- Email: [email protected]
On the wear behavior and damage mechanism of bonded interface: Ceramic vs resin composite inlays
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
Contents lists available at ScienceDirect
Journal of the Mechanical Behavior of Biomedical Materials journal homepage: www.elsevier.com/locate/jmbbm
On the wear behavior and damage mechanism of bonded interface: Ceramic vs resin composite inlays
T
Ping Yua, Yuhuan Xionga, Peng Zhaoa, Zhou Xub, Haiyang Yua, Dwayne Arolac,d,e, Shanshan Gaoa,∗ a
State Key Laboratory of Oral Diseases, National Clinical Research Center for Oral Diseases, West China Hospital of Stomatology, Sichuan University, Chengdu, China Shanghai Putuo District Eye & Tooth Disease Control and Prevention Hospital, Shanghai, China c Department of Materials Science and Engineering, University of Washington, Seattle, WA, USA d Departments of Oral Health Sciences, School of Dentistry, University of Washington, Seattle, WA, USA e Departments of Restorative Dentistry, School of Dentistry, University of Washington, Seattle, WA, USA b
A R T I C LE I N FO
A B S T R A C T
Keywords: Inlays Bonded interface Damage mode Ceramic Resin composite Wear
Advances in adhesive technologies have increased indications for the use of inlays. Decrease in the bonded interface integrity due to wear has been cited as the main cause of its failure. However, this process of interface degradation and the influence of inlay material on damage mechanism appear to be poorly understood. Thus, we aimed to compare the wear behavior and interface damage between ceramic and resin composite inlays bonded to enamel under sliding contact and use the experimental findings to support recommendation of the appropriate inlay material. Bonded interface specimens involving tooth enamel and either ceramic or resin composite inlays were prepared and subjected to reciprocating wear tests up to 5×104 cycles. The wear track profiles and morphologies were characterized after increments of cyclic sliding contact using white light interferometry and scanning electron microscopy, respectively. Optical microscopy was used to evaluate sub-surface cracks and their propagation within the samples. A finite element analysis was used to analyze the stress distributions of the bonded interfaces. Composite inlays showed higher wear depth than the ceramic in the early stage (N ≤ 5×102 cycles), while no significant difference was found at the later stage. For ceramic inlay a greater portion of the contact load was concentrated in the ceramic structure, which facilitated cracks and chipping of the ceramic inlay, with rather minimal damage in the adjacent interface and enamel. In contrast, for the resin composite inlay there was larger stress concentrated in the adjacent enamel, which caused the development of cracks and their propagation to the inner enamel. The restoration material could contribute to the stress distribution and extent of damage within enamel-inlay bonded interfaces. A tough ceramic appears to be more effective at protecting the residual dental tissue.
1. Introduction The loss of substantial tooth structure most frequently occurs in posterior teeth as a result of primary caries, fractures and aging related degradation of existing restorations (Fron et al., 2013). With the development of improved adhesive technologies, inlays have been the preferred treatment option in such cases due to the greater emphasis on preservation of tooth structure (Holberg et al., 2013). Various types of restorative materials are available for use as inlays, yet ceramics and resin composites are presently the most widely used due to their excellent aesthetic performance. In comparing these two, dental ceramics exhibit superior mechanical properties and wear resistance, while resin
composite restorations are less brittle and more “user-friendly” (Höland et al., 2008; Moszner and Klapdohr, 2004; Papadogiannis et al., 2008; Yap et al., 2004). According to the modified US Public Health Service (USPHS) criteria, besides ideal anatomical form, good color match and smooth surface texture, a successful restoration should have perfect marginal integrity (Santos et al., 2016). While clinical evaluations have commented that an obvious deterioration in marginal adaptation was widely observed in both ceramic and resin composite inlays, especially in the occlusal margins (Kraemer and Frankenberger, 2005; Krämer et al., 2006, 2015; Peumans et al., 2010, 2013). For instance, Peumans et al. reported only 21.7% of the ceramic inlays exhibited an excellent
∗ Corresponding author. State Key Laboratory of Oral Diseases, National Clinical Research Center for Oral Diseases, West China Hospital of Stomatology, Sichuan University, Chengdu, 610041, China. E-mail address: [email protected] (S. Gao).
https://doi.org/10.1016/j.jmbbm.2019.103430 Received 25 July 2019; Received in revised form 9 September 2019; Accepted 12 September 2019 Available online 12 September 2019 1751-6161/ © 2019 Elsevier Ltd. All rights reserved.
