Statement of problem
The change in fit after ceramic layering of additively manufactured cobalt-chromium alloy frameworks has not been evaluated extensively.
The purpose of this in vitro study was to compare the fit of cobalt-chromium alloy fixed dental prostheses fabricated by different techniques before and after ceramic layering.
Material and methods
A cobalt-chromium alloy master model was prepared to receive a 3-unit fixed dental prosthesis. Sixty cobalt-chromium alloy frameworks (N=60) were manufactured by using 3 manufacturing techniques (n=20): selective laser melting (Mediloy S-Co), soft milling (Ceramill Sintron), and conventional casting as the control group (Girobond NB). The replica technique was used to measure the marginal and internal discrepancies. The frameworks were then layered with ceramic, and the same fit measuring procedure was repeated. The results were compared before and after ceramic layering within each group. The data were analyzed using the Levene, ANOVA, and paired-samples t tests (α=.05).
A significant difference was found within groups before and after ceramic layering for selective laser melting ( P =.006) and soft milling ( P =.009) but not for conventional casting ( P >.05). No statistical difference was reported in the marginal region for conventional casting group ( P =.155) in contrast with the soft milling and selective laser melting groups ( P =.003). Soft milling showed increased gap values in the marginal ( P =.006) and occlusal regions ( P =.004).
Ceramic layering increased the discrepancy of the laser-sintered and milled frameworks, increasing the marginal discrepancy.
The alteration in fit after the ceramic layering of cobalt-chromium alloy 3-unit prostheses may influence their clinical longevity.
The target of contemporary dentistry is to provide restorations with esthetics and function. Metal-ceramic prostheses are still considered the standard, especially for multiple-unit fixed dental prostheses (FDPs). High risk of ceramic chipping and core fracture has been reported for ceramic-layered zirconia, whereas long-term evaluation of monolithic zirconia is still to be assessed. With metal-ceramic restorations, cobalt-chromium (Co-Cr) has largely replaced nickel-chromium because of allergic reactions to nickel.
The conventional casting method has been considered technique sensitive, and metal shrinkage may occur during the liquid to solid phase transition, with pores or defects generated within the cast structure. Simplicity and reduced costs and manufacturing time are advantages of computer-aided design and computer-aided manufacturing (CAD-CAM) techniques , with a standardized process for producing predesigned restorations. , CAD-CAM technology comprises 2 major categories: subtractive and additive manufacturing.
Co-Cr frameworks can be milled from either partially or fully sintered industrial blocks. Milling from fully sintered blocks may produce restorations with more precision but with high wear rates of the milling tools and time-consuming processing. , Using partially sintered blocks may increase the efficiency of the milling process. Enlarged restorations must be produced to compensate for sintering shrinkage and to ensure adequate restoration fit. , , However, whether this compensation is effective is not yet clear, especially in the production of long-span FDPs. , Compared with the casting method, the subtractive technique may reduce flaws and pores but with a greater waste of materials. The milling technique has followed a standardized process for producing predesigned restorations, with reduced milling time for the soft-milled materials. ,
Additive manufacturing technology is promising, allowing the production of complex designs. , Stereolithography, fused deposition modeling, powder bed fusion (PBF), and inkjet printing are additive manufacturing techniques used in dentistry ; of which, PBF is currently the most common. PBF uses a heat source to consolidate the metallic powder, layer by layer, to produce a 3-dimensional object from a CAD. PBF includes selective laser sintering (SLS), selective laser melting (SLM), and electron-beam melting. In SLS, coherent solids are produced from metal powder at temperatures below their melting point, whereas in SLM and electron-beam melting, the metal powder is completely melted. Optimal SLS process parameters (temperature, time, geometric structure of the powder particles, powder mix composition, powder density, and protective atmosphere composition in the sintering furnace) are essential to minimize the porosity of the elements produced. Unfortunately, completely eliminating the porosities with partial melting and sintering lower than the melting point may not be possible.
Metallic structures may also be fabricated by SLM, with successive fine layers of metal powder fused by using CO 2 or fiber lasers (Nd:YAG or Yb:YAG). , The powder is melted by a focused laser beam to generate products with lower blister occurrence and better superficial finish than those made by SLS. However, additional heat treatment may be required to eliminate the high internal stresses caused by the thermal gradients during manufacturing. , SLM technology is currently the most popular approach to metal additive processing in dentistry.
