The effect of phosphate on the solubility of calcium carbonate and of bicarbonate on the solubility of calcium and magnesium phosphates.

In 1925, Holt, La Mer, and Chown (1) reported that the apparent solubility product [Ca++13[P04=J2 was greater in a solution of inorganic salts approximating the composition of serum than it was in solutions of sodium chloride of the same ionic strength. Very shortly thereafter, Sendroy and Hastings (2) recalculated values for pK’, Ca,(PO& from the data of Holt, La Mer, and Chown and also reported extensive data of their own. According to Sendroy and Hastings, the solubility of calcium phosphate was the same in “serum salt” solutions as in sodium chloride solutians and was between those values calculated from the data of Holt, La Mer, and Chown for the solubility in sodium chloride and in “serum salt” solutions. Sendroy and Hastings suggested (p. 821) that the latter difference might be due to the “use of what may be uncertain values for the first or the third stoichiometric dissociation constant of phosphoric acid at such low pcllH+,” but they seem to have made no attempt to calculate whether or not any conceivable error in these dissociation constants could have accounted for the observed lOO-fold change in the apparent solubility product. We have made such calculations and do not find that any conceivable error in any, or all, of the values for the dissociation constants of phosphoric acid could produce any significantly different effect upon the ratio of POh’ to total phosphorus at pH 5.2 or 7.7, which represent the extreme values in the two sets of experiments by Holt, La Mer, and Chown. The “serum salt” solutions employed by Holt, La Mer, and Chown contained, besides calcium hydroxide and phosphoric acid, sodium, potassium, and magnesium chlorides and sodium bicarbonate. We have since reported evidence for the existence, in solution, of undissociated calcium and magnesium carbonate, bicarbonate, and phosphate (3, 4). However, the amounts of such complexes that could have been present could account for only a small part of the observed greater solubility in “serum salt” compared with that in sodium chloride solutions. The composition of the “serum salt” solutions used by Sendroy and Hastings is not given. They do not seem to have contained magnesium.