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
were prepared with leucite-reinforced ceramic (IPS e.max CAD, Ivoclar Vivadent, Schaan, Liechtenstein) or resin composite (Tetric N Ceram Bulk Fill, Ivoclar Vivadent, Liechtenstein). The preparations involved use of 4.9% hydrofluoric acid gel (IPS ceramic etching gel, Ivoclar Vivadent, Schaan, Liechtenstein), 35% phosphoric acid gel (Eco etch, Ivoclar Vivadent, Liechtenstein) and self-adhesive resin cement (Rely X U200, 3M, America).
marginal adaptation compared to 70.7% at baseline after 24 months (Peumans et al., 2013). The percentage of marginal deficiencies increased to 98% for ceramic restorations and between 84% and 92% for composite restorations after only 8 years of clinical use (Kraemer and Frankenberger, 2005; Peumans et al., 2010). Degradation of the bonded interface integrity gradually increased with the extended time, which could cascade into various complications, such as marginal discoloration, secondary caries, dentin sensitivity, pulpitis, and even fracture of the restoration (Santos et al., 2016). Therefore, the interface integrity could be considered a foundational requirement for long-term effectiveness of inlay restorations. Conversely, an accelerated decrease in integrity is one of the primary contributions to restoration failures. Analyzing the bonded interface degradation process under simulated chewing forces could help to predict the longevity and clinical performance of restorations. Mechanical strength of inlay restorations is important for their durability in clinic applications. The differences in properties between ceramic and resin composite may result in different wear behavior of inlay-restored teeth as they would influence the stress concentration the bonded interface between the tooth and restoration (Flávia Zardo Trindade et al., 2018; Thordrup et al., 2001). To date, various investigations in comparing ceramic and resin composite inlays have been reported, including both clinical and laboratory studies. The results of clinical studies are inconsistent; there is no definite conclusions on which material performs better and has the best chance for long-term survival (Fron et al., 2013; Grivas et al., 2014; Lange and Pfeiffer, 2009; Morimoto et al., 2016; Pol and Kalk, 2011). In vitro tests have largely been focused on the effects of the restorative material on the stress distribution in tooth structures and fracture strength. Similar to a comparison of findings from the clinical studies, the results are sometimes contradicting. According to some authors, teeth restored with ceramic inlays achieve higher fracture resistance than those restored with composite inlays (Belli et al., 2005; Costa et al., 2014; Magne, 2007; Magne and Belser, 2003; Manhart et al., 2001; Mesquita et al., 2006; TE, 2013; Yamanel et al., 2009), while others have reported that glass-ceramic inlay materials elevates the risk of fracture (Ausiello et al., 2004; Pest et al., 2006; Soares et al., 2004; St-Georges et al., 2003). Yet, there are also reports that the fracture resistance of teeth restored with ceramic and resin composite inlays is similar (Da et al., 2004; Shor, 2003). The difference of mechanical strength may be related to the inlay material, bonding operation, test method, etc. However, fracture strength test could reflect the overall strength of restored teeth, rather than the resistance of inlay bonded interface to masticatory stress, which exhibit limited significance in the clinical prediction of inlay restorations. Hence, despite the significance of this issue, the wear behavior and damage mechanism of inlay bonded interface are still unclear, let alone the difference between ceramic and resin composite inlays. Therefore, the process of damage initiation and evolution degradation behavior and the damage mechanisms of enamel-inlay (ceramic and resin composite) bonded interfaces were investigated and understood in this study. Furthermore, stress propagation and potential crack growth inside adhesively restored teeth under occlusal loading have been simulated. The aim of the study was to develop an evidencebased understanding of the evolution in enamel-inlay bonded interface integrity that could help improve the inlay bonded interface integrity and support an informed choice of inlay materials.
