Method: Cavities preparations (3mm wide, 5mm long and 2mm depth) on the buccal surface of 80 bovine incisors were made. A conventional (Ketac Molar Easymix - KM) and a resin-modified (Vitremer - VT) GIC were used. The cavities were divided into 8 groups according to substrate and superficial treatment. Groups: 1 - KM ND without UA, 2 - KM CA without UA; 3 - KM ND with UA, 4- KM CA with UA, 5 - VT ND without UA, 6 - VT CA without UA, 7 - VT ND with UA and 8 - VT CA with UA. The induction of caries-like lesions was performed by the microbiological method using Streptococcus mutans. The UA on the GIC was held for a period of 30s. The specimens were stored for 30 days in water, after this period the specimens were sectioned mesio-distal direction. Knoop hardness readings (25Kg, 15s) were performed in 5 linear points until depth of 200 µm from material/substrate interface. Three sequences of measurements in the pulpal wall of the specimen were made, one in the central region of the cavity and the other two 1000 mm in each side from the central measure. Data were statistic analyzed by ANOVA and of Tukey's test (α = 0.05).
Result: Noncarious dentin presented better results when compared to that caries affected dentin. Furthermore the ultrasound application resulted in higher hardness values for both materials and for both surfaces studied.
Conclusion: It was concluded that the ultrasound application increased the hardness of dentin and to encourage remineralization of caries affected dentin.