Phosphates in suspensions of alkaline, basaltic soils.

Abstract

The phosphate solid-solution activities in the A and A₁ horizons of ten alkaline soils from basalt were calculated in order to assess their conformity to solid-solution behavior. Nine soils were from the Cape Verde Islands and one from Arizona. Four of the Cape Verdean soils belonged to the same series (Ponta) but presented different levels of P fertilization. The other soils were not fertilized. All soils behaved in conformity with solid-solution theory. Three assumptions were made: (1) Surfaces of minerals are similar because of weathering; (2) Partial equilibrium was reached; (3) The mole fraction of the total solid composition is similar to the X of the surface of the solid. A preliminary test was conducted for optimization of experimental conditions, which turned out to be: 24-25°C; 1:5 soil solution ratio; 0.01M CaCl₂ extracting solutions. The relation of pH vs. time was used as the indicator of change and suspensions were kept for 21 days until pH stabilization. After filtration, electrical conductivity, soluble Ca, Mg and P were measured. A second experiment was done on selected samples, in order to study the influence of added Ca on the Solid Activity Coefficient (SAC). SAC and ionic solid activity coefficients for phosphates were computed from the data and the equation IAP = g X Ksp. Ksp of different Ca phosphates were taken as reference. Best results were found with octocalcium phosphate and bobierrite as reference for calcium and magnesium phosphates respectively. The curves of log SAC vs. P(added) were linear, indicating confirmation of solid solution behavior. Fertilized and virgin soils presented different slopes for those curves. Based on octocalcium phosphate, the basaltic soils yielded phosphate ionic solubility coefficients on the range 1 to 4 which are very low compared with similar coefficients calculated in the literature for non basaltic soils. The amorphous nature of basaltic minerals was considered as an explanation for the solid-solution behavior observed. Results show reasonable conditions for magnesium phosphate formation and no influence of added Ca on the linearity of log SAC vs. P(added) plots. However Ca levels corresponding to I.S. around 0.028 produced a considerable drop on the values of SAC when bobierrite was taken as reference.