2.2. Specimen preparation Thirty bonded-interface specimens were prepared for each inlay material according to the following procedures. The teeth were inspected to exclude those with caries, cracks and fluoric mottle. The crowns were separated from the roots and then cut into sections along the mesial-distal axis with a diamond-coated band saw (Struers Minitom; Struers, Copenhagen, Denmark) under continuous water coolant. A total of sixty tooth sections were prepared. The leucite-reinforced ceramic (IPS e.max CAD) was cut to sections of 6×6×4 mm3 and sintered according to the manufacturer's protocol. At the same time, the resin composite (Tetric N Ceram Bulk Fill) was applied incrementally to fill the mold cavity between two glass plates and lightcured for 60 s at a time with an LED-type light source (Bluephase, 800 mW/cm2, Ivoclar Vivadent). After 24 h, the resin composite stripes were released from the mold and cut to sections with cross section of 6×6 mm2 and length of approximately 4 mm. Thirty ceramic blocks and thirty composite blocks were fabricated in total. The sixty tooth blocks were randomly divided into two groups, and then bonded to ceramic or resin composite sections. The interior faces of the tooth sections and one of the 6×4 mm2 ceramic/composite surfaces were ground with #280 mesh silicon carbide (SiC) abrasive papers (Struers, Copenhagen, Denmark) under continuous water irrigation to prepare the surfaces for bonding. After cleaning with alcohol, the abraded surfaces of the tooth enamel were etched with the 35% phosphoric acid gel (30 s for enamel, 15 s for dentine), rinsed for 10 s and air dried. The abraded surfaces of ceramic blocks were etched with 4.9% hydrofluoric acid gel for 20 s, rinsed for 30 s with water spray and dried with oil-free air. The RelyX U200 resin cement was then mixed following the manufacturer's instructions and applied to the entire ceramic or composite surface. Subsequently, the inner surface of the tooth section was positioned on the resin cement (Fig. 1A), taking care that the occlusal surface of the tooth section was level with the face of the ceramic/ composite block. After removing excess adhesive, light curing was then performed for 60 s with the LED-type light source. To control the resin cement thickness, indicator papers with a thickness of approximately 50 μm were utilized. The sixty bonded specimens were stored in artificial saliva for 24 h at 37°C and then embedded in an epoxy resin (Struers, Copenhagen, Denmark) with the occlusal surface facing outward (Fig. 1A). After solidification, the exposed surface of each specimen was ground with SiC papers in a sequence of decreasing abrasive sizes (P280, P800, P1200, P2400, and P4000 mesh number) under continuous water irrigation. Then, all of the specimens were highly polished sequentially using DiaDuo-2 (Struers, Copenhagen, Denmark) with 3 μm particle size for 5 min, followed by OP-S NonDry (Struers, Copenhagen, Denmark) with particle size of 0.04 μm for 5 min. The polishing was performed on a dedicated instrument (Struers, Copenhagen, Denmark) with felt cloths (Dac, Struers, Denmark) under constant water lubrication. Finally, the specimens were ultrasonically cleaned for 5 min in deionized water (KQ-50B, Shumei, Kunshan, China) and inspected under an optical microscope (BX51RF, Olympus, Tokyo, Japan) to measure the interface width at magnification of 20×. The width of the interfaces was all between 30 μm and 80 μm (Fig. 1B).
2. Materials and methods 2.1. Materials The bonded interface specimens were prepared using freshly extracted human third molars (donor age: 18 years to 25 years). Both the collection and the experimental procedures involving the teeth were approved by the Research Ethics Committee. Model inlay preparations
2.3. Wear tests Reciprocating wear tests were conducted on occlusal surface of the 2
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 1. Details of the specimen configuration. A: schematic diagram of the specimen preparation (section view), B: optical microscope view of the bonded interface specimen surface. The d in figure B indicates the width of bonded interface. The red two-way arrow in figure B shows the position of the reciprocating wear test.
and OP-S NonDry for 5 min as previously described under constant lubrication (Xu et al., 2019; Yu et al., 2017). Finally, the specimens were ultrasonically cleaned for 5 min in deionized water. Then all the specimens were examined by optical microscopy to observe subsurface damage and cracks extending beneath the worn surface.
bonded interface specimens (Fig. 1B) using a commercial machine (Rtec instruments inc, MFT5000, USA), with a ball-on-flat contact mode. Silicon nitride balls with a diameter of 6.35 mm were used as the antagonist (Jung et al., 2000; Krejci et al., 1999). All tests were performed at room temperature under simulated artificial saliva lubrication. The testing parameters included an imposed contact load of 20 N, a reciprocating amplitude of 2 mm (Lawson et al., 2013) and a frequency of 1 Hz. The displacement midpoint of the articulations was located at the bonded interface. A total of 5×104 cycles were performed and three intervals of analysis were employed over the whole cycle in order to analyze the evolution of interface degradation, including after 5×102, 5×103 and 5×104 cycles (Xu et al., 2019; Yu et al., 2017). Ten samples were used to evaluate the wear characteristics for each interval (3x10 = 30 specimens per group). The wear scars were scanned using a white light interferometer operating in the vertical scanning mode equipped with the tribometer. From each scan obtained, the wear depths of the bonded interface area were calculated, including adhesive, nearby enamel and inlay material. The locations of measurement for the enamel and inlay material were 30 μm in distance from the adhesive border. Then, three-dimensional topography maps of the wear tracks were reconstructed using the commercial software (Gwyddion 2.30, Department of Nanometrology, Czech Metrology Institute), which accompanies the interferometer.
