Abstract
Objectives
CAD/CAM-technologies aim for a standardized, accurate production of dental restorations out of high strength materials (zirconia). The three-dimensional internal fit of CAD/CAM-manufactured zirconia copings was evaluated in vitro to verify the realizability of this aim.
Methods
The analysis was based on ceramic master dies of prepared teeth and corresponding virtual CAD surfaces. Five copings per die were manufactured with two different CAD/CAM-technologies: milling and grinding. The internal fit was determined by a three-dimensional replica technique by optical digitization and computer-assisted analysis.
Results
Mean internal gaps were 134/84 μm (SD 78/28) for molar and 93/69 μm (SD 56/35) for premolar copings (milling/grinding) using a digitizable silicone for the replicas representing the cement space; they were statistically significant regarding tooth and CAD/CAM-system ( p < 0.001).
Significance
All zirconia copings showed an internal accuracy of fit where the gap widths ranged within the current clinical recommendations. However, there still is room for improvement and further standardization of CAD/CAM-technologies.
1
Introduction
Computer-aided design and computer-aided manufacturing (CAD/CAM) technologies were developed as an alternative to the conventional casting method with the aim of producing dental restorations in a standardized, reproducible, and efficient way as well as being able to process new dental materials . A major production requirement is accuracy since the accuracy of fit of dental restorations appears to be a major factor for the long-term survival of such restorations . Traditional cast restorations are completely handmade whereas with the CAD/CAM technologies many production steps are virtual and apply design software.
With CAD/CAM techniques, copings and frameworks for all-ceramic restorations can be made out of Y-TZP zirconia (yttria-stabilized tetragonal zirconia polycrystals) satisfying the increasing demand for esthetic restorations . Due to its high stability and toughness, the machining of densely sintered zirconia proves to be time- and material consuming and may lead to surface damages of the ceramic material . For this reason, many manufacturers recommend the processing of zirconia in a presintered state. Restorations have to be sintered to their final dimensions after being machined. The resulting fit depends on two variables: the accuracy of the CAD/CAM technology applied and the control of the three-dimensional (3D) sintering shrinkage. The latter, in turn, depends on the kind of raw material used and on its pre processing .
Conventional methods that seek to evaluate the marginal or internal fit of restorations do so only in two dimensions: in different sections, the gaps between the restoration and the die are usually measured by a microscope at 4–24 points . However, a minimum number of 50 measurements per tooth were determined to be necessary in order to fully assess the marginal gaps . So far, most of the studies, including those focusing on zirconia restorations, do not meet this requirement and therefore lack consistency .
Attempts were made to evaluate the fit three-dimensionally using the replica technique or “elastomeric putty-wash technique” with a low-viscosity silicone that fills the space between restoration and die (“replica” = a duplication of the cement space) . The silicone film can be digitized with an optical system ; its thickness can be evaluated photometrically , and its density and weight can be measured . Current computer-aided techniques have a greater potential for accurate and consistent gap measurements while they also provide more and extensive information due to their non-destructive and unsectioned nature.
Due to the absence of a clinically suitable and consistent evaluation method, a clear correlation between the accuracy of fit and the survival rate of fixed dental restorations has been impossible to date. Gaps of up to 150 μm are suggested and commonly considered as being clinically tolerable . Furthermore, the thickness of the cement layer should be as uniform and as thin as possible .
Therefore, the aim of this study was to evaluate the accuracy of fit by examining posterior zirconia copings made with different CAD/CAM technologies applying a 3D-replica method. The test hypothesis was that for the analyzed copings neither the CAD/CAM system used nor the shape of the tooth restored nor the type of silicone material used for the replica technique had any influence on the evaluation of the internal fit.
2
Materials and methods
The experimental design comprised two optically digitizable ceramic master dies and their corresponding virtual surfaces. Based on these dies, ten (five for the molar and five for the premolar) zirconia copings each were CAD/CAM-made with two different systems, one using dry milling (Cercon ® , DeguDent GmbH, Hanau, Germany) and the other one applying water-cooled grinding (Cerec ® , Sirona Dentals Systems, Bensheim, Germany). The internal fit was analyzed by a 3D-replica technique and the use of both, a conventional and an optically digitizable low-viscosity silicone.
2.1
Manufacturing of the ceramic master model
The ceramic master dies and their corresponding virtual surfaces were fabricated by reverse engineering. First, the dies were made of an optically digitizable zirconia–alumina dispersion ceramic (inocermic GmbH, Hermsdorf/Thuringia, Germany; Software: ce.novation, ilmcad GmbH, Ilmenau, Germany; Digitizing: ODKM 97, Fraunhofer Institute for Applied Optics and Precision Engineering IOF, Jena, Germany) in a CAD/CAM process. The preparations were suitable for ceramic restorations (shoulder and rounded edges). Then, the corresponding virtual die surfaces were created without using the original CAD surface but data of the resulting ceramic master dies after applying optical digitization again.
