Provisional restoration is an important process between tooth preparation and permanent setting procedure during fixed prosthetic treatment.1 Techniques for fabricating provisional restoration are divided into direct and indirect methods.2 The indirect method can avoid the damage of pulp and the surrounding tissue from heat and residual resin monomer.2 In addition, the polymerization shrinkage can be reduced because indirect method is progressed on a stone cast.

Recently, the use of computer-aided design/computer-aided manufacturing (CAD/CAM) technology has been increasing to fabricate provisional restoration by indirect method.3 The CAD/CAM technology has reduced working time and amount of labor compared to the conventional method.4

There are subtractive and additive approaches to fabricate provisional restoration by CAD/CAM.5 The subtractive approach is three-dimensional milling of a resin block using a computer numeric controlled (CNC) machine.5 Since the resin blocks were polymerized with high degree of conversion, they have better physical properties and accuracy than direct methods.678 It is, however, impossible to produce a very complicated restoration due to the technical limitations of the milling tools.910

The additive method, also called 3D printing or rapid prototyping, is a method of building a three-dimensional model by stacking layers of materials.3 The additive method makes it easy to fabricate complicated designs that could not be made by milling. It is also economical as only the necessary amounts of materials are used.10

The provisional restoration may be adjusted for correction of contour, proximal contact, and occlusion based on the treatment plan, improvement of aesthetics in the case of discoloration or defects, relining to improve the fit of the margin area, and repairing of damage.1112 Although it would be ideal to remake the provisional restoration, it is economically advantageous to repair it at the chairside.13 If the provisional restoration fabricated by CAD/CAM method needs chairside modification, conventional monometacrylate or dimethacrylate based provisional materials are commonly used. A strong bond between the newly added material and the existing material is one of the most important factors for long-term use of the restoration. In addition, provisional restoration is exposed to various masticatory forces and occasionally to parafunctional habit, such as bruxism and clenching. Thus, it is important to fully understand the mechanical properties of provisional materials. Fracture toughness predicts the clinical performance of dental materials. High fracture toughness is essential for long-term stable use.14 There is, however, a lack of information on the fracture toughness of CAD/CAM provisional materials and compatibility with conventional materials.

The objectives of this study were (i) to evaluate the effect of surface treatment and repair material on the shear bond strength when CAD/CAM provisional materials (3D printing materials, CAD/CAM resin blocks) were repaired with conventional provisional materials, (ii) to measure the fracture toughness, and (iii) to compare their properties to conventional provisional materials. The null hypotheses to be tested are that 1) various surface treatments and types of repair materials do not affect to the shear bond strength in CAD/CAM provisional materials, and 2) shear bond strength and fracture toughness of CAD/CAM provisional materials are less than those of conventional materials.

All of provisional restoration materials used in this study are presented in Table 1. Disk-shaped specimens of 7.5 mm in diameter and 2.0 mm in thickness for each CAD/CAM material were fabricated by using different manufacturing machines (Fig. 1A).1516 Nextdent C&B was fabricated using a 3D printer (W11, Bio3D, Seoul, Korea) by digital light processing (DLP) method. ZMD-1000B Temporary was fabricated using another 3D printer (Zenith U, Dentis, Daegu, Korea) by stereolithography (SLA) method. Yamahachi polymethyl methacrylate (PMMA) disk and Huge PMMA block was milled using a milling machine (Arum 5X-200, Doowon, Daejeon, Korea). Each specimen was embedded in a polytetrafluoroethylene (PTFE) mould with an outer diameter of 20 mm, an inner diameter of 8.5 mm, and a height of 11 mm using acrylic resin (Ortho-jet, Lot No. 144217BG, Lang, Wheeling, IL, USA).1516 Luxatemp and Protemp 4 were injected into the mould with automixing gun applicators, while Jet and Alike were mixed with a clean plastic spatula for 20 seconds according to the manufacturer's instructions, and then immediately placed into the mould located on a flat glass plate. A vinyl strip was placed onto the mould and gentle hand pressure was applied to remove excess material. The specimens were allowed to polymerize at 23 ± 1℃ for 60 minutes. One hundred twenty specimens for each CAD/CAM material and sixty for each conventional material were fabricated. The CAD/CAM specimens were then divided into two groups of 60 specimens according to the repair material.

