In-vitro performance of CAD/CAM crowns with insufficient preparation design

Abstract Purpose To compare the debonding and fracture force of different CAD/CAM composite crowns with respect to the influence of water storage (0d vs. 90d/37 °C) and types of cementation (adhesive vs. self-adhesive). Methods Extracted human molars were prepared with a worst-case preparation scenario providing a nonretentive design (height ~4 mm; angle ~15°) and reduced fitting (250 µm). After digitalization, 72 crowns (n = 8 per group; circular wall thickness 1.5 mm / occlusal thickness ~2.5 mm) were milled from the composites (CS, LU), one experimental composite (EX), a resin-infiltrated ceramic (VE), and a feldspar ceramic reference (VM). The crowns were adhesively bonded (Scotchbond Universal + Rely X Ultimate, 3M), and two groups (EX, VE) were additionally cemented with a self-adhesive cement (RelyX Unicem, 3M). After 90-d water storage, thermal cycling and mechanical loading (TCML) were performed. Restorations, which failed during storage or TCML, were analyzed using scanning electron microscopy, and surviving restorations were loaded to fracture. To evaluate storage effects, two materials (EX, LU) were investigated without water storage. Results CS (7×) and LU (2×) exhibited debonding during 90-d storage. LU (5×) debonded during TCML. Cement remained on the inner sides of the crowns in all cases. EX and VE survived storage and TCML without failure or debonding. Two specimens of VM exhibited cracks after TCML. Fracture forces varied between 720 N and 2155 N. Solely the results between VE and VM were not significantly different ( p  = 0.204). Debonding effects due to water storage were material dependent. Fracture forces in tendency ( p  > 0.117) were higher for self-adhesive cementation. Conclusions Debonding and stability of CAD/CAM crowns were material dependent. Water storage affected debonding, and cementation marginally influenced performance and fracture force.