[OLD] Gibbsite and Diaspore solubility calculations

[webmaster]

From: James Laurinat

Subject: Gibbsite and Diaspore solubility calculations

I have been using REACT, with the default thermo.dat database, to calculate gibbsite and diaspore solubilities in strong basic solutions (NaOH = 1 to 15 molal). The calculated solubilities do not closely match measured solubilities reported by A.S. Russell, J.D. Edwards, and C.S. Taylor, "Solubility and Density of Hydrated Aluminas in NaOH Solutions," J. of Metals, p.1123-1128 (October, 1955). Moreover, when I use a modified database to input log K's to match calculated and measured solubilities, the resulting log K's are not constant but vary with composition. These problems make it difficult to use GWB to predict aluminum solubilities. Does anyone have suggestions as to why I am having these difficulties? My input to REACT consists of:

Temperature (deg C)

Na+ (molal)

Al+++ (molal)

balance on H+

suppress all (mineral formation)

Is the problem that I am trying to use the database beyond its intended pH or ionic strength range? Should I specify the solution's ionic strength using either the simax or timax command? I have been using the default of 3 molal. The manual states that extrapolation beyond this value is dangerous. Also, does anyone know the specific data source on which the thermo.dat gibbsite and diaspore solubility correlations are based?