Surface milled by CAD-CAM system Vs laboratorial methods to simulate the milled surface: Effect on the resin bond strength to lithium disilicate glass-ceramic

Abstract

The aim of this study was to assess the resin microshear bond strength (μSBS) and the surface topography of lithium disilicate ceramic milled by CAD/CAM and to compare it with in-lab methods that simulate the topography of milled surfaces; thus, assessing the feasibility of those in-lab methods for in vitro microshear studies. The study was composed of 2 control groups (CAD/CAM group – discs were milled by a CAD/CAM system; POL – polished ceramics surface) and 4 testing groups (SiC – manual grinding simulation with #60 silicon carbide paper; WS – manual grinding with #60 wood-sandpaper; DB – grinding with a fine cylindrical diamond bur; MANDREL – grinding with a CAD/CAM bur coupled to a mandrel). After production, the ceramic samples were crystallized. Surface treatment with hydrofluoric acid (HF) and silane was performed on the bonding surface, then the starch matrices were filled by the resin cement and light-activated. The specimens (n= 45) were submitted to μSBS after 24 h of storage in water (37° C) and the failure pattern was analyzed. Topographic analysis, surface roughness, contact angle, and fractal dimension were performed. One-way ANOVA and Tukey's post-hoc tests were run for roughness, while Bonferroni’ post hoc was used for μSBS data. Fractal dimension data was qualitatively assessed. No difference was found in the μSBS between in-lab simulation methods and the milled group, while the polished group presented the lowest μSBS values. The CAD/CAM milled specimens presented the highest roughness values, while the POL the lowest ones. In regards to surface complexity, no difference was found among all simulated groups and the CAD/CAM group, with an exception for the POL group. All conditions presented an increase in surface complexity after HF etching. Therefore, the resin bond strength to lithium disilicate glass-ceramic is not influenced by the in-lab simulation method, meaning they are predictable approaches when evaluating the resin bonding to the tested ceramic; the surface polishing underestimates the bond results in ∼40%.