A physicochemical bench-scale caries model.

A diffusion cell comprising two compartments separated by a commerical membrane of known ion permselectivity was used as an experimental model to study factors which may affect caries formation. One compartment (the "lesion") contained an excess of hydroxyapatite or fluorapatite crystals, and its solution was kept near saturation by stirring. An unsaturated acidic calcium phosphate solution flowed continuously through the other compartment (the "plaque-saliva"), thus providing the driving force for dissolution of the crystals as modified by the permeability of the membrane and/or the presence of fluoride. Calcium, phosphate, fluoride, and chloride concentrations, pH, and membrane potential were measured at steady state. The permselective character of the membrane profoundly affected the solution in the "lesion": The initial Ca/P ratio of 1.6 was shifted, at steady state, to less than 1.1 ("neutral" membrane), to less than 0.48 (cation-permselective membrane), and to 3.3 (anion-permselective membrane). All the membranes caused Ca and PO4 concentrations in the "lesion" to increase (e.g., with the cation-permselective membrane, the Ca 22-fold and the PO4 75-fold). Substantial membrane potentials were encountered in all cases. The pH of the steady-state solution was always less than before diffusion was initiated; the average decrease was 1.1 units. In every case, the pH, and the Ca and PO4 concentrations of the F- -containing solution were less than those in the F- -free solution. The changes induced by the "neutral" membrane were unexpectedly large. In two cases, H3PO4 was diffusing against its chemical potential gradient at steady state. The results are relevant for understanding and for modeling of the caries process.