Introduction

Due to ongoing progression of computer-aided design and computer-aided manufacturing (CAD/CAM) techniques and high strength ceramic materials, the current use of full-ceramic restorations has increasingly developed.¹ Among current CAD/CAM ceramic restorations, due to being esthetically pleasant and resistance to fracture, zirconia crowns are being used widely nowadays.² Among the basic requirements for a successful long-term fixed dental crowns, marginal and internal adaptations are of great importance.^6-8

The cement space between the crown and the underlying tooth should not exceed the clinically acceptable limits.⁹ A cement space of 50 to 100μm is considered the acceptable range for internal gap.¹⁰ Beside increasing the chance of cement wash out and recurrent caries, it has been reported that the internal gaps of more than 70μm could decrease the fracture resistance of zirconia crowns.⁸ Also, a marginal discrepancy of 73 to 150μm has been reported to be clinically desirable.^11-13

Marginal and internal adaptations of monolithic and layered zirconia crowns and fixed partial dentures have been the subject of various studies. Different parameters that could possibly affect the adaptation of the zirconia crowns in these studies include finish line design, impression technique (conventional vs. digital), scanning system, fabrication system (CAD/CAM system), zirconia block type (green vs. pre-sintered), sintering process, veneering technique, firing cycles numbers, and also the design of fixed partial denture.^14-35 The porcelain framework design might influence the crown adaptation, as well as its esthetic and strength. Therefore, this study aimed to evaluate the effect of four different framework designs on the marginal and internal fits of CAD/CAM zirconia crowns.
Materials and Methods
A sound maxillary right central incisor extracted due to periodontal problems was used in this in vitro study. The tooth was mounted in a resin pattern block (ACROPARS, Marlic Medical Industries Co, Iran) 3mm apically to the CEJ. Condensational silicone material (Speedex putty, COLTENE, Berlin, Germany) was used to make indices from the tooth crown. Also, the crown segment of tooth was scanned by a scanner (Open Technologies Dental, Italy, Brescia). Using a high-speed handpiece and a torpedo diamond bur, the tooth was prepared 1mm coronally to the CEJ, with 1.5mm incisal reduction and 1mm axial walls reduction with a 6-degree taper and a heavy chamfer finish line.

After scanning the prepared tooth (Figure 1), four different frameworks were designed using EXOCAD software (Exocad Dental CAD, Darmstadt, Germany) with 60micron cement space. Four groups each containing 10 samples was formed according to the framework designs.

Group 1: consisted of simple 0.5mm thickness zirconia cores without any anatomical contour (Figure 2, A).

Group 2: included zirconia cores with a cut back design contour and a 3mm lingual collar and proximal struts up to the half of the proximal wall height (Figure 2, B).

Group 3: consisted of zirconia cores with anatomic core contour. To create this group first an acrylic resin pattern of the full contour tooth was made using the previously made putty index. Then the resin pattern was cut back according to the mean enamel thickness in five different points (mesial, distal, buccal, lingual, incisal) (Figure 2, C). Mean enamel thickness was calculated using CBCT of five sound central maxillary teeth. Then, the cutback resin pattern was scanned for designing the zirconia anatomic core.

Group 4: consisted of full anatomical contour monolithic zirconia crowns (Figure 2, D).

The milling process was carried out using a green zirconia blank (KATANA, Kuraray Noritake Dental Inc, Japan). Then, veneering porcelain (Zr-FS, GC, Europe A.G., E.U.) was applied by an experienced technician on the first three core groups. A previously made putty index was used to ensure the same full contour for all the crowns.

Prepared tooth surfaces and internal/marginal surfaces of crowns were scanned by a scanner (ATOS, GOM GmbH, Braunschweig, Germany) with an accuracy of 1 million dots per scan. This scanner has two cameras that receive the reflected beam and with its own algorithms (Triple Scan Principle and Dynamic Referencing) produces a 3D image which can be used by various softwares. After scanning the specimens, the crowns and the tooth were aligned in CATIA software (Dassault Systèmes SE, Vélizy-Villacoublay, France) (Figure 3) to analyze the internal and marginal fit between the crown and the tooth (Figure 4). Internal gap was measured in six areas (mesiopalatal, mesiobuccal, distopalatal, distobuccal, midpalatal, and midbuccal) and marginal gap was measured in eight areas (mesiopalatal, mesiobuccal, midpalatal, midbuccal, distobuccal, distopalatal, mesial, and distal). Each point was measured for ten times and the average measurement of ten points in each area was considered as the gap of that area.

The collected data were analysed using R 3.6.0 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) software using 2-way ANOVA and Tukey tests. The statistical significance level was determined to be at 0.05.
Results
The results showing the mean marginal and internal gaps of 4 different zirconia core designs crowns at the six measuring points are presented in Table 1. According to one-way ANOVA analysis, the mean internal gap was the highest in anatomic core design group (75.33 μm) and the lowest in monolithic crown group (49.07 μm) (P=0.007). Although, the mean marginal gap difference in the cut back design and monolithic groups was statistically significant (P=0.043), the differences between the other groups were not significant (P>0.05).

Table 1: Mean (and standard deviation) marginal and internal gaps of zirconia cores by 6 measuring areas.

