Calcium-Magnesium Exchange Selectivity of Wyoming Montmorillonite in Chloride, Sulfate and Perchlorate Solutions

The need to consider CaCl and MgCl complexes in Ca-Mg exchange reactions on montmorillonite was evaluated by examining cation selectivity in various anion backgrounds. The Ca-Mg exchange isotherms on SVVy-1 Wyoming bentonite were determined in 0.0125 M SO4, 0.025, 0.050, and 0.100 M Cl and 0.03 M perchlorate background. Calculated selectivity values for the reaction Ca + MgX^ Mg + CaA" were Kv = 1.15, 1.11 and 1.10 for the SO,,, Cl and perchlorate media, respectively. There was no trend in K, with exchangeable Ca with any anion or with ionic strength in Cl media. The Ca preference in SO4 and perchlorate media indicates that montmorillonite preference for Ca is not an artifact of complexation in Cl media. These A", values are also consistent with earlier published Ca-Mg selectivity of the same clay in perchlorate media. The CaCI and MgCl complexes comprised an insignificant amount of the exchangeable cations. P of the effect of irrigation water on soil solution composition is dependent on the cation exchange process and thus the selectivity that a soil complex has for a particular ion. The exchange reactions involving Na-Ca-Mg have been studied extensively in bulk soils and soil clays as well as in relatively pure specimen clays. From all these investigations it is apparent that the exchange reactions characterized for specimen clays cannot yet be used to predict exchange behavior in soil clays. Solution chemistry models thus require that soil cation-exchange constants be measured for the individual soils and under the chemical conditions to be simulated. The site specificity of exchange constants is one of the obstacles to generalization of predictive soil water models. Poor prediction of cation exchange constants in soils is due to several factors. Clays may not behave as simple mixtures, as evidenced by the inability to predict exchange constants of a mixture from those of the pure clays (Krishnamoorthy and Overstreet, 1950). Organic matter and oxides also have exchange capacities and their exchange constants are usually much different from those of clays. Samples of the same clay mineral from different locations also have different exchange constant values, due in part, to variation in structural composition The possibility that ion pairs are important species on the exchange complex has been discussed by several researchers (Bower and Truog, 1940; Babcock et al., 1959). Recently, Sposito et al. (1983a,b), concluded that CaCl and MgCl are important in montmorillonite exchange in Cl media. They noted that almost all exchange experiments with specimen clays D. L. Suarez and M. F. Zahow, U.S. Salinity Lab., USDA-ARS, 4500 Glenwood Dr., Riverside, CA 92501. M. F. Zahow is from Soil Salinity and Alkalinity Laboratory, Baccos, Alexandria, Egypt. We gratefully acknowledge funding provided M. F. Zahow by the Cooperative Arid-Lands Agric. Res. Program supported by U.S.AID. Contract NEB-0170-AOO-0047-00. Received 4 Jan. 1988. "Corresponding author. Published in Soil Sci. Soc. Am. J. 53:52-57 (1989). have been conducted in Cl background solutions. Sposito et al. (1983a), estimated that for Na-Ca exchange in their Cl media experiments, up to half of the exchangeable Ca was in the form CaCl and that for Na-Mg exchange up to about 30% of the exchangeable Mg was in the form MgCl. If these ion pairs are adsorbed, then Ca-Mg exchange in a Cl background is actually a quarternary system (or even a quinternary system if H competition is considered). Sposito et al. (1983b), further hypothesized that the reported preference of montmorillonite for Ca over Mg was due to the use of Cl. The increase in adsorbed charge with increasing exchangeable Ca or Mg (when the counter ion was Cl), was considered an artifact of attributing all adsorbed Ca to the species Ca. This hypothesis was supported by the equal preference of montmorillonite for Ca and Mg in the presence of a perchlorate (C1O4) background (which is assumed not to form complexes), reported by Sposito et al. (1983b). The present study was initiated to examine Mg-Ca exchange in SO4 and C1O4 (perchlorate) media and Mg-Ca exchange in Cl media under different ionic strengths. Our objective was to evaluate the importance of Ca and Mg chloride complexes in Ca and Mg exchange reactions and the applicability of selectivity values determined with one anion to other anion systems. If the preference for Ca over Mg in the Cl system is due to the preference of CaCl over MgCl then Ca and Mg should have equal preference in the SO4 and C1O4 media. Additionally, Ca preference in Cl media should increase with increasing ionic strength as the percent Ca complexed and the activity ratio CaCl/ Ca increases. These predictions have important implications for saline soils and especially for the reclamation of sodic soils by using Ca salts. MATERIALS AND METHODS The Na-montmorillonite, SWy-1, (Crook County, WY) was obtained from the Clay Mineral Society Source Clay Repository. This SWy-1 bentonite was also used by Sposito et al. (1983a,b) in their studies of Ca-Mg, Na-Ca and NaMg exchange. In 0.25 L tared bottles, 32.0 g of clay were reacted with 0.200 L of 1 MNaCl. The bottles were shaken for 15 min then centrifuged for 10 min at 3400 X g and 25 °C. The supernatant was decanted and the procedure repeated. Each bottle was subsequently reacted with 0.05 M NaCl and treated in a similar manner. The clay was next mixed with 20 L of 0.02 MNaCl and left standing for 16 h. The <2-jtm size fraction was decanted, centrifuged and dried for 36 h at 60 °C. A subsample of the prepared clay was Mg saturated and equilibrated with glycerol at 30% humidity. A prepared oriented slide gave x-ray diffraction basal reflections in agreement with published data for Wyoming montmorillonite (Brown, 1972). We detected no other minerals except a trace of quartz in our prepared SWy-1 clay. Montmorillonite standards were spiked with 0 to 10% vermiculite by weight, Mg saturated and equilibrated with glycerol at 30% humidity. Based on the estimated detection limit of 0.5% we consider that SWy-1 contains < 0.5% vermiculite.