Improper marginal fit was the most reported factor in restoration failure, and ceramic layering may induce changes in restoration fit. , , , Alloy choice, framework design, and the coefficients of thermal expansion of the metal and the ceramic may influence prosthesis fit. Ceramic shrinkage on firing has been reported to alter prosthesis adaptation. , , , The creep of low-density alloys under high temperature may distort FDPs. In addition, the effect of heating and cooling on CAM alloys has not been fully assessed. Furthermore, the author is unaware of studies that have compared the marginal and internal discrepancies of 3-unit FDPs produced by Co-Cr materials fabricated by different techniques before and after ceramic layering.
Therefore, the purpose of this study was to evaluate the effect of ceramic layering on the marginal and internal fit of SLM and presintered Co-Cr compared with a conventionally cast Co-Cr in 3-unit FDPs. The null hypothesis was that the application of the layering ceramic would not affect the adaptation of the restorations.
Material and methods
A master model was prepared to receive a 3-unit FDP from preparations of a typodont first right maxillary premolar and molar (Frasaco A3; Frasaco). A standardized preparation with a 360-degree 1.2-mm chamfer, a 2-mm occlusal reduction, and a total occlusal convergence of 8 degrees was made on the abutment teeth. A polyether impression (Impregum; 3M) was made, poured in melted hard wax (Preci-DIP; Yeti Dental Products), and cast in Co-Cr alloy (Solidur Co-Cr; Yeti Dental Products) by using the conventional lost-wax technique as per the manufacturer’s instructions.
The dies were digitized with a scanner (Ceramill Map 400; Amann Girrbach AG), and an FDP was designed by using a software program (Ceramill Mind; Amann Girrbach AG). The prosthesis was designed to generate a 1.5-mm space for the ceramic, with a preset cement space of 50 μm starting at 1 mm from the finish line. The data were transferred to a milling machine (Ceramill Motion 2; Amann Girrbach AG) to generate 20 frameworks from soft Co-Cr (Ceramill Sintron; Amann Girrbach AG) (Group SM) and 20 frameworks from wax (Ceramill wax; Amann Girrbach AG). The SM frameworks were then fully sintered in a furnace (Ceramill Argotherm; Amann Girrbach AG) at 1280 °C for a total time of 5 hours. To generate the control group, the milled wax patterns were cast in Co-Cr alloy (Girobond nb; Amann Girrbach AG) as per the manufacturer’s instructions.
To produce the SLM specimens from a metal powder (Mediloy S-Co; Bego), the files of the 3-dimensional model were sent to a laser machine (Sisma; Mysint100) with the following specifications: scan speed of 7 m/s, focus diameter of 40 μm, lamination thickness of 25 μm, and Yb-fiber laser power (wavelength approximately 1060-1100 nm), laser power output (200 W), and surface power density (25 kW/mm 2 ). For the powder selective melting process, a nitrogen protective gas was used as suggested by the manufacturer. The printed specimens were submitted to a heat treatment at 650 °C in a furnace (LT 15/12/P330; Nabertherm). Within 12 minutes, the temperature was increased to 800 °C and held for 15 minutes. The temperature was then lowered to 550 °C within 15 minutes.
The marginal and internal fit was measured by using the silicone replica technique ( Fig. 1 ). The framework was filled with a light-body polyvinyl siloxane (Aquasil Ultra-light; Dentsply Sirona) and seated on the die with a 40-N vertical force on the pontic central fossa. The copings were removed with the polymerized polyvinyl siloxane attached to the intaglio surface, which was supported with a medium-body polyvinyl siloxane (Aquasil Ultra Medium; Dentsply Sirona) and examined for flaws. The polyvinyl siloxane specimens were sectioned in the buccolingual and mesiodistal planes at the deepest point of the central fossae, guided by an acrylic resin template (Unifast Trad; GC) ( Fig. 2 ).
The polyvinyl siloxane specimens were gold sputtered (Sputter Coater 108 Auto; Cressington Scientific Instruments) in a vacuum by using a standard metal evaporation technique and examined with a scanning electron microscope (AIS2100C; Seron Technologies) at ×100 to ×150 magnification and 20 kV ( Fig. 3 ). Eighteen measuring points were selected (M1 to M9), starting from the most distal point from the pontic in the mesial-distal plane and from the middle of the buccal to the middle of the lingual surfaces in the buccal-lingual plane for the premolars and molars ( Fig. 4 ). The marginal measurement was calculated in all planes as the mean of the measurements made on M1, M2, M8, and M9; the axial measurement as the mean of M3 and M7; and the occlusal measurement as the mean of M4, M5, and M6.