2.7. Finite element analysis In order to analyze the differences in stress distribution caused by the different inlay materials, two-dimensional finite element models consisting of tooth enamel, adhesive and inlay (ceramic and resin composite) were developed using a stress analysis software ANSYS 10.0 (ANSYS, Inc., Canonsburg, USA). All of the materials were treated as isotropic and homogeneous. The elastic modulus and Poisson's ratio of the meshed models are listed in Table 1 (Ausiello et al., 2017; Selma et al., 2016; Takenaka et al., 2015; Tolidis et al., 2012; Wendler et al., 2017). The interfaces between different parts of the models were defined as perfectly bonded and nodal displacements at the bottom of the models were constrained in all directions. The width of the adhesive interface was set as 50 μm. A vertical load of 20 N was applied at different locations of the occlusal surface of the bonded interface model through a rigid cambered contact surface. This contact surface was defined with radius of curvature equivalent to that of the Silicon nitride antagonist ball. The placement of the contact load was varied as a function of distance from the bonded interface, and the principal stress was plotted and analyzed.
2.4. Statistical analysis Statistical analyses were performed using the software SPSS 18.0. A one-way ANOVA and LSD Fisher test were performed to assess the difference in wear depths between the tooth enamel, inlay material and adhesive. The difference between the two groups of inlay materials was analyzed using a Student's t-test. A p-value of less than 0.05 was considered statistically significant.
3. Results 3.1. Wear behavior A comparison of typical 3D topography and vertical profiles of wear tracks is presented in Fig. 2 after the three different periods sliding contact. Profiles of the wear tracks were obtained from near the center of the tracks and perpendicular to the bonded interfaces along the path of articulation (Fig. 2A). As evident from this figure, the wear depth and wear volume increased with number of testing cycles. The wear track
2.5. Surface morphology In order to analyze the overall progression of wear and damage mechanisms, representative specimens from each group were ultrasonically cleaned and examined using a scanning electron microscope (SEM, INSPECTE, Czech, Republic) after 5×102, 5×103, and 5×104 cycles of testing.
Table 1 Elastic modulus and Poisson's ratio that were used in the finite element simulations.
2.6. Subsurface evaluation After completion of the wear test, three specimens from each group were ground with SiC abrasive papers perpendicular to the bonded interface to reveal the subsurface region beneath the center of the wear scars. When reaching near the center of the tracks, the surfaces were ground with SiC papers in a sequence of decreasing abrasive sizes (P280, P800, P1200, P2400, and P4000 mesh number) and polished sequentially using DiaDuo-2 (Struers, Copenhagen, Denmark) for 5 min 3
Elastic modulus (GPa)
Poisson's ratio
References
Enamel Adhesive (U200)
84.1 10.76
0.30 0.30
Ceramic (emax) Resin composite (TNB)
102.7 12.1
0.215 0.30
Selma et al. (2016) (Takenaka et al., 2015; Tolidis et al., 2012) Wendler et al. (2017) Ausiello et al. (2017)
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 2. A comparison of typical 3D topography and vertical profiles for the ceramic and composite groups after different testing increments. A: 3D topography of wear tracks after 5×102 cycles; B-D: vertical profiles of wear tracks after 5×102, 5×103 and 5×104 cycles, respectively. The black and red solid lines in figure A indicate where the vertical profile is obtained; the red dotted lines in figure B-D indicate the bonded interface.
