Monazite and xenotime solubility in granitic melts and the origin of the lanthanide tetrad effect

The solubility of synthetic, pure rare earth phosphates with monazite or xenotime structure (LaPO4 to LuPO4) in hydrous haplogranitic melts was measured at 2 kbar and 800-1,100 degrees C. Experiments were run for up to 2 months to attain equilibrium. Monazite and xenotime solubility decreases with increasing phosphorus concentration in the melt. Published equations for monazite solubility in felsic melts, which do not explicitly include phosphorus concentration in the melt, should therefore be treated with caution. The effect of phosphorus can be quantitatively modeled if one assumes that monazite partially dissolves as ionic and molecular species in the melt (REE3+ and REEPO4). Equilibrium constants for the dissolution reactions as well as quantitative data on speciation were derived from the solubility data. Monazite and xenotime solubility strongly increases with the peralkalinity of the melt. This effect is mostly due to an increase in the solubility of the ionic species, which are probably stabilized by non-bridging oxygen atoms in the melt. In peraluminous melts, the solubility of monazite and xenotime is nearly constant. Fluorine has no major effect on monazite and xenotime solubility; in fact, the solubility appears to slightly decrease with increasing fluorine content. The solubility of rare earth phosphates is not a simple continuous function of atomic number or ionic radius. Rather, the solubility shows a tetrad-like pattern with several local maxima of solubility at individual rare earth elements. The solubilities of neighboring rare earth elements sometimes differ by more than a factor of two; these effects are far outside any analytical error. The tetrad pattern is particularly clearly seen in some of the peralkaline melts and in the fluorinerich metaluminous melts. Some features, however, such as a solubility maximum at ytterbium, are seen in virtually all melts. The lanthanide tetrad effect in some highly evolved granites may therefore be a result of monazite and xenotime fractionation. The solubility of monazite and xenotime in silicate melt probably shows the tetrad effect, because of the very unusual coordination of the rare earth elements in these phosphate minerals, which is different from the coordination of the melt and therefore causes different crystal field interactions with the partially filled f orbitals of the rare earths. The tetrad effect in granites cannot be used as an indicator of fluid/rock or fluid/melt interaction, since it can be experimentally reproduced in the absence of any fluids.