2.2
Manufacturing of the ceramic copings
At the time of the study, it was only possible to make copings by means of the dry milling CAD system. Therefore, only these results were included.
Ten monophase impressions (Impregum ® , 3M ESPE, Seefeld, Germany) of each ceramic die (molar and premolar) were taken with individualized (Kerr ® , Kerr Italia, Italy), non-perforated metal impression trays. After that, one gypsum die per impression was poured with a low-expansion die stone (Esthetic-rock 285 ® , Dentona, Dortmund, Germany). The 20 dies (ten molars and ten premolars) were randomly assigned to four groups ( N = 5). For each gypsum die of every group, one zirconia coping was made by one of the two systems. In both systems, the integrated laserpoint scanners were used (Digitizing unit of Cercon brain, DeguDent GmbH, Hanau, Germany or Cerec inLab, Sirona Dentals Systems, Bensheim, Germany, respectively). According to the manufacturer’s instructions, it was not necessary to powder the plaster dies for scanning with the dry milling system due to their apricot color while for digitization with the water-cooled grinding system, titanium oxide powder coercively had to be used as contrast powder to eliminate reflections (Cerec ® Powder and Cerec ® Propellant, Vita Zahnfabrik, Bad Säckingen, Germany).
For the computer-aided design of the copings, specific system parameters, mainly ‘spacer’, can be adjusted. ‘Spacer’ regulates the cement space and thus the internal fit. Since the copings were supposed to offer the best possible fit that can be achieved with the particular CAD/CAM system, best-fit parameters were determined in preliminary tests. At first, the copings were produced with standard parameters, and the resulting fit was examined as it would be done in a dental laboratory or office (check friction, rotation, margin). The ‘spacer’ parameters (milling: molar 10 μm, premolar 20 μm, spacer surface 70%; grinding: −100 μm) were then adjusted in order to improve the fit until, based on a subjective evaluation, the decision ‘suitable for clinical use’ could be made by an experienced dentist (OM). All copings had a thickness of 0.5 mm.
Using the determined parameters, the copings were virtually designed (Cercon ® Art 1.1, Degudent GmbH, Hanau, Germany and Cerec ® inLab Ver. 2.70 R 2201, Sirona Dental Systems, Bensheim, Germany) and either dry milled out of Y-TZP zirconia blanks (Cercon ® Base 12 Degudent GmbH, Hanau, Germany) or wet ground out of Y-TZP zirconia (Vita In-Ceram YZ 20/15, Vita Zahnfabrik, Bad Säckingen, Germany). Before manufacturing the copings, the machines were calibrated, and new tools were inserted. The partially sintered ceramic needed sintering according to the manufacturer’s recommendations (Cercon ® Heat, Degudent GmbH, Hanau, Germany and Zyrkomat ® , Vita Zahnfabrik, Bad Säckingen, Germany). None of the copings was seated on either the gypsum or ceramic master die prior to the investigation of the accuracy of fit.
2.3
Evaluation of fit: making replicas of the cement space
A 3D replica technique was used to assess the accuracy of fit. In two different examination parts, the gap between each coping and the respective ceramic master die (either molar or premolar) was filled with one of the two types of low-viscosity addition silicone. In the first part of the examination, a conventional light-body silicone (Dimension Garant L ® , 3M ESPE, Seefeld, Germany) was used to make the replicas, and in the second part, an optically digitizable silicone (KwikkModel fluid ® , R-Dental, Hamburg, Germany) was used for the replicas.
All copings were axially loaded over a cotton roll with a weight corresponding to a continuous force of 20 N. This process was controlled by an ordinary scale placed underneath the ceramic master die during the setting time of the silicone. Low-viscosity silicone oil (Silikonöl Typ350, Caesar & Lötz, Hilden, Germany) was used as separating agent; a very fine film of it was spread by compressed air on the inside of the copings. Then, the copings were carefully removed leaving the silicone on the die. Since the projected light has to be reflected completely in order to achieve accurate measurements, the conventional silicone needs to be matted (Occlu Plus-Spray, Hager und Werken, Duisburg, Germany) prior to the optical digitization with white-light fringe projection (ODKM 97; IVB GmbH, Fraunhofer-Institute for Applied Optics and Precision Engineering, Jena, Germany) offering a measurement uncertainty of ∼8 μm .