differences in the wear track morphologies between the two groups that reflected the mechanisms of degradation. After 5×102 cycles of sliding contact (Fig. 4), the wear tracks were wider in the enamel side than the inlay material, and the enamel was covered by wear debris (triangles). Besides some similarity in features of the groups restored with ceramic and resin composite, there were also some distinct differences. For example, cracks were evident in the ceramic surface (Fig. 4A2), especially in the immediate vicinity of the bonded interface (arrows), which partly extended to the boundary. While the edge of the ceramic inlay adjacent to the bonded interface was damaged and exhibited chipping (circles), the enamel-adhesive interface appeared intact. For the resin composite inlay shown in Fig. 4B, the width of the wear scar at the bonded interface was wider than nearby in the enamel or inlay, suggesting a scuffing motion at the interface. Characteristics of the tooth enamel adjacent to the bonded interface were different from the surrounding area in the wear track. Specifically, chipping and cracks (arrows) were apparent in the adjacent tooth enamel, and the adhesive bond integrity between the tooth enamel and adhesive appears degraded. In comparison, abrasive wear of the resin composite was evident, but the bonded interface appeared largely intact with no signs of damage (Fig. 4B1). With an increase in to 5×103 cycles, the wear scar on the side of the inlay material broadened and exceeded that on the enamel side (Fig. 5). For the ceramic inlay, the wear tracks exhibited shallow plough marks on the surface and a small amount of wear debris was evident on the enamel surface (Fig. 5A). In comparison to features evident after 5×102 cycles, the bonded interface appeared to be intact, with minimal evidence of damage except for a few cracks concentrated in the ceramic just adjacent to the interface (Fig. 5A2). In the resin composite group, a mass of wear debris covered the majority of the enamel surface and obscured details of the microstructure. The surface of the resin composite showed evidence of exfoliated particles (Fig. 5B). The extent of damage in the enamel also appeared substantially lower than at 5×102 cycles, and the bonded interface exhibited higher integrity overall (Fig. 5B2). As the number of cycles increased to 5×104 cycles, the wear scars evolved in size with growth preferentially on the side of the inlay material (Fig. 6). In fact, the wear scar width on the inlay side exceeded that on the enamel side, regardless of the inlay material. For the ceramic group, deep and dense plough marks were noted running
profiles for the two material groups exhibited unique features at specific stages of the testing evaluated. At the initial stage of contact wear (5×102 cycles), wear of the enamel exceeded that of the inlay, regardless of the material (Fig. 2B). The vertical profiles at this stage of the wear history exhibited a discontinuity in the vicinity of the bonded interface. As evident from the comparison in Fig. 2B, wear in the resin composite inlay appeared more extensive than in the ceramic throughout the wear scar, but also in the adjacent tooth enamel. With continuation of the sliding contact to the next interval (5×103 cycles), the wear tracks appeared relatively smooth, with no abrupt discontinuities, implying the wear response was approaching steady state (Fig. 2C). With further progression of the sliding contact, wear of the inlay materials increased rapidly. In the tertiary period of the loading history (5×104 cycles), the wear track profiles showed that the wear depth of the inlay material exceeded that of the tooth enamel (Fig. 2D). A comparison of the wear track depths at the three intervals of the wear testing is shown in Fig. 3. Specifically, the depth of wear in the enamel, adhesive and inlay material are compared after 5×102 cycles, 5×103 cycles and 5×104 cycles respectively. In the early stage of sliding contact (5×102 cycles), the wear depth was location dependent. In the ceramic group the depth was ordered in decreasing magnitude from the adhesive > adjacent enamel > adjacent ceramic. In the resin composite group, the depth was ordered in decreasing magnitude from the enamel > adhesive > nearby resin composite. The wear was significantly greater in the enamel, adhesive and inlay material for the tooth specimens restored with resin composite inlay. However, with evolution in the cyclic loading process, there was no significant difference in the extent of wear between two groups or among different testing locations in each group. With further progression in cyclic contact the inlay materials adjacent to the interface exhibited the highest wear depth, followed by the adhesive and nearby enamel. Interestingly, after 5×104 cycles there was no significant difference in the extent of wear in the bonded interface between the groups restored with ceramic and resin composite inlays. 3.2. Wear morphology Micrographs that document the morphology of typical wear tracks after the three different periods of sliding contact are shown in Figs. 4–6 after the three intervals of sliding contact. Overall, there were 4
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 3. A comparison of wear depths for the ceramic and composite groups at the three intervals of the wear testing. A: 5×102 cycles; B: 5×103 cycles; C: 5×104 cycles. *,**,*** The same symbol denotes that there is no significant difference in wear depth between tooth enamel, adhesive and inlay material for each material group; # indicates there are significant difference in wear depth between the two material groups.
are shown in Fig. 7. As evident from these images, the enamel-adhesive and inlay-adhesive interfaces appeared intact in both groups. For the ceramic inlays, a few microcracks were found within the ceramic adjacent to the bonded interface (Fig. 7A). For the composite group, cracks were clearly observed within the enamel and extended adjacent to the bonded interface (Fig. 7B).