According to the Tukey test the monolithic crown group had the least marginal gap in all the measured areas, except for the distal area. There were statistically significant differences between monolithic crown group and cut back design group in mesiopalatal, midpalatal, and distopalatal areas (P = 0.024, 0.009, and 0.032, respectively).

Based on the two-way ANOVA analysis, there were significant differences among 6 measuring points. The interactive effect of the two factors was not significant, indicating that there are no statistical differences in design groups and measuring areas combinations (Table 2).

Table 2: Results of two-way ANOVA analysis.

R Squared = .201 (Adjusted R Squared = .167)

Discussion

In partial fixed dentures, the effect of the prosthesis design (straight vs. curved) has been studied on the marginal gap.²² However, in single zirconia crowns proper framework architecture has been evaluated only in term of its effect on esthetic and fracture strength of the restoration.³⁶ Therefore, this study aimed to evaluate the effect of three different designs of layered zirconia copings and the anatomic design (monolithic) on the marginal and internal adaptation of finalized zirconia crowns. While the importance of marginal gap lies in the prevention of carious and periodontal disease,⁷ a thin and uniform internal gap is necessary for resistance, retention and strength of the zirconia restorations.³⁷

According to the findings, internal and marginal gaps of the non-veneering (monolithic) group were significantly lower than the three veneering groups. However, there was no significant difference between the veneering groups. Marginal gap in monolithic group (41.50±5.15μm) was significantly lower than cut back design group (63.66±20.5μm). However, all these gaps were within clinically acceptable ranges.¹² Also, internal gap of the monolithic group was significantly lower than the cut back and anatomic core design groups (70.73±14.30 and 72.31±18.33μm, respectively). Internal inaccuracy of zirconia restorations has been generally attributed to the cooling process, firing contraction, and also the geometry of the milling burs.^19,23,24 However, using additive manufacturing techniques could overcome the limitation caused by the size and shape of burs used in the subtractive CAD/CAM milling method.³⁸ The most amount of veneering porcelain was used in the simple core design group and the smallest amount was used in the anatomic core design group. Therefore, it seemed that the amount of veneering porcelain did not necessarily affect the gap. Furthermore, high standard deviation (SD) seen in this study might be due to various marginal fit of each crown at different locations.³⁹

One of the factors that could affect the marginal accuracy is the material used for crown fabrication. Kunii et al.¹⁴ studied the effect of sintering on the marginal and internal gaps of the KATANA zirconia crowns (the same zirconia material as used in this study). They reported a lower marginal gap (3.6±5.8μm). However, the amount of axial gap was similar to that of the monolithic group in this study, and the occlusal gap was higher than all the groups in our study.

It has been reported that the veneering process including layering, pressing or CAD-on techniques could increase the marginal gap. However, a higher marginal gap has been associated with the hand layering technique.²⁵ The reasons suggested for this finding include the higher mismatch of coefficients of thermal expansion (CTE) of coping and veneering porcelain. Also, the shrinkage of veneering porcelain due to the more firing cycle numbers in this technique could lead to lifting of the margin off the die.^26,27Also, it has been reported that repeated firing cycles could change the CTE of both core and veneering porcelain.^40,41Other studies have also showed an increase in the marginal gap after veneering process due to compressive stresses created in the coping after melting and contraction of veneering porcelain particles.^19,28,29 The anisotropic contraction of zirconia copings and the inconsistent firing shrinkage of the veneering porcelain could have also played a part in the results.³⁰ However, this simple contraction force might not quite explain the marginal gap developed after veneering process due to the high flexural strength and fracture toughness of zirconia framework.⁴² Therefore, a more reasonable explanation could be development of microcracks during the milling process of pre-sintered zirconia block. They could then cause stress points after subjecting to the wet veneering porcelain and grow during firing resulting in 3-5% volume increase due to zirconia phase transformation. This compressive stress together with the CTE incompatibility between zirconia coping and the veneering porcelain could cause the marginal gap.^{27, 31,43}

Differences in the studies regarding the increasing the marginal gap after veneering process might be attributed to different veneering thickness and uniformity of porcelain mass (greater marginal gaps in buccal and palatal margins in comparison to mesial and distal margins), different core thicknesses, different number of firing cycles, different coping shape (regarding the tooth shape), different materials used for core and veneering, operator manual skills, geometric complexity of the restoration (single crown versus fixed partial dentures), and also various methods used for measuring the gaps. ^{21, 28,32,44,45} To eliminate the errors associated with non-digital measuring methods, crowns were digitally scanned using GOM ATOS scanner and the marginal and internal gaps were measured by CATIA reverse engineering software in this study. Measuring internal gap is only possible through digital measurement, internal microscopes or replica technique. ^46-48

Another factor reported to affect the marginal fit of zirconia restoration is cement space which has been suggested to not to be less than50- 60 micron.^33,34 Excess cement space could also decrease the fracture strength of the prosthesis and add to the failures of the veneering porcelain.^49,50 On the other hand, according to Yilmaz et al³⁵ decreasing the cement space could improve  the marginal gap of monolithic crowns.

 One of the limitations of this study was not subjecting the samples to the aging process to simulate intraoral conditions. Also, more studies should be conducted to evaluate the marginal and internal gaps of frameworks before application of veneering porcelain.

Conclusion

Within limitations of this study, the amount of internal and marginal gaps of zirconia crowns with four different coping designs were within clinically acceptable range. Also, the monolithic group had the lowest marginal and internal discrepancies.