The specimens were randomly divided into six groups of 10 specimens and subjected to various surface treatments for clinical situation. Group with no treatment was set as a control group. Groups subjected to grinding with a 220-grit silicon carbide (SiC) paper (Waterproof Abrasive Paper, Daesung Abrasive Co., Seoul, Korea) and further surface treatments were set as experimental groups. After grinding, the specimens were subjected to ultrasonic cleaning in distilled water for 5 minutes. The surface treatments applied to the specimen were as follow:

Additional specimens of each material were prepared for surface examination. The surface of each treated specimen was observed at × 500 using a scanning electron microscope (FE-SEM, S-4700, Hitachi, Tokyo, Japan).

The repair procedure was performed using a stainless steel (SS) mould of 3 mm in diameter and 6 mm in height. The SS mould was placed on the treated surface and the fresh repair material was injected into the mould. The CAD/CAM materials were repaired with the Jet and Luxatemp, while conventional materials were repaired with their own materials. A vinyl strip was placed onto the mould and gentle hand pressure was then applied to remove excess material. The repair resin was then polymerized at 23 ± 1℃ for 60 minutes. The specimens treated with the bonding agent were light cured for 10 seconds using an LED light curing machine (BeLite, B & L Biotech, Ansan, Korea).

After polymerization, the specimen was removed gently from the SS mould (Fig. 1B), and then stored in distilled water at 37℃ for 24 hours before measuring the shear bond strength. After 24 hours, specimens were immediately fixed on a shear test jig and firmly mounted on a universal testing machine (Instron, Model 3345, Canton, MA, USA). The knife-edge rod was then loaded at the crosshead speed of 0.5 mm/min on the bonding interface (Fig. 1C). The maximum load was recorded when fracture occurred. The shear bond strength was then measured by applying the obtained values into the following formula:

where σ is the bond strength (MPa), F is the load at failure (N), and A is the repaired area (mm²).

The fracture surface of each specimen was examined with a stereoscopic microscope (BH-2, Olympus, Tokyo, Japan) and divided into adhesive, cohesive, and mixed failure. Adhesive failure was determined when less than 10% of the repair resin remained. Cohesive failure was determined when more than 50% of the provisional base material was fractured. Mixed failure was the intermediate between the two failure mode.17

The fracture toughness was measured by single edge v notched beam (SEVNB) technique. A 36 mm × 8 mm × 4 mm rectangular column-shaped specimens for each material were prepared (Fig. 2A). At the center of the specimen, a diamond wheel (Diamond disc, Edenta AG, AU/SG, Switzerland) was used to form a precrack of 3 mm in length at a circumferential speed of about 15,000 rpm. After dropping the glycerol into the precrack, a microcrack with a length of 0.1 to 0.4 mm was formed by a sliding motion using a scalpel blade (Surgical blade, Ribbel, New Delhi, India) (Fig. 2B). The length of the crack was measured in three parts of the specimen and the average value was obtained. The ratio of the height to the crack length (a/h) was in the range of 0.45 to 0.55.18 After ultrasonic cleansing, the specimens were stored at 37℃ for 7 days. They were taken out and stored at 23℃ for 1 hour before mechanical testing. After 1 hour, fracture toughness test was performed at a cross-head speed of 0.5 mm/min in a universal testing machine. Maximum load to fracture was recorded and the fracture toughness was calculated by the following equation18:

where P_max is the maximum load (kN), l is the span length (mm), and h, b are specimen height and width (mm).

The mean values and standard deviations of the results were calculated. The data were analyzed using one-way ANOVA and multiple comparison Scheffés post hoc test. For all statistical analyses, a significance level of 0.05 was used (SPSS 25, SPSS Inc., Chicago, IL, USA).

Shear bond strength of each material and mean and standard deviation of all materials are presented in Table 2. It was noted that bonding did not occur when the bis-acrylic resin was used as a repair material after Single Bond universal treatment.

In Nextdent C&B and ZMD-1000B Temporary, group S/SA showed the highest mean shear bond strength values. The mean shear bond strength values of both materials were also higher when Luxatemp was used as repair material.

