CHANGE OF b-DIMENSION WITH SWELLING OF MONTMORILLONITE

-The effect of swelling on the b-dimension of Na-saturated, Upton, Wyoming montmorillonite was re-examined using powdered silicon as an internal standard. After being corrected with reference to the internal standard, the b-dimension did not increase through the entire range of swelling as reported previously. It increased up to a water content of ~ 3.0 g/g montmorillonite and remained constant thereafter. However, the diffraction peak of the internal standard shifted towards lower angles at higher water contents. This shift was attributed to a relaxation of the wetter samples away from the knife edge of the diffractometer. Presumably, a similar relaxation occurred in the earlier study and was responsible for the apparent increase in b-dimension at water contents above ~ 3.0 g/g. The b-dimension of Upton montmorillonite saturated with different cations was determined using powdered silicon and a colloidal quartz impurity as internal standards. The water content at maximal swelling decreased as the b-dimension increased. I N T R O D U C T I O N Low, Ravina and White (1970) and Ravina and Low (1972) reported that the b-dimension of Na-montmorillonite increases continuously as it swells from the air-dried to the maximally swollen state. However, questions have arisen about the validity of their X-ray analysis and so the present study was undertaken. In this study, the b-dimension of each montmorillonite was determined using colloidal quartz or powdered silicon as an internal standard. Thus, the results could be corrected for any geometrical diffraetometer aberrations that might occur. MATERIALS AND M E T H O D S The < 2 #m fraction of Upton, Wyoming montmorillonite was saturated with different cations by batch washing, initially in 1 N chloride solutions of the cations and finally in deionized water, by the method of centrifugation and decantation. The maximal swelling of each of the resulting homoionic montmorillonites was then determined in triplicate by placing a one-g sample of the dry montmorillonite on the wet, sintered glass filter of a Gooch crucible, setting the crucible in a tray of deionized water and allowing the montmorillonite to absorb the water. To prevent evaporation, the crucible was covered by a plastic sheet with a perforation in it. When equilibrium was attained, as indicated by constancy in the weight of the cruciblel the water content of the swollen montmorillonite was determined gravimetrically. For X-ray powder diffraction analysis, an air-dried or swollen sample of homoionic montmorillonite was 201 confined in the "window" of a powder mount by a glass film on one side and a Mylar film on the other. Before sealing the Mylar film in place with rubber cement, the surface of the sample was made level with the flat surface of the mount by means of a razor blade. Then, after bringing this surface into coincidence with the knife-edge of a Phillips X-ray diffractometer, arranged for symmetrical reflection, the sample was scanned automatically at a speed of 1/8 ~ (20)/min using Ni-filtered Cu Ks radiation. The slit system consisted of a 1 ~ divergence slit, a 1 ~ receiving slit and a 0.1 nun scatter slit. Each sample contained natural colloidal quartz as an impurity. Initially, it was used as an internal standard to correct the results for any geometrical diffractometer aberrations. For this purpose, its diffraction peak was assumed to be at 60 ~ (20), the location of the corresponding peak for pure crystalline quartz. Later, spectrographically standardized, powdered (< 140 mesh) silicon was used as an internal standard. Small amounts of it were mixed thoroughly with the sample, either in the air-dried or swollen state, and diffractograms of the mixtures were obtained as before. The diffraction peak of the silicon was assigned a value of 56.12 ~ (20). When the colloidal quartz was used as an internal standard, duplicate diffractograms were obtained on 5-10 samples of each montmorillonite. When the powdered silicon was used instead, triplicate diffractograms were obtained on three samples of each montmorillonite. After locating the resulting (060) diffraction peak of the montmorillonite with reference to the corresponding diffraction peak of the internal standard, the respective 202 [SRAELA RAVINA and PHILIP F. Low b-dimension was calculated. The average b-dimension is believed to be accurate to within +0.001 A. RESULTS AND DISCUSSION As noted by Norrish (1954), there are two ranges of clay swelling, namely: (1) the range of water adsorption between 0 and 0.5 g water/g clay in which the interlayer spacing varies, often in stepwise fashion, from ~ 9.5 ~ to ~ 20 ~ and (2) the range of water adsorption above 0.5 g water/g clay in which the interlayer spacing varies continuously from --~40A upwards. Evidently, interlayer spacings between 20A and 40 h are forbidden. The first range, which is referred to as the range of crystalline swelling, has been studied extensively. The second range has received relatively little attention. Our study is concerned largely with the latter range. The water content of the different homoionic montmorillonites at maximal swelling is reported in Table 1. Note that the exchangeable cations have a profound effect on the water content. This effect has been observed previously by Falconer and Mattson (1933) and Winterkorn and Bayer (1934). The b-dimension of the air-dried montmorillonites was determined initially with the colloidal quartz as internal standard. However, because of the possibility that the lattice dimensions of the colloidal quartz might change with the adsorption of water, the Table 1. The water content at maximal swelling of Upton montmorillonite saturated with different cations Water water content content Cation (g H20/g clay) Cation (g HzO/g clay) Li + 23.3 Be + + 2.8 Na + 24.4 Mg + + 2.6 K + 9.3 Ca + + 3.8 Rb + 2.0 Sr + + 2.5 Cs + 0.8 Ba + + 2.4 b-dimension was determined again with powdered silicon as the internal standard. The results are reported in Table 2. The data in Table 2 indicate that the values obtained for the b-dimension depend on the internal standard. Further, the data indicate that the b-dimension is related to the radius of the exchangeable cation when the cation is monovalent. This relation is illustrated graphically in Figure 1. A similar relation was found by Leonard and Weed (1967) for dioctahedral vermiculite dried at 350~ Hence, it appears that the exchangeable cation perturbs the underlying crystal structure and that the degree of perturbation depends on ionic size. In keeping with the concepts of Radoslovich and Norrish (1962) and Lahav and Bresler (1973), we assume that the oxygen triads rotate until half the oxygens coordinate with the interlayer cations. As mentioned earlier, Table 2 shows that the values obtained for the b-dimension with colloidal quartz as internal standard are different from those obtained with powdered silicon as internal standard. Figure 1 shows that the difference becomes smaller as the crystallographic radius of the exchangeable cation decreases. This difference would remain constant if it were due entirely to an error in locating the diffraction peak of one (or both) of the internal standards on an absolute scale. Therefore, either the lattice dimensions of the colloidal quartz changed with water content as affected by the exchangeable cation (Table 2); or the two sets of data were obtained under different conditions, e.g. under different relative humidities and, hence, at different water contents as affected by the exchangeable cation. To accomplish the primary objective of the present study, we measured the b-dimension of the Na-saturated, Upton montmorillonite at different water contents using powdered silicon as an internal standard. The results, corrected with reference to this standard, are given in Table 3. They show that the b-dimension of the montmorillonite increases with water content up to a water content of 1.0-3.0 g/g and remains conTable 2. The water content and b-dimension of air-dried, Upton montmorillonite saturated with different cations b-dimension Crystallographic Water quartz as silicon as Exchangeable radius of cation t content int. stnd. int. stnd. cation (A) (g/g clay) (A) (A) Li § 0.78 8.966 Na + 0.98 0.0165 8.969 8.968 K + 1.33 0.0059 8.971 8.973 Rb + 1.49 8.973 8.977 Cs + 1.65 0.0032 8.976 8.984 Be + + 0.34 8.960 Mg + + 0.78 8.960 Ca + + 1.06 0.0641 8.960 8.973 Sr + + 1.27 0.0589 8.960 Ba + + 1.43 0.0414 8.960 t Data from Stillwell (1938). Effect of swelling on montmorillonite 203