The Sorption of the Rare Earth Element, Nd, Onto Kaolinite at 25°C

Abstract

The question of the fractionation of rare earth metals (REE) between clay minerals and coexisting aqueous solutions is one of interest to the geochemical community. Research interest in the marine geochemistry of the REE, for instance, is partly driven by the need to better understand REE metal budgets in seawater and/or the geochemical cycle of REE. This, in turn, is linked to the proposition that models of hydrothermal fluxes associated with spreading ridges must consider the roles of secondary minerals as sinks for the REE, especially in view of the fact that smectite and chlorites are important products of low-temperature alteration of mid-oceanic ridge basalts (Chamley 1989). Clay minerals are also common products of hydrothermal alteration, weathering and diagenesis, and petrographic evidence indicates that some REE redistribution occurs under these low-temperature conditions (McLennan 1989 and references therein). Indeed, some petrographic evidence shows that REE remohilization may occur during diagenesis and that adsorption by clay minerals plays important roles in such fractionation processes (Awwiner and Mack 1991; Zhao et al. 1992). A limited number of experimental studies have reported coefficients for the partition of REE between various minerals and coexisting aqueous solutions. Koeppenkastrop and De Carlo (1992) studied the sorption of REE, in the presence of seawater (pH = 7.8), onto the following synthetic phases: vernadite (~MnO~), hydroxylapatite [CaI0(PO4)(OH)2], amorphous goethite and crystalline goethite. The fractionation trends observed by them included the preferential uptake of light rare earth element (LREE) relative to heavy rare earth element (HREE), the existence of a positive Ce anomaly for sorption onto vernadite and a relatively enhanced uptake of Nd by hydroxylapatite. According to Koeppenkastrop and De Carlo (1992), this enhanced uptake of Nd by hydroxylapatite and the positive anomaly for Ce sorption on vernadite suggest that the mechanism of sorption in the former was due to substitution for Ca 2§ and to oxidative scavenging of Ce 3+ in the latter. Nonetheless, mechanistic models of REE sorption have not yet evolved and the question of the relevant adsorption model (isotherm) on these phases is not resolved. Beall et al. (1979) measured distribution coefficients for the sorption of La, Sm, Yb, Am, Cm and Cf on kaolinite, montmorillonite and attapulgite at (room temperature); the experiments were conducted in NaC1 solutions (0.25-4 M) buffered to a pH of 5 with acetate solutions. Their study suggested that metal uptake decreased with ionic strength, although the extent to which this might be due to the effects of competing metal complexation reactions was not addressed. At each ionic strength and for a given clay mineral, the values of the distribution ratios measured for both the lanthanides and actinides were only marginally different. Amongst the minerals, however, the actinides demonstrated a strong preference for attapulgite. Prel iminary experimental investigations (Mecherri et al. 1990) into the sorption of Nd onto orthoclase and calcite at 50 ~ have also been reported. According to them, Nd sorption onto orthoclase is best modeled by a Langmuir-type isotherm whereas a Freundlich-type isotherm characterized the sorption onto calcite. In addition, Mecherri et al. (1990) observed that only a small fraction of the sorbed metals could be desorbed and that increasing pH tends to decrease the partition of Nd onto the solid phases. The focus of the studies discussed above has understandably been the determination of the distribution coefficients for various lanthanides. Besides the obvious petrologic utility of partition coefficients, they also have important practical applications. Such data are