The aim of this in vitro study was to investigate the marginal fit of cemented zirconia copings manufactured after digital impression with Lava™ C.O.S. in comparison to conventional impressions with polyvinyl siloxane. A flowchart of the experimental procedures is given in Fig. 1. One typodont plastic tooth (left first molar, KaVo Dental, Biberach/Riß, Germany) was prepared with a circumferential reduction of 0.8–1.2 mm, occlusal reduction of 1.5 mm, chamfer finish line of 0.8 mm and convergence angle of the axial walls of 6°. The master die was replicated 40 times using a high quality vinyl silicone (DUOSIL D, SHERA Werkstoff-Technologie, Lemförde, Germany) and highly precise model resin (Mirapont®, Hager & Werken, Duisburg, Germany). Afterwards, the replicated dies were randomly divided into two groups according to the impression-taking technique (digital or conventional). The dies of each group were fixed in plaster blocks (five dies in each block) and marked in order to facilitate the identification in further steps.
Conventional impressions were taken with a polyvinyl siloxane impression material (Imprint II Garant®, 3 M ESPE, Seefeld, Germany) in a one-step technique using individual trays fabricated with cold-curing material (SR Ivolen, Ivoclar Vivadent, Liechtenstein). The impressions were inspected by the same operator under a microscope at 10× magnification (Stemi DV4 Spot, Carl Zeiss Microscopy, Jena, Germany) and then poured with a type IV plaster (SHERAHARD-ROCK, SHERA Werkstoff-Technologie, Lemförde, Germany). From the plaster models a saw-cut model was fabricated and each of the plaster model dies was individually digitalized with a laboratory dental scanner (3Shape Scanner D700, Wieland Dental + Technik, Pforzheim, Germany).
Digital impressions were taken with the Lava™ C.O.S. (Software version 3.0.2), an intra-oral digitizing system that creates the impressions by means of continuous 3D video images. These video images are possible due to three sensors that simultaneously capture the dies from three different perspectives. The dies were slightly dusted with titanium oxide powder for optical scanning (Lava™ Powder for Chairside Oral Scanner, 3 M ESPE, Seefeld, Germany) with the corresponding sprayer (Lava™ Sprayer, 3 M ESPE, Seefeld, Germany) directly before the scanning process. The data sets were sent to the company 3 M ESPE via Internet and were made available for download in the Lava C.O.S. Lab Software (3 M ESPE, Seefeld, Germany) 24 h later. The dies were then virtually cut and marked.
Scanned data from the digital and conventional group were transmitted to the 3Shape DentalDesigner™ software (Wieland Dental + Technik, Pforzheim, Germany) in order to design the copings. The marginal fitting parameters were set to 0.01 mm thickness to a level of 1 mm above the margin and the cement space was set to 0.04 mm. The data of the virtually constructed copings were then transferred into the Zenotec CAM basic software V 2.2.17 (Wieland Dental + Technik, Pforzheim, Germany) for computation of the milling paths and the milling strategies. The milling process took place in a ZENOTEC mini milling machine (Wieland Dental + Technik, Pforzheim, Germany) from a Zenostar Zr Translucent blank (Wieland Dental + Technik, Pforzheim, Germany). The enlarged copings were sintered in a Cercon® heat plus furnace (DeguDent GmbH, Germany) for 8 h at 1350 °C. The copings of both groups were placed on their respective dies and checked for irregularities under a microscope at 10x magnification (Stemi DV4 Spot, Carl Zeiss Microscopy, Jena, Germany).
For cementation, the copings were seated on their respective dies using zinc oxide phosphate cement (HOFFMANN’S CEMENT quick setting, Hoffmann Dental Manufaktur, Berlin, Germany) after try on. The copings were held under constant finger pressure for 10–15 min until the cement was set. The cement was mixed to the manufacturer’s specified powder/liquid ratio with a spatula on the rough side of a pre-cooled mixing glass slab to a fixation consistency. After removing the cement residues from the margins the dies were detached from the plaster blocks.
After 24 h all specimens were dehydrated and degreased in absolute ethanol for 2 days and subsequently infiltrated with a light-curing resin (Technovit 7200®, Kulzer & Co., Wehrheim, Germany) in ascending grades of ethanol/resin mixtures. Finally the specimens were embedded in pure resin. Each embedded specimen was cut in mesio distal direction with a diamond-coated band saw (EXAKT 300, Exakt, Norderstedt, Germany) for obtaining two parts of the same size.
Thin ground sections of approximately 30 μm thickness were prepared from one half of the specimen using the cutting-grinding technique of Donath [24] for the measurements using the optical microscope (Olympus BX51, Olympus, Japan) Fig. 2b). Images with a resolution of 0.16 μm per pixel were captured on 40x magnification using the Olympus dotSlide 2.4 - digital microscopy system (Olympus, Tokyo, Japan).
The remaining half of the dissected ceramic specimens was prepared for measurements using the TM-1000 tabletop scanning electron microscope; Hitachi, Krefeld, Germany Fig. 2a). Images in 200×, 300×, and 400× magnification were taken and analyzed using the Hitachi’s SEM software Ver. 03–02.
Marginal gap (MG) (Fig. 3A (b) and B (b)) and the absolute marginal discrepancy (AMD) (Fig. 3A. (a) and B (a)), were measured in accordance with Holmes et al. [25]. Marginal gap is defined as the perpendicular distance between the internal surface of the restoration and the preparation line for overextended copings or the perpendicular distance from the restoration margin to the tooth surface for underextended copings. The absolute marginal discrepancy represents the distance between the restoration margin and the preparation line. By rounded preparation margins the measuring point was determined by extending the main contours of the die and drawing an angle bisector as shown in Fig. 3c.
For both groups on each specimen MG and AMD were measured with an optical microscope and SEM on the mesial and distal aspect (Fig. 4). The values for either MG or AMD for the mesial and distal aspect of each slice were summed up and mean values for MG and AMD were calculated. Consequently, the groups were as follows: i) AMD OM, ii) AMD SEM, iii) MG OM and iv) MG SEM.
Five additional dies and their respective copings were manufactured in order to compare the results of MG and AMD when another technique, namely the replica technique was used. For this a light body silicone impression material (Imprint II Garant®, Light Body) was used to fill the copings before they were seated on the respective dies by applying constant finger pressure simulating the cementation process. After five minutes the copings were removed together with the adhered impression material, which indicates the gap between die and coping. To be able to perform the measurements of MG and AMD a regular body material with a different colour (Imprint II Garant®, Regular Body) was placed in the copings. After setting the silicone replicas were removed, sectioned in a mesio distal direction and the MG and the AMD were measured using the optical microscope. Since the replica technique is a non-destructive method for both, the dies and the copings, MG and AMD could be also measured after cementation using the zinc oxide phosphate cement as described above.
For each of the measurements (AMD OM, AMD SEM, MG OM and MG SEM) Student’s t-tests between the digital and conventional technique were performed. Statistical outcome was adjusted for multiple testing using Holm’s method [26]. Intraclass correlation coefficient (ICC) was used to assess agreement of SEM and optical measurements. R 2.15.1 (R Core Team 2012) and ggplot2 0.9.2.1 were used for all computations.