Clays and Clay Minerals, Vol. 47, No. 5, 555-566, 1999. DETERMINATION OF ILLITE-SMECTITE STRUCTURES USING MULTISPECIMEN X-RAY DIFFRACTION PROFILE FITTING BORIS A. SAKHAROV, HOLGER LINDGREEN, 1 ALFRED SALYN, AND VICTOR m. DRITS Institute of Geology, Russian Academy of Sciences, Pyzhevsky per D.7., 109017 Moscow, Russia Clay Mineralogical Laboratory, Geological Survey of Denmark and Greenland, Thoravej 8, DK2400 Copenhagen NV, Denmark Abstraet--A procedure for structural investigations by X-ray diffraction of mixed-layer structures incor- porating swelling layers has been developed. For each sample, specimens saturated with different cations (Na, Mg, and Ca), are analyzed both as air-dried and as glycolated. One structural model fitting all the observed patterns then provides the structure of the sample. Samples tested include: Illite-smectite (I-S) minerals from Kazachstan (a rectorite), Dolna Ves in Slovakia, Kinnekulle in Sweden, the North Sea, and Scania in Sweden. The fitting of the patterns of the Kazachstan rectorite demonstrated that the instrumental parameters applied in the modeling were correct. For the I-S minerals from Slovakia and Kinnekulle the observed patterns were fitted with one two-component I-S model. However, the Ca-saturated and air-dried specimen of the Kinnekulle bentonites had two types of swelling interlayers. For the Slovakian I-S with Reichweite = 2, an alternative two-phase I-S plus I-V (V = vermiculite) model fitted the experimental X-ray diffraction patterns equally well. The I-S mineral from Scania is in fact a three-component I-T-S (T = tobelite) and the North Sea sample is a four-component I-S-V-V', one type of the swelling layers having swelling characteristics intermediately between smectite and vermiculite. In addition to layer types and distribution, interlayer compositions, such as the amount of interlayer glycol and water and of fixed and exchangeable cations, were determined. Key Words---lllite-Smectite, Simulation, Structure, Swelling, X-Ray Diffraction. INTRODUCTION Determination of the structure of mixed-layer min- erals containing illite and swelling layers is important because these are common minerals and diagenesis and weathering changes their structure (Shutov et al., 1969a, 1969b; Perry and Hower, 1970; Weaver and Beck, 1971; Hower et al., 1976). By X-ray diffraction (XRD), the interstratification of illite-smectite (I-S) minerals is usually estimated from peak-migration curves showing the position of basal reflections versus the proportion and mode of interstratification of layer types in the mixed-layer structure (Drits and Sakharov, 1976; Srodofi, 1980, 1981, 1984; Watanabe, 1981, 1988; Reynolds, 1980, 1988; Tomita et al., 1988; Moore and Reynolds, 1989; Drits et al., 1994). The peak-migration technique can, however, only be used for two-component I-S with random (R = 0, where R is the Reichweite parameter) or maximum ordering for R = 1, 2, or 3, but not for I-S with segregated I or S layers or with intermediate degrees of ordering. Fur- thermore, peak-migration curves have sofar mainly been used for glycolated I-S and are usually based on the assumption that all smectite interlayers contain two glycol layers and that their swelling properties do not depend on the exchangeable cation. In addition, the mica layers are usually assumed to be K-bearing and the thickness to be 9.98 or 10 ,~. The most effective technique for determination of the structural parameters of mixed-layer minerals is based on comparison between experimental XRD Copyright 9 1999, The Clay Minerals Society curves and curves calculated for structural models having different proportions and distributions of illite and smectite layers (Reynolds and Hower, 1970; Drits and Sakharov, 1976; Reynolds, 1980; Moore and Reynolds, 1989; Drits and Tchoubar, 1990). However, the simulation of XRD patterns requires many struc- tural and instrumental parameters (e.g., types, struc- tures, compositions, thicknesses, and distribution of layers; mean thickness and thickness distribution of coherent scattering domains (CSD); particle orienta- tion; divergence of initial and diffracted XRD beams) which are known only approximately. For example, the swelling of smectite interlayers depends on the ex- changeable cations, the intercalated organic com- pounds, and the charge of the 2:1 layers, and these parameters require extensive work to determine (Brin- dley, 1966; MacEwan and Wilson, 1980; Drits et al., 1997a). The effect of these numerous variables is that several structural models may seem to fit the experi- mental data equally well, especially if the experimen- tal and calculated patterns are not superposed. The goal of this work is to demonstrate that deter- mination of reliable and detailed structural models for mixed-layer minerals can be obtained by the multi- specimen profile fitting procedure: First, for the pattern of each specimen, close agreement between positions, intensities, and profiles of the reflections in the cal- culated and experimental diffractograms must be ob- tained. Second, for one and the same sample, calcu- lated XRD patterns for one structural model must fit 555