^254: J. A. GRAY AND M. D. FRANCIS 



At present, the onl\' mechanism bv which all the different effects 

 observed experimentallv can be satislactorih explained (Gray et al., 

 1962 ) is the precipitation of some of the reaction products as phases 

 that are different from the original hydrox\ apatite. These phases 

 form on the h\ droxxapatite surface and limit further reaction or, at 

 equilibrium, completely prevent access of hydrogen ions to the en- 

 amel or egress of calcium and phosphate ions to the bulk solution 

 phase. The existence of these phases has been proxed by equilibrium 

 solubility measurements, as will now be described. 



Detection of the surface phases as described in the previous para- 

 graph by currently available physical methods is very difficult be- 

 cause, in the first place, the la\'ers formed are onh' a few angstrom 

 units thick (approximateh^ 25 A), on the basis of calculations from 

 chemical and phvsical data. Secondlv, isolation of the solid samples 

 is at present impracticable because any change of the equilibrium 

 solution in contact with the surface will tend to change the phase. 

 If, however, the original postulation is correct, then the solution in 

 equilibrium with the surface coating should have a composition 

 satisfving the solubilitv product for the phase existing on the hy- 

 droxyapatite surface. These measurements of solubilitv product 

 were made in simple acidic solutions (i.e., not decalcification me- 

 diums). Brieflv, the experimental work was as follows (Francis, in 

 preparation). Samples of dental enamel (human and bovine, both 

 intact and powdered) and s\'nthetic hvdroxvapatite (Victor Chem- 

 ical Works) as well as dicalcium phosphate were agitated in solu- 

 tions of lactic or acetic acid buffers in a pH range from 4.6 to 6.5. 

 The slurries were filtered or centrifuged to remove the solids, and 

 the solution pH, calcium concentration, and phosphate concentra- 

 tion were measured. Determinations as a function of exposure time 

 were made to confirm that equilibrium conditions had been attained. 

 Corrections for complexing of calcium bv lactate and acetate anion 

 species were required, but ionic strength corrections could be ig- 

 nored. Several ratios of solution to solid, and addition of calcium 

 (as CaCL.) and phosphate (as NaHL>P04) initiallv to the acid solu- 

 tion, were also studied. Regardless of conditions, the Ksi- consistentlv 

 observed was that of dicalcium phosphate (Van Wazer, 1958), as 

 summarized in Table III. This proved that tlie solution was in equi- 



