Dice Question Streamline Icon: https://streamlinehq.com

Dependence of excess entropy of mixing on Al–Si order in alkali feldspar

Determine whether the excess entropy of mixing s^{ex} in the NaAlSi3O8–KAlSi3O8 alkali feldspar solid solution increases with increasing Al–Si order of the tetrahedral framework, by characterizing s^{ex} as a function of potassium site fraction X_K and the degree of Al–Si order across ordered and disordered states, to resolve the ambiguity indicated by conflicting experimental reports.

Information Square Streamline Icon: https://streamlinehq.com

Background

The paper evaluates mixing thermodynamics of the NaAlSi3O8–KAlSi3O8 (alkali feldspar) solid solution using a neural network potential, with explicit consideration of three Al–Si framework ordering states (fully ordered, T1-disordered, and fully disordered). From simulations at 1073.15 K, the authors find that the excess entropy of mixing s{ex} increases with increasing Al–Si order, whereas experimental literature presents conflicting results.

Specifically, some datasets (e.g., low albite–microcline) suggest higher s{ex} for ordered compositions, while others (e.g., adularia with intermediate disorder) report very high s{ex} that contrasts with measurements on disordered specimens. This inconsistency leads the authors to state that it is not clear from experiments whether an increase of s{ex} with increasing Al–Si order should be expected. Clarifying this dependence is important for accurately modeling the solvus and mixing behavior of alkali feldspar and for interpreting thermodynamic data across different disorder states.

References

It is not clear from the experiments, whether this should be expected as s{\mathrm{ex}} of the low albite-microcline of Delbove is certainly higher than the s{\mathrm{ex}} for the various disordered specimen but the 1991 data of Hovis and coworkers on adularia (intermediate disorder) reaches very high s{\mathrm{ex}} which contrasts with Heuser.

Thermodynamics of alkali feldspar solid solutions with varying Al-Si order: atomistic simulations using a neural network potential (2407.17452 - Gorfer et al., 24 Jul 2024) in Results and Discussion — Miscibility and non-ideality (Figure 5c)