Alkali diffusion and exsolution in alkali feldspar

In the precursor project we studied microstructure evolution during spinodal decomposition of alkali feldspar in experiment and theoretically. With the successor project we want to extend this study by (i) addressing nucleation- growth as an alternative reaction pathway for exsolution and (ii) investigating in detail Na+/K+ interdiffusion in alkali feldspar. (i) To induce exsolution by nucleation and growth, isothermal anneals will be done at conditions where the homogeneous alkali feldspar is metastable. Experiments will be done at high pressure in the 1 to 2 GPa range, where the metastable region in the alkali feldspar binary extends to high temperature, and the overall reaction kinetics is fast. The exsolution microstructure and microchemistry will be investigated using electron beam microanalytical techniques with high spatial resolution including FEG-SEM, FEG-EMP and TEM combined with FIB sample preparation. The numerical model that was developed in the precursor project will be complemented with a nucleation module to extract process parameters from comparing modelled and observed microstructures and phase compositions. (ii) From experiments in the precursor project we find that the geometries of concentration profiles produced from Na+/K+ exchange between alkali feldspar and NaCl-KCl salt melt deviate from what is expected for simple interdiffusion. Exchange fronts propagating through the crystal show a sharpness, which depends on the composition difference between the pristine and the exchanged domains of the grain but does not change with time. We hypothesize that this may be due to strain occurring at the interface between exchanged and pristine crystal due to lattice mismatch (coherence-strain). The mechanisms by which this occurs are, however, elusive. Solving the question of coherence strain effects on cation interdiffusion is fundamental for the understanding of interdiffusion and exsolution in mineral systems. We suggest doing cation exchange experiments using different starting materials including isotopically labelled KCl with well-defined geometry. The microchemistry, defect density and lattice distortion at the exchange fronts will be investigated using high resolution analytical techniques, and conceptual models for the influence of coherency strain on Na+/K+ interdiffusion in alkali feldspar will be developed and parameterized.

The main achievement of the project was the experimental calibration of Na-K interdiffusion in potassium-rich alkali feldspar including diffusion anisotropy and the composition dependence of the interdiffusion coefficient. To this end Na-K cation exchange experiments between a NaCl-KCl salt melt and gem-quality alkali feldspar machined to plates with polished faces in specific crystallographic orientations were made in the temperature range of 800 to 1000 degrees Centigrade at ambient pressure. Diffusion profiles were analyzed in six different crystallographic directions, and the full diffusivity tensor was determined, allowing for calculation of the interdiffusion coefficient for any direction. At intermediate compositions with atomic Na over (Na+K) ratios of 0.70 to 0.95 Na-K interdiffusion is only weakly composition dependent. In contrast, at more potassium-rich compositions the Na-K interdiffusion coefficient increases with increasing K content. With this calibration the accuracy of diffusion modelling of alkali feldspar is substantially improved and kinetic calculations involving alkali feldspar are more accurate than with previous calibrations. Due to the strong composition dependence of the lattice parameters of alkali feldspar, any composition heterogeneity associated with diffusion produces elastic stress in the crystal, which eventually may induce fracturing. In samples with well-defined geometry, regular fracture patterns develop, where the characteristic spacing between cracks is related to the strength of the compositional perturbation. This effect was used to extract the so-called stress intensity factor, which describes the brittle behaviour of the material. If fracturing does not occur the elastic strain and stress associated with a composition perturbation is stored in the lattice. It was shown for the first time that this can be made visible using electron diffraction techniques. In comparative studies chemically induced fracturing and the influence of stored elastic strain on diffraction patterns was also investigated in calcite aggregates.