The crystal chemistry of hydrotalcite-like compounds is investigated by powder diffraction methods such as Rietveld structure refinement, radial distribution function analysis, atom-specific radial distribution analysis obtained from anomalous diffraction data, and EXAFS spectroscopy. The topology of the brucite-type layer as well as the layer stacking arrangement and the intralayer bonding are determined for Mg,Al, Mg,Ga, and Ni,Al systems. The occurrence of long-range cation ordering in materials with M(II)/M(III) cation ratios close to 2.0 is observed only for systems with similar cation radii, and it is associated with a corrugation of the octahedral layer. The lack of ordering for systems with highly different cation radii is ascribed to the layer compression exhibited by these compounds, which prevents the layer distortion. The stacking arrangement is random for the solids investigated, except for the Mg,Al system which shows a preference for the rhombohedral polytype. It is proposed that this behavior is related to the extent of local directional bonds between the oppositely charged layers.

The structure of the phases obtained upon dehydration and decomposition of hydrotalcite-like compounds is investigated by several experimental techniques. A reaction mechanism is proposed encompassing a change in coordination of the M(III) cations during the dehydration step. The formation of a 3-dimensional structure occurs upon the subsequent decomposition of the interlayer anions and dehydroxylation of the octahedral layers. In the decomposed material the cations are trapped in the interstices of a regular oxygen cubic close packed lattice and exhibit a considerable disorder. Strains develop during the decomposition, which are likely related to the observed increase of surface area. The thermal stability of the decomposed materials is connected to the reduced cation diffusivity in the oxygen lattice.

When dehydroxylation of kaolinite powder is carried out in the usual way, the linear relations anticipated for first-order kinetics and for the Arrhenius plot of log k versus 1/T are satisfied only very approximately. Factors relating to the form of the specimen, (shape, size, compaction, container, etc.) are shown to be very important. A method is developed for obtaining data for a specimen in the form of an infinitely thin disc. The first-order kinetic relation and the Arrhenius relation are then linear, and the latter gives an activation energy of 65 K cal./mol. The dehydroxylation process is shown by x-ray analysis to proceed crystal by crystal and this leads to an interpretation of the first-order kinetics. The x-ray method is used to study the distribution of reacted and unreacted material throughout a disc of material. Although isothermal conditions are employed, large differences are found between the interior and exterior of a partially dehydroxylated disc. These effects ~e attributed to the influence of a water vapor atmosphere within the heated disc. (1) INTRODUCTION The loss of weight when kaolinite is fired to temperatures exceeding about 450~ under normal atmospheric conditions is commonly ascribed to "dehy-dration" and the water involved in the reaction is designated "structural water." Neither term is correct, as the crystal lattice loses hydroxyl groups. The process is better described as "dehydroxylation" and it can be represented chemically by the equation OH+OH=H2OI' +O The mechanism is most probably one of proton migration so that if two protons momentarily find themselves associated with the same oxygen ion, there is a probability that a water molecule will be formed and will detach itself from the lattice. The present work was undertaken in the hope that a kinetic study of the process accompanied by x-ray examination would provide more detailed information about the process than is currently available. The crystal structure of kaolinite (Brindley and Robinson, 1946; Brindley and Nakahira, 1956) is so wet| known that no detailed description is required here. It suffices to recall that the four (OH) groups of the structural unit AI2Si2Os(OH)4 are located wholly in the octahedral sheet of the structure, three of them occupying positions on the outside of the layer structure and one within the structure. A previous study of the dehydroxylation of kaolinite 1 Contribution no. 56-35 from the College

No abstract

Abstract. The present work is a follow-up of the investigation on the decomposition reaction of kaolinite as a function of the defectivity of the starting material and the temperature of reaction. In the present work we study the high temperature reaction of mullite synthesis from kaolinite, from the starting point of the results obtained in the first part.Time resolved energy-dispersive powder diffraction patterns have been measured using synchrotron radiation in isothermal conditions. The apparent activation energy for mullite nucleation and growth is found to be related to the defective structure of the starting kaolinite, which thus must have an influence on the chemical homogeneity of the amorphous intermediate phase.The analysis of the kinetic data indicate that the initial reaction mechanism is controlled by mullite nucleation, while as the reaction proceeds it shifts towards a grain growth-limited process which is intermediate between phase boundary and diffusion controlled. The order of the reaction obtained from standard analysis of the isothermal kinetic data is lower in the case of the ordered kaolinite KGa-1, in agreement with a rate limiting process more strongly limited by diffusion.For each sample there is a small but significant decrease in the order of the reaction at higher temperature: we interpret the change as related to the variation of the diffusion process in the amorphous phase due to the growing grains of mullite and cristobalite.The values of the activation energies and induction times are comparable neither to a model of mullite formation from a monophasic gel, nor mullite formation from a diphasic gel, being intermediate between the two. We can infer that the amorphous precursors from natural kaolinites can be considered pseudo-monophasic geMike phases, approaching the monophasic gel-like behaviour as the defectivity of the initial kaolinite increases.