Solvent-induced dimensional changes in EDTA-demineralized dentin matrix

The purpose of this study was to test the null hypothesis that the re-expansion of dried matrix and the shrinkage of moist, demineralized dentin is not influenced by polar solvents. Dentin disks were prepared from midcoronal dentin of extracted human third molars. After complete demineralization in 0.5M of EDTA (pH 7), the specimens were placed in the well of a device that measures changes in matrix height in real time. Dry, collapsed matrices were created by blowing dry N(2) on the specimens until they shrank to a stable plateau. Polar solvents [water, methanol, ethanol, n-propanol, n-butanol, formamide, ethylene glycol, hydroxyethyl methacrylate (HEMA), or mixtures of water-HEMA] as model primers then were added and the degree of re-expansion measured. These same solvents also were applied to moist, expanded matrices and the solvent-induced shrinkages measured. Regression analysis was used to test the correlations between matrix height and Hansen's dispersive, polar, hydrogen bonding, and total solubility parameters (delta(d), delta(p), delta(h), delta(t)). The results indicate that water-free polar solvents of low hydrogen bonding (H-bond) ability (e.g., neat HEMA) do not re-expand dried matrices and that they shrink moist matrices. When HEMA was mixed with progressively higher water concentrations, the model water-HEMA primers expanded the dried matrix in proportion to their water concentrations and they produced less shrinkage of moist matrices. Solvents with higher H-bonding capacities (methanol, ethanol, ethylene glycol, formamide, and water) re-expanded the dried matrix in proportion to their solubility parameters for H-bonding (delta(h)). They also induced small transient shrinkages of moist matrices, which slowly re-expanded. The results require rejection of the null hypothesis.