METHODS: Class II butt joint and internal bevel cavities were prepared on the proximal surfaces of 120 caries free extracted human molar teeth. They were randomly divided into 3 groups of 40 each. Bonding intermediates Calibra and RelyX ARC were applied respectively to two of these groups, third group was left untreated. The 3 groups were then further divided into two groups of 20 each and restored with Dispersalloy and Oralloy Magicap S respectively. Ten specimens of each group were thermocycled. All teeth were stored in basic fuchsin (0,5%), for 8 hours, sectioned (two 1mm sections) and evaluated for dye penetration using a light microscope (40x magnification). Mean microleakage scores were analysed using the chi-squared test at a confidence level of 95%.
RESULTS: Microleakage of non-bonded amalgams was for both cavity preparations significantly higher (p < 0,0001) than that of the bonded amalgams. Microleakage of bonded amalgam restorations in the butt joint cavity was not significantly different from those in the internal bevel under various thermal stress conditions (p > 0,05). Sealing abilities of Calibra and RelyX ARC did not differ significantly (p > 0,05). Thermal stress did not affect the bond adversely in both cavity preparations (p > 0,05).
CONCLUSION: Within the limitations of this experiment it was concluded that, Class II cavities for bonded amalgams may be prepared with either a butt joint or an internal bevel, bonding of amalgam would reduce the leakage of high-copper amalgam restorations and thermal stress would not affect the bond adversely.