2.4
Data acquisition, processing and alignment
Each silicone replica of the cement space was digitized first while still sitting on the ceramic master die. Without removing the ceramic die from the digitizer, the thin silicone film was gently removed with forceps. Then, the ceramic master die alone was optically digitized, thus being in an identical position in the measuring device and sharing identical measuring coordinates with the replica. The data sets of the premolar and the molar ceramic master die as well as of the replicas representing the inside of the 20 copings still contain outliers and stray points which have to be removed in order to improve the data quality. The procedure was described in detail in an earlier publication . A uniform distribution of the measuring points over the entire replica surface is required for proper data comparability. For this reason, all datasets were reduced to an equal point distance of 50 μm (Surfacer ® 10.6, SDRC Imageware, Ann Arbor, MI, USA).
While the replica data and the ceramic master die data were in the same coordinate system, the corresponding virtual die surfaces have their own coordinate system. The data had to be aligned in a common coordinate system before analyzing them. The ceramic master die data from the same measuring coordinate system as the respective replica data can be used as ‘means of transportation’ for the alignment of the replica data to the corresponding virtual surfaces since close to identical shapes can be aligned with the smallest possible error. While this applies to the measured ceramic master die data and the corresponding virtual surfaces, the replica data, of course, will deviate and thus cannot be aligned. The quality of the alignment is described by the RMS (root mean square) error .
2.5
Quantitative 3D analysis and statistics
Subsequently, the internal fit was calculated as difference between the replica data and the corresponding virtual surface of the respective ceramic master die ( Fig. 1 ). A total of 40 replicas (or rather their digitized data), where 20 were made with conventional silicone and 20 with digitizable silicone, were analyzed with regard to 3D differences showing the internal accuracy of fit (minimum, maximum, average, SD).
The statistical analysis was performed by a one-way variance analysis (ANOVA, at α = 0.05, SPSS 12.0, SPSS Inc., Chicago, USA). Furthermore, the standard deviations of the measurements with both silicones and both CAD/CAM-systems were evaluated (Excel ® 2007, Microsoft Corporation, Redmond, USA).
2
Materials and methods
The experimental design comprised two optically digitizable ceramic master dies and their corresponding virtual surfaces. Based on these dies, ten (five for the molar and five for the premolar) zirconia copings each were CAD/CAM-made with two different systems, one using dry milling (Cercon ® , DeguDent GmbH, Hanau, Germany) and the other one applying water-cooled grinding (Cerec ® , Sirona Dentals Systems, Bensheim, Germany). The internal fit was analyzed by a 3D-replica technique and the use of both, a conventional and an optically digitizable low-viscosity silicone.
2.1
Manufacturing of the ceramic master model
The ceramic master dies and their corresponding virtual surfaces were fabricated by reverse engineering. First, the dies were made of an optically digitizable zirconia–alumina dispersion ceramic (inocermic GmbH, Hermsdorf/Thuringia, Germany; Software: ce.novation, ilmcad GmbH, Ilmenau, Germany; Digitizing: ODKM 97, Fraunhofer Institute for Applied Optics and Precision Engineering IOF, Jena, Germany) in a CAD/CAM process. The preparations were suitable for ceramic restorations (shoulder and rounded edges). Then, the corresponding virtual die surfaces were created without using the original CAD surface but data of the resulting ceramic master dies after applying optical digitization again.
2.2
Manufacturing of the ceramic copings
At the time of the study, it was only possible to make copings by means of the dry milling CAD system. Therefore, only these results were included.
Ten monophase impressions (Impregum ® , 3M ESPE, Seefeld, Germany) of each ceramic die (molar and premolar) were taken with individualized (Kerr ® , Kerr Italia, Italy), non-perforated metal impression trays. After that, one gypsum die per impression was poured with a low-expansion die stone (Esthetic-rock 285 ® , Dentona, Dortmund, Germany). The 20 dies (ten molars and ten premolars) were randomly assigned to four groups ( N = 5). For each gypsum die of every group, one zirconia coping was made by one of the two systems. In both systems, the integrated laserpoint scanners were used (Digitizing unit of Cercon brain, DeguDent GmbH, Hanau, Germany or Cerec inLab, Sirona Dentals Systems, Bensheim, Germany, respectively). According to the manufacturer’s instructions, it was not necessary to powder the plaster dies for scanning with the dry milling system due to their apricot color while for digitization with the water-cooled grinding system, titanium oxide powder coercively had to be used as contrast powder to eliminate reflections (Cerec ® Powder and Cerec ® Propellant, Vita Zahnfabrik, Bad Säckingen, Germany).
