EXPERIMENTAL-STUDY OF ANORTHITE DISSOLUTION AND THE RELATIVE MECHANISM OF FELDSPAR HYDROLYSIS

Steady-state dissolution rates of anorthite (Ang,) were measured as a function of aqueous Si, Al, and Ca concentration at temperatures from 45 to 95 degrees C and over the pH range 2.4 to 3.2 using a Ti mixed-f low reactor. All dissolution experiments exhibited stoichiometric diss olution. The concentration of aqueous Si, Al, and Ca ranged from simil ar to 7 x 10(-5) to similar to 1 x 10(-3) molal, similar to 6 x 10(-5) to similar to 3.4 x 10(-3) molal, and similar to 5 x 10(-5) to simila r to 0.1 molal, respectively, corresponding to calculated anorthite ch emical affinities ranging from similar to 115 to similar to 65 kJ/mol. Measured anorthite dissolution rates at constant temperature are prop ortional to a(H+)(1.5), where a(H+) designates the activity of the hyd rogen ion, and consistent with an apparent activation energy of 18.4 k J/mol. Anorthite dissolution rates are independent of aqueous Al conce ntration, which is in contrast with the alkali feldspars, whose consta nt pH, far from equilibrium rates are proportional to a(Al+3)(-0.33) ( Oelkers et al., 1994; Gautier et al., 1994; E. H. Oelkers and J. Schot t, unpubl. data). This difference suggests a distinctly different diss olution mechanism. For the case of both types of feldspars it appears that Al is more readily removed than Si from the aluminosilicate frame work. Because it has a Si/Al ratio of 3, the removal of Al from the al kali feldspar framework leaves partially linked Si tetrehedra. Removal of Si still requires the breaking of Si-O bonds, and thus the overall alkali feldspar dissolution rate is controlled by the decomposition o f a silica-rich surface precursor. The variation of alkali feldspar di ssolution rates with aqueous Al activity stems from the fact that the formation of this precursor requires the removal of Al. In contrast, b ecause it has a Si/Al ratio of 1, the removal of Al from the anorthite framework leaves completely detached Si tetrehedra. As a result, the removal of Si does not require the breaking of Si-O bonds, the rate co ntrolling precursor complex is not formed by the removal of Al, and th e overall dissolution rate is independent of aqueous Al concentration at far from equilibrium conditions. It can be inferred from these resu lts that the variation of far from equilibrium aluminosilicate dissolu tion rates on aqueous Al depends on the number and relative strength o f different bond types that need to be broken for mineral hydrolysis.