Methods: Symmetrical dentin pairs were prepared from 40 human freshly extracted molars. The experimental (eroded) samples were subjected to six demineralization and remineralization cycles/day and the control (non-eroded) samples were kept in a remineralization solution for six days. All samples were brushed for 15 seconds, two times/day. Four different bonding systems were analyzed: Group 1 Optibond FL, Group 2 Excite, Group 3 One Coat SE Bond and Group 4 G-Bond. All restorations were made with Tetric Evo Ceram. The samples were stored in a dark humidity chamber at 370C for 24 hours and then thermally cycled 1000 times. The shear bond strengths were measured with a crosshead speed of 1mm/min. The fracture analysis was performed using SEM. The data were evaluated statistically with and without pre-test failures (PTF). Wilcoxon Signed Ranks test, Mann-Whitney U test, Pearson Chi-Square test, Fisher's Exact t-test and Kappa tests were used for statistical analysis (p=0.05).
Results: The mean bond strength values for eroded groups were 15.47± 11.54 and 18.20± 10.31, for non-eroded groups 18.48± 12.58 and 21.12± 11.14 MPa with and without PTF, respectively. The non-eroded Group 1 and 2 exhibited higher bond strengths than 3 and 4. No statistically significant differences were found among the eroded groups. Among failure types, no statistically significant differences were found between the eroded and non-eroded groups.
Conclusion: It was concluded that erosion influenced bond strength of composite restorations to dentin.