In Yamahachi PMMA disk and Huge PMMA block, on the other hand, the mean shear bond strength values were higher when Jet was used as repair material. Both materials showed the highest mean shear bond strength values in group S/SA like 3D printing materials.

Jet and Alike showed the highest mean shear bond strength values in the group S/SA, while Luxatemp and Protemp 4 showed the highest value in group S/HF.

The SEM images of CAD/CAM materials are presented in Figs. 3, 4, 5, 6. Different topographic patterns were observed depending on the fabricating method and the surface treatment. The 3D printing material was layered and the CAD/CAM resin block was milled during the fabricating procedure, so lines on the surface of the specimen were observed in a certain direction in the control groups. In the groups grinded with SiC paper, long grooves, which are macroretentive features, were observed. When hydrofluoric acid gel was applied, smooth edges, dissolved fillers, and irregular porous forms were observed especially noticeable in bis-acrylic resins. In sandblast treatment after SiC paper grinding, micromechanical retention systems, such as irregular surface and many channels, were observed.

The failure modes of all groups in this study were shown graphically in Fig. 7. When the 3D printing material was repaired by Luxatemp and the CAD/CAM resin block was repaired by Jet, cohesive failure was more frequent. It was also observed when mechanical surface treatment, especially sandblasting was treated.

The mean fracture toughness values of the materials were presented in Table 3. Yamahachi PMMA disk showed the highest mean fracture toughness value and Huge PMMA block was the second highest. The 3D printing materials showed the lowest values, but there was no statistical difference from Alike and Luxatemp (P > .05).

The provisional restoration may be adjusted for repair or relining procedure at the chairside and the repaired provisional restoration should be maintained until definitive treatment is placed. To improve the bond strength, the surface roughness should be increased to promote mechanical interlocking, or the bonding agent may be applied to increase the wettability and induce chemical bonding.19 Choice of the proper repair material is also critical point. Thus, in this study, mechanical methods including silicon carbide (SiC) paper grinding, 5 µm Al₂O₃ sandblasting, and 4% hydrofluoric acid etching and chemical methods including the use of bonding agent and adhesive containing silane were chosen for increasing surface roughness. The reason for choosing these treatments was that they could be simply and frequently used in all dental clinics.161920212223 Especially, grinding with carbide or diamond bur is the essential process for removing the contamination on surface and preparing a uniform space for the added material when the provisional restoration was repairing or relining. Thus, grinding was selected as the base treatment and included in all groups. For grinding procedure, SiC paper was selected to replace a carbide or fine diamond bur.24 It has the advantage that it can give more uniform and stable force and is easy to apply than carbide or diamond bur in laboratory research. Lee et al. proposed the use of 220-grit SiC paper for this purpose.25 Combination of three or more treatments was also excluded as it was economically disadvantageous in clinical situation. Conventional methacrylate-based provisional materials such as Jet and Luxatemp were selected as repair materials. The reason for choosing these materials is that they had been verified over the long-term and were one of the most widely used materials in clinical practice. They were also more economical than Alike and Protemp 4, respectively.

The shear bond strength of group SiC was higher than that of group CON in most of the materials because of macro retentive features such as long grooves on the surface.19 However, group CON of Jet showed opposite results. It is because a number of air bubbles that could be effective in increasing surface roughness have been generated, even though vibrator is used for mixing process. But in other groups of Jet, it was confirmed that the surface bubbles were mostly removed by SiC paper grinding. There was also the possibility that grinding debris reduced roughness.26

Sandblasting of the composite resin before repair procedure increases the bond strength by forming micro retentive features on the surface.19272829 Some studies have reported that the micromechanical retention was the most important factor in repairing the composite resin.2730 In this study, it was confirmed that the group S/SA of all materials increased the shear bond strength as compared with the group SiC. Therefore, it is considered that sandblasting after SiC paper grinding is an effective surface treatment method to increase the bond strength.

