Acknowledgement of oxygen and carbon dioxide (0.5O2 +CO2) was cathode gas. Porous electrode plates of nickel This work was supported by the Presidium of for anodes and of lithium-nickel oxide for cath- Russian Academy of Sciences grant No. 26 “Hydrogen LOW-TEMPERATURE SYNTHESIS O HIGHLY DISPERSE LITHIUM GAMMA-MONOALUMINATE 383 energetics” (project “New composition materials with 4 S. W. Kwon, S. B. Park, Ibid., 35, 8 (2000) 1973. lithium and proton conductance for solid state 5 V. P. Isupov, R. P. Mitrofanova, and L. E. Chupakhina, electrochemical devices”). Zhurn. prikl. khimii, 67, 11 (1994) 931. 6 V. P. Isupov, R. P. Mitrofanova, L. E. Chupakhina, REERENCES and A. Yu. Rogachev, Khimiya v interesakh ustoichivogo razvitiya, 4, 3 (1996) 215. 1 8. A. Hummel, J. Am. Ceram. Soc., 34 (1951) 235. 7 R. Dronskowski, Inorg. Chem., 32, 1 (1993). 2 S. W. Kwon and S. B. Park, J. Nucl. Mater., 24, 6 (1997) 131. 8 V. P. Isupov and L. E. Chupakhina, Zhurn. prikl. khimii, 3 S. W. Kwon, E. H. Kim, S. B. Park, J. Mat. Sci. Lett., 70, 3 (1997) 402. 18 (1999) 931.