parallel to the sliding direction in both the enamel and ceramic surface (Fig. 6A). Furthermore, the enamel-adhesive and adhesive-ceramic interfaces appeared intact, although obvious cracks were identified (arrows) in the ceramic near the bonded interface (Fig. 6A1). In the composite group, the surface characteristics within the adhesive layer and nearby resin composite appeared consistent with those from earlier observations in the wear history; exfoliated particles were identified in the resin composite. At the bonded interface, the extent of damage in the enamel varied, with some obvious cracks (arrows) noted in the enamel adjacent to the bonded interface (Fig. 6B1). A few cracks extended to the bonded interface. Nevertheless, the interface between the adhesive and enamel/resin composite still appeared to remain intact (Fig. 6B2).
3.4. FEA results Results of the FEM were evaluated in terms of the location and the homogeneity of the stress distributions. Stress fields for the ceramic and resin composite are shown in Fig. 8A and B, respectively. Within the range of contact from the bonded interface area, it was noted that the stress distributions were highly dependent on the location of contact load. Clearly the stresses were lowest when contact was far from the bonded interface (not shown). When the contact load approached the bonded interface, a non-uniform stress distribution develops due to the
3.3. Crack propagation Subsurface damage and cracks extending beneath the wear tracks
Fig. 4. Representative SEM micrographs of the worn surfaces after 5×102 testing cycles. A: ceramic group, B: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. The white dotted square outlines the bonded interface in the wear scar and is shown in the right image; red triangles indicate the wear debris; red arrows indicate cracks; red circle indicates the chipping.
5
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 5. Representative SEM micrographs of the worn surfaces after 5×103 testing cycles. A: ceramic group, B: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. The white dotted square outlines the bonded interface in the wear scar and is shown in the right image; red triangles indicate the wear debris; red arrows indicate cracks.
4. Discussion
modulus mismatch, and a stress concentration is posed at the corner of the material with highest elastic modulus (Fig. 8A2 and 8B2). In comparing the stress distributions at the bonded interface area, there are distinct differences between the ceramic and resin composite inlay materials. For the ceramic inlay, a majority of the contact stress is transmitted to the inlay, which causes shielding of the adjacent tooth structure (Fig. 8A1). The point of maximum principal stress is at the edge of the inlay near the bonded interface (Fig. 8A2). In contrast, for the resin composite inlay a larger portion of the contact stress is transmitted to the enamel due to the comparatively lower elastic modulus of the resin composite. As a result, the stress in the adjacent enamel exceeds that in the resin composite (Fig. 8B1 and 8B3). The maximum principal stress in the bonded interfaces involving resin composite develops in the enamel structure adjacent to the bonded interface (Fig. 8B2).
Advances in adhesive technologies have increased indications for the use of inlay restorations in dentistry. With the escalation in aesthetic demands, the most commonly used materials for inlay restoration are ceramics and resin composites. Long-term clinical studies found that the failure of inlay restorations was primarily associated with a decrease in the bonded interface integrity (Frankenberger et al., 2008; Krämer et al., 2008). Therefore, the in vitro study was performed to compare the wear behavior and damage mechanism of bonded interfaces involving ceramic and composite inlay restorations using two clinically representative materials. The film thickness of cement is one of the critical factors essential to the clinical success of fixed prostheses and an adequate internal adaptation of a restoration is considered as a decisive factor for longevity. Threshold values for internal fit dimensions have not been determined (Guess et al., 2014). The International Standard for Water-based cement (ISO9917-1, 2007) defines the critical thickness of the cement layer is
Fig. 6. Representative SEM micrographs of the worn surfaces after 5×104 testing cycles. A: ceramic group, B: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. The white dotted square outlines the bonded interface in the wear scar and is shown in the right image; red arrows indicate cracks. 6
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 7. A vertical cross-section of the specimens beneath the wear scars after 5×104 cycles of sliding articulation. A: ceramic group, B: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. Red triangles indicate the cracks.
