In-vitro polishing of CAD/CAM ceramic restorations: An evaluation with SEM and confocal profilometry

The objective of this In-vitro investigation was to analyze and compare the surface after polishing of 63 all-ceramic restorations fabricated out of monolithic zirconium dioxide (ZIR), lithium disilicate (LS) or feldspathic ceramic (FC) under standardized laboratory conditions with different protocols. The primary outcome was defined as the roughness (R_a/S_a) of different ceramic surfaces after distinctive polishing procedures.

The study set-up consisted of three main groups: ZIR, LS, FC (20 crowns each), every group divided into two sub-groups (10 crowns each) depending on the polishing method. The untouched glazed surface of one crown per material served as a control. Every crown displayed a defined supra-contact at the palatal cusp which was removed with a fine grain (38-45 μm) diamond bur. Surface polishing was carried out with either a two-step system (one kit for zirconium dioxide (ZIR₂), another kit for lithium disilicate (LS₂) and feldspatic ceramic (FC₂)), or a three-step system (ZIR₃, LS₃, FC₃) under standardized conditions. Roughness parameters (R_a and S_a) were measured by means of confocal profilometry. Specimens were also visually inspected with scanning electron microscopy (SEM). Statistical analyses were performed using SPSS Statistics software.

Visual examination of the specimens using SEM showed several inhomogeneities on the glazed surface of the control samples, i.e. pores and particles. On every test sample, the grinding curves of the diamond bur were still recognizable. Polishing revealed similar median R_a 0.491 μm (ZIR₂) and 0.434 μm (LS₂) after two-step polishing (p = 0.754), and 0.311 μm (ZIR₃) and 0.208 μm (LS₃) after three-step polishing (p = 0.917). Surface roughness in group FC measured 0.889 μm (FC₂) after the two-step polishing process and 0.903 μm (FC₃) following three-step surface refinement. No significant difference was detectable between surface roughness of glazed controls compared to either polished surfaces with two-step or three-step treatment within one material. ZIR and LS presented significantly lower median roughness R_a after two-step and three-step procedures than test samples of FC, measured subsequent to either of the polishing methods (p = 0.016, p = 0.009).

The surface roughness of ZIR, LS and FC crowns after the use of chairside polishing kits was comparable with the roughness measured before occlusal adjustment. A two-step procedure showed as good results as a three-step process. A smoother surface was obtained for ZIR and LS compared to FC with both polishing protocols.