Hydrofluoric acid gel (HF) etches glass particles in the matrix to form porous surfaces and gaps to provide micromechanical retention.31 In this study, it was found that the HF etching did not affect or rather reduced the bond strength in PMMA. It may result from the low amount of glass particles and the high hydrophobic monomer in PMMA. In addition, HF etching softened the PMMA surface that resulted in decreasing the surface roughness.2032 On the contrary to PMMA materials, HF etching in bisacrylic resins such as Luxatemp and Protemp 4 increased the shear bond strength, as they contained more than 24% filler in the volume ratio of resin. It was verified from the results that all of bis-acrylic resins showed the highest mean shear bond strength values in group S/HF. The etching effect of HF depends on the amount and type of filler. If the components of the resin are not known accurately, routine use of HF should be avoided.203132

Many studies have reported that the bond strength is increased when bonding agents is used.192233343536 The polymers of the bonding agent flow into the microcrack of the polymer matrix to form a micro-entanglement.1937 The bonding agent also improves the wettability of exposed filler particles and inhibits the formation of air voids.193638 Some studies have shown that bond strength is increased when silane coupling agent is applied with bonding agent.1939 Silane also improves the wettability of the adhesive system.3140 In this study, the silane-free adhesive and Single bond universal, which was an all-in-one product with silane, were used as bonding agents. The results of this study showed that both of the bonding agents had almost no effect on bond strength. In the Single Bond universal treated group, the silane did not have any additional effect to bond strength and the bonding did not even occur when the bis-acrylic resin was used as the repair material. The reasons for this phenomenon are that: 1) the pH of the Single Bond universal is low by about 2.7 due to the presence of functional monomer.41 2) there are few or no inorganic fillers in PMMA that react with silane.31 3) the Single Bond universal contains HEMA, a hydrophilic monomer that absorbs water, which cause swelling and weak bonding.4243 Another interesting phenomenon is that the similar bond strength was obtained when the CAD/CAM materials were repaired with the bis-acrylic resin after the adhesive treatment. It is assumed that the adhesive flowed into the surface of the specimen was responsible for most of the bond strength than repair material.

The 3D printing materials showed a high bond strength when it was repaired with Luxatemp. Because Luxatemp flowing into the matrix formed a more complex network and a harder bond than Jet. On the other hand, the CAD/CAM resin blocks showed the opposite results. As the resin blocks had a high degree of convergence and were composed of cross-linked PMMA, the bis-acrylic resin with a high molecular weight monomer was difficult to penetrate into the matrix.

The bond strength after repair procedure is determined by the number of unconverted C=C double bonds that can react with the repair material.31 As time passes after the polymerization, the number of C=C double bonds is decreased and cross-linking of the polymers is increased. It is then difficult for fresh monomer to penetrate into the matrix.16 The specimens of CAD/CAM materials used in this study had been prefabricated, while the specimens of conventional provisional materials were examined just after one hour of polymerization. Although conventional provisional materials have a lot of the unconverted C=C double bonds, there was no statistically significant difference from CAD/CAM materials in the shear bond strength. As a result, the value of shear bond strength is clinically acceptable, when CAD/CAM materials were repaired at the chairside.

In clinical practice, the fracture of provisional restoration is caused by crack propagation that started from the surface. The fracture toughness is an appropriate parameter to measure the ability to resist various masticatory forces.44 Fahmy and Sharawi44 showed the highest fracture toughness of PMMA, followed by bis-GMA resin and PEMA. Balkenhol et al.45 reported that the bond strength of the cross linked Protemp 3 Garant was higher than the not-cross linked Trim. Tahayeri et al.46 reported that Nextdent C&B has low physical properties due to a little amount of fillers but it is within the range that can be used in clinical practice. CAD/CAM resin blocks showed the highest fracture toughness. Although 3D printing materials showed the lowest values, it showed no statistically difference with Alike and Luxatemp (P > .05). Thus, it is considered that the CAD/CAM materials have suitable fracture toughness for clinical use as compared to conventional materials.

Within the limitation of this study, the following conclusions were drawn. In order to successfully repair the CAD/CAM provisional restoration, mechanical surface treatment and appropriate repair material according to the CAD/CAM material type should be selected. The CAD/CAM provisional materials have proper mechanical properties for clinical use as compared to conventional materials.