25 μm. It is difficult to achieve a cement thickness less than 25 μm in clinical practice due to the limitations in machining accuracy of inlay restorations. In fact, measured cement thickness values for inlay restorations varied substantially between various studies, with a range from 20 to 150 μm (Homsy et al., 2017; Pott et al., 2016; SenerYamaner et al., 2016). Some scholars have advocated that the clinically acceptable interface width limit for reducing the effect of shrinkage stress was 100 μm (Krämer et al., 2000). The interface width of all specimens in the present study were 30–80 μm, which were all within the clinically acceptable range. According to the experimental results, distinct differences in the depth and morphology of wear and damage were evident at the bonded interface area between the ceramic and resin composite inlay groups. The bonded interface typically consists of three primary parts, namely the tooth structure, adhesive and the restoration. The differences in mechanical properties between these materials influences the stress distribution of the bonded interface as well as both the type and extent of damage that develops (Xu et al., 2017). For the group restored with ceramic, the elastic modulus of the inlay was higher than the enamel and adhesive, which supports the tooth structure. At the early stages of sliding contact, the contact area between the restored tooth sample and the antagonist ball was essentially
a ball-on-flat configuration. As a consequence, the contact stress was very high. And due to its brittle nature, the ceramic possesses only moderate resistance to localized tensile and shear stresses (Kuijs et al., 2006). Thus, under the high contact stress, numerous fatigue cracks (Fig. 4A) occurred on the ceramic surface. Within the immediate vicinity of the bonded interface, the stress is not equally distributed between the enamel and ceramic. Due to its higher modulus, the ceramic inlay bears a larger portion of the contact load, and the stress in the enamel structure is lower (Fig. 8A1). When contact occurs at the bonded interface the maximum principal stress is concentrated in the ceramic (Fig. 8A2), promoting damage there preferentially. Consistent with the FEA results, the results of the experiments showed dense and long cracks in the ceramic adjacent to the bonded interface. Damage within the nearby enamel surface was limited to characteristics of abrasive wear (Fig. 4A). At the bonded interface proper, the adhesive underwent more severe wear than the nearby enamel and ceramic due to its inferior wear-resistance (Furuichi et al., 2016; Ishikiriama et al., 2015; Prakki et al., 2007). Meanwhile, without the support of the adhesive, the edge of the ceramic was exposed and exhibited chipping as a result of sliding contact (Fig. 4A2) (Nishide et al., 2011). At 5×103 cycles, the bonded interface area exhibited less damage (Fig. 5A), which could be attributed to two reasons. First, the contact
Fig. 8. The stress distributions at different distances from the enamel-inlay bonded interface. A1-A3: ceramic group, B1-B3: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. The width of the adhesive layer is 50 μm. 7
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Specifically, the resin composite inlay shed stress to the nearby enamel, which resulted in cracks and fracture. With progression of sliding contact, the dominant cracks propagated to the inner enamel region, parallel to the bonded interface (Fig. 7B). Cracks and related damage that forms under cyclic loading could leave the interface more susceptible to biofilm degradation by enabling bacteria to access the underlying dental tissue, and possibly secondary caries and tooth discoloration (Khvostenko et al., 2015). Cracks that extend more deeply in enamel could cause postoperative sensitivity and pulpitis [52], or even lead to catastrophic fracture of the tooth under higher cyclic stress via cyclic crack growth. Conventional thoughts are that postoperative sensitivity and secondary caries is attributed to microleakage at the margins of resin composites, and caused by polymerization shrinkage or the different thermal expansion coefficients between the resin composite and tooth structure (Morimoto et al., 2016). However, the results of this study suggest that secondary caries at the margins are not necessarily all associated with marginal gaps. Alternatively, cracks may develop in the enamel near the bonded interface and along the enameladhesive interface as initiated by early sliding contact and progressive growth. That is an additional contributing mechanism for the occurrence of postoperative complications. Another interesting finding was that wear of the adhesive exceeded that of the enamel and the restorations within the first 5×102 cycles. However, wear of the adhesive was intermediate (between that of the enamel and inlay) at the later stage of the evaluation. This response is quite different from previous studies, which reported that the cement undergoes the most wear in the bonded interface area (Belli et al., 2009; Takamizawa et al., 2015; Torii et al., 1999). The difference in findings could be associated with the adhesive layer thickness (Kawai et al., 1994). In this study, the adhesive thickness ranged from 30 to 80 μm. The antagonist ball was in direct contact with the adhesive at the early stage and caused rapid wear due in the adhesive to its inferior wear resistance. But with evolution in the sliding contact, the antagonist ball was prevented from causing further abrasion of the adhesive by the adjacent enamel and inlay material. In previous studies the interface gap was as much as 300 μm (Belli et al., 2009; Takamizawa et al., 2015; Torii et al., 1999). In that scenario, the adhesive could undergo more substantial contact with the antagonist ball and undergo preferential wear more continuously. According to these results, a close marginal fit serves to diminish the wear of luting cements in clinical function. This study investigated and compared the sliding contact damage and wear behavior of enamel-inlay bonded interface within two types of inlay material (ceramic and resin composite). A finite element analysis enabled an evaluation of the stress distribution at the bonded interface area under contact, and more comprehensive understanding of the experimental results. The combined findings of experiments and simulations suggest that the ceramic inlay is more effective at protecting the residual dental tissue, although no significant difference was found in wear depth between the two materials at the later stage. Of course, a real oral environment that includes temperature changes, pH variations, bacterial biofilm and other challenges may also affect the wear behavior of bonded interface. While, due to the complexity and uncontrollability of oral environment, in vitro studies are easily performed and cheaper than clinical studies. Laboratory tests under very controlled conditions remain an important predictive tool for clinical performances despite their limitations (Wulfman et al., 2018). Also, aiming at the limitation, further in vivo studies are currently underway to explore the importance of these additional factors.