For the computer-aided design of the copings, specific system parameters, mainly ‘spacer’, can be adjusted. ‘Spacer’ regulates the cement space and thus the internal fit. Since the copings were supposed to offer the best possible fit that can be achieved with the particular CAD/CAM system, best-fit parameters were determined in preliminary tests. At first, the copings were produced with standard parameters, and the resulting fit was examined as it would be done in a dental laboratory or office (check friction, rotation, margin). The ‘spacer’ parameters (milling: molar 10 μm, premolar 20 μm, spacer surface 70%; grinding: −100 μm) were then adjusted in order to improve the fit until, based on a subjective evaluation, the decision ‘suitable for clinical use’ could be made by an experienced dentist (OM). All copings had a thickness of 0.5 mm.
Using the determined parameters, the copings were virtually designed (Cercon ® Art 1.1, Degudent GmbH, Hanau, Germany and Cerec ® inLab Ver. 2.70 R 2201, Sirona Dental Systems, Bensheim, Germany) and either dry milled out of Y-TZP zirconia blanks (Cercon ® Base 12 Degudent GmbH, Hanau, Germany) or wet ground out of Y-TZP zirconia (Vita In-Ceram YZ 20/15, Vita Zahnfabrik, Bad Säckingen, Germany). Before manufacturing the copings, the machines were calibrated, and new tools were inserted. The partially sintered ceramic needed sintering according to the manufacturer’s recommendations (Cercon ® Heat, Degudent GmbH, Hanau, Germany and Zyrkomat ® , Vita Zahnfabrik, Bad Säckingen, Germany). None of the copings was seated on either the gypsum or ceramic master die prior to the investigation of the accuracy of fit.
2.3
Evaluation of fit: making replicas of the cement space
A 3D replica technique was used to assess the accuracy of fit. In two different examination parts, the gap between each coping and the respective ceramic master die (either molar or premolar) was filled with one of the two types of low-viscosity addition silicone. In the first part of the examination, a conventional light-body silicone (Dimension Garant L ® , 3M ESPE, Seefeld, Germany) was used to make the replicas, and in the second part, an optically digitizable silicone (KwikkModel fluid ® , R-Dental, Hamburg, Germany) was used for the replicas.
All copings were axially loaded over a cotton roll with a weight corresponding to a continuous force of 20 N. This process was controlled by an ordinary scale placed underneath the ceramic master die during the setting time of the silicone. Low-viscosity silicone oil (Silikonöl Typ350, Caesar & Lötz, Hilden, Germany) was used as separating agent; a very fine film of it was spread by compressed air on the inside of the copings. Then, the copings were carefully removed leaving the silicone on the die. Since the projected light has to be reflected completely in order to achieve accurate measurements, the conventional silicone needs to be matted (Occlu Plus-Spray, Hager und Werken, Duisburg, Germany) prior to the optical digitization with white-light fringe projection (ODKM 97; IVB GmbH, Fraunhofer-Institute for Applied Optics and Precision Engineering, Jena, Germany) offering a measurement uncertainty of ∼8 μm .
2.4
Data acquisition, processing and alignment
Each silicone replica of the cement space was digitized first while still sitting on the ceramic master die. Without removing the ceramic die from the digitizer, the thin silicone film was gently removed with forceps. Then, the ceramic master die alone was optically digitized, thus being in an identical position in the measuring device and sharing identical measuring coordinates with the replica. The data sets of the premolar and the molar ceramic master die as well as of the replicas representing the inside of the 20 copings still contain outliers and stray points which have to be removed in order to improve the data quality. The procedure was described in detail in an earlier publication . A uniform distribution of the measuring points over the entire replica surface is required for proper data comparability. For this reason, all datasets were reduced to an equal point distance of 50 μm (Surfacer ® 10.6, SDRC Imageware, Ann Arbor, MI, USA).
While the replica data and the ceramic master die data were in the same coordinate system, the corresponding virtual die surfaces have their own coordinate system. The data had to be aligned in a common coordinate system before analyzing them. The ceramic master die data from the same measuring coordinate system as the respective replica data can be used as ‘means of transportation’ for the alignment of the replica data to the corresponding virtual surfaces since close to identical shapes can be aligned with the smallest possible error. While this applies to the measured ceramic master die data and the corresponding virtual surfaces, the replica data, of course, will deviate and thus cannot be aligned. The quality of the alignment is described by the RMS (root mean square) error .
2.5
Quantitative 3D analysis and statistics
Subsequently, the internal fit was calculated as difference between the replica data and the corresponding virtual surface of the respective ceramic master die ( Fig. 1 ). A total of 40 replicas (or rather their digitized data), where 20 were made with conventional silicone and 20 with digitizable silicone, were analyzed with regard to 3D differences showing the internal accuracy of fit (minimum, maximum, average, SD).