To produce similarly layered FDPs, a template (Ceramill PMMA; Amann Girrbach AG) was generated after the preparation of a layered FDP with a ceramic thickness of 1.5 mm and an anatomic shape. Two ceramic dentin layers and a glaze were applied by following the firing protocols recommended by the manufacturer. To guide the ceramic layering, a silicone mold (Pala Lab Putty; Kulzer GmbH) was made on the acrylic resin template. The ceramic layering (IPS Style; Ivoclar Vivadent AG) was carried out as per the manufacturer’s instructions. After adding a separator (VITA Modisol; VITA Zahnfabrik), the silicone mold was adapted to the model for the second and final dentin layer. The ceramic was fired in a furnace (VITA Furnace; VITA Zahnfabrik) as per the manufacturer’s instructions. The final ceramic form was finished, and the metal margins were polished with rubber wheels (Green Coarse; Dedeco International Inc). The fit measuring replica technique was repeated. The materials used for this study are described in Table 1 .
|Girobond NB||Amann Girrbach AG||Lost wax/cast||IPS Style Ivoclar Vivadent AG||20|
|Ceramill Sintron||Amann Girrbach AG||Soft milled CAD-CAM||IPS Style Ivoclar Vivadent AG||20|
|Mediloy S-Co||Bego||Laser Sintered||IPS Style Ivoclar Vivadent AG||20|
The sample size was based on previous studies which compared the prostheses fit before and after ceramic layering and where a sample size of 10 per group was sufficient to demonstrate a significant difference in fit alteration. , , , In the present study, 20 specimens were assigned to each group. The measurements were made by using a software program (ImageJ; National Institutes of Health). One image was selected for calibration at each imaging session, and 1 investigator (E.E.D.) performed all the measurements. The intraclass correlation coefficient test showed a high intraobserver agreement (0.892), from the repeated measurements of 18 points on several abutments within 5 days of assessments. A statistical software program (IBM SPSS Statistics, v23.0; IBM Corp) was used for the analysis. Before and after means were compared by using paired-samples t tests. The discrepancy values between materials, assessed as the mean differences before and after ceramic layering, were compared by using ANOVA, and the Bonferroni test was used to compare between materials to account for multiple testing (α=.05).
A significant difference was found in fit values before and after ceramic layering in groups SLM ( P =.006) and SM ( P =.009). The difference was significant for the molars ( P =.007) before and after ceramic layering, whereas for the premolars, the difference was not statistically different ( P =.288) ( Table 2 , Fig. 5 ). A significant difference was found between material groups ( df =2, F=5.996, P =.003) but not between premolars and molars ( df =117, F=4.882, P =.099).
|Total Difference||N||Before||After||Mean||Standard Error||P|
|Mediloy S-Co||40||1262 ±167||1358 ±373||276||51||.006|
|Ceramill Sintron||40||1195 ±220||1481 ±508||308||82||.009|
|Girobond NB||40||1256 ±295||1232 ±332||-25||58||.092|
|MP||120||1222 ±228||1345 ±352||-122||48||.288|
|MM||120||1726 ±248||1486 ±488||-253||61||.007|
When the effect of ceramic layering was evaluated at the different measured points within each group, no statistical difference was found in the marginal region for the cast group ( P =.155), but a significant difference was found between the SM and SLM groups ( P =.003). The SM group showed an increase of the discrepancy values in the marginal ( P =.006) and occlusal regions ( P =.004) ( Table 3 , Fig. 6 ).
|Ceramill Sinton||Before||75 ±34||75 ±28||93 ±36||217 ±48||230 ±63||222 ±78||118 ±40||101 ±28||75 ±38|
|After||115.±68||102 ±54||96 ±41||277 ±108||299 ±63||270 ±112||121 ±38||115 ±45||105.3 ±63|
|Mediloy S-Co||Before||71 ±28||66 ±26||80 ±33||227 ±30||240 ±46||247 ±70||115 ±36||101 ±28||100 ±63|
|After||107 ±56||91 ±47||83 ±31||283 ±88||314 ±85||289 ±87||111 ±35||109 ±40||128 ±52|
|Girobond NB||Before||83 ±39||79 ±35||52 ±19||235 ±66||266 ±68||257 ±69||96 ±29||86 ±37||95 ±48|
|After||94 ±48||79 ±39||63 ±27||230 ±77||258 ±76||236 ±76||84 ±23||89 ±35||94 ±39|