area between the wear track and the antagonist ball increased, which resulted in a decrease in the contact stress. And secondly, the wear tracks became relatively smooth with no abrupt discontinuities (Fig. 2B1), which made the stress distribute more uniformly at the bonded interface. With additional contact cycles, at the bonded interface, the concentrated stress in the ceramic facilitated accelerated wear relative to the nearby tooth enamel in the following stage. The difference in wear depth between the enamel and ceramic caused aggravated accumulation of stress in the ceramic structure near the bonded interface. Cracks were apparent in the ceramic immediately adjacent to the bonded interface (Fig. 6A1), which could lead to ceramic restoration fracture (Kuijs et al., 2006), and its prominence in the failure of porcelain inlays (Banditmahakun et al., 2006). In clinic practice, monitoring the bonded interface and eliminating the wear differences on both sides of the interface could help to reduce the risk of ceramic inlay restoration fractures. Through a comparison of the wear tracks morphologies after 5×103 cycles and 5×104 cycles, it appears that damage of the bonded interface is largely associated with the dissimilar abrasion resistance between tooth enamel and ceramic. Therefore, matching the wear resistance of the tooth enamel and ceramic restorative is an important consideration when choosing restorative materials. It may be a more appropriate goal in the development of “next generation” dental ceramics, rather than maximizing the toughness and wear resistance. In contrast to the ceramic inlay, the elastic modulus of the resin composite is similar to the adhesive, and significantly lower than that of tooth enamel. As a result of the low elastic modulus, a greater portion of the contact load was concentrated in the enamel structure, especially at its boundary with the bonded interface (Fig. 8B2). Consistent with the FEA results, the wear track in the resin composite at 5×102 cycles was flared at the immediate interface, due to its lower modulus and hardness (Fig. 4B). The enamel adjacent to the bonded interface was severely damaged, exhibiting cracks and brittle fracture, while the surrounding area of the wear track in the resin composite only showed exfoliated particles (Fig. 4B2). These particles and lower elastic modulus of the resin composite are responsible for the larger wear depth in the tooth enamel near the bonded interface. Owing to the inferior wearresistance, the resin composite inlays showed higher wear depth than the ceramic inlays in the early stage of sliding contact. But with the increase in sliding contact cycles to 5×103, damage of the bonded interface became less prominent and more consistent with that in the ceramic inlays (Fig. 5B1). With further increase in cycles, the wear depth of the resin composite inlays exceeded that of the enamel due to its inferior wear-resistance. In addition, cracks appeared again within the adjacent enamel structures due to the stress concentration (Fig. 6B1). It was unexpected to find that there was no difference in the wear depth of the two inlay materials after the initial wear (N > 5×102 cycles) despite the greater wear resistance of the ceramic relative to the resin composite. This could be attributed to the difference in stress distribution at the bonded interface. In the ceramic inlay, the stress accommodated by the stiff ceramic accelerates damage and wear in the nearby ceramic. That resulted in similar wear depth with that of the resin composite and greater depth of wear than in the nearby tooth enamel at the later stage of evaluation. Hence, wear of the bonded interface is not necessarily decided by the wear resistance of the inlay material, but by the difference or “mismatch” in mechanical properties of the tooth structure and restorative material. Apart from the wear responses of the two inlay groups, their damage and cracking behavior were very different during the sliding contact history. For the ceramic inlay, degradation of the bonded interface occurred mostly in the ceramic