Subsolidus phase relations in the nepheline-kalsilite system at 0.5, 2.0, and 5.0 kbar

This paper presents: (1) X-ray diffraction data and refined unit-cell parameters for (Na,K)₃K(AlSiO₄)₄ nephelines and kalsilites synthesized at 0.5 kbar; (2) experimental data on the nepheline-kalsilite solvus in the Na₃K(AlSiO₄)₄−K₄(AlSiO₄)₄ system from 400 to 1000°C at 0.5, 2.0, and 5.0 kbar; (3) thermodynamic mixing-parameter equations for (Na,K)₃K(AlSiO₄)₄ crystalline solutions; and (4) calculated nepheline-kalsilite solvi and subsolidus activity-composition relations for the Na₃K(AlSiO₄)₄−K₄(AlSiO₄)₄ system at pressures in the range 0.5–5.0 kbar.

X-ray diffraction data and refined unit-cell parameters for (Na,K)₃K(AlSiO₄)₄ nephelines and kalsilites crystallized at 0.5 kbar indicate that d_20.1 spacings and unit-cell parameters for the synthetic nephelines, as well as d_10.2 spacings and unit-cell parameters for the synthetic kalsilites, are a linear function of XK₄(AlSiO₄)₄. However, the a unit-cell dimensions and unitcell volumes for the synthetic nephelines are systematically smaller than those for nephelines crystallized at one atm by Smith and Tuttle (1957) and Donnay et al. (1959). The discrepancies are attributed to slightly Na₂O-deficient starting materials and additional Na₂O volatilization during experimentation in these two earlier investigations.

Three different types of solid starting materials were used to delimit the nepheline-kalsilite solvus at ∼100°C intervals from 400 to 1000°C at 0.5, 2.0, and 5.0 kbar. At a given pressure and temperature the three starting materials yielded very similar results indicating that equilibrium was closely approached in all experiments. Collectively the experimental data suggest that nepheline-kalsilite solvi are slightly asymmetric toward Na₃K(AlSiO₄)₄. Between 400 and 800°C there is good agreement between our solvus data obtained at 0.5 kbar and the solvus data obtained by Tuttle and Smith (1958) at one atm, 480 bars, and 981 bars, but there are significant discrepancies at higher temperatures.

Comparisons between our solvus data and calculated solvi indicate that values of Margules, van Laar, and quasichemical mixing parameters for (Na,K)₃K(AlSiO₄)₄ crystalline solutions are a linear function of temperature but not of pressure. Relative activities for the Na₃K(AlSiO₄)₄ and K₄(AlSiO₄)₄ components of the crystalline solutions calculated from the Margules and van Laar solution models are nearly identical and significantly different from activities calculated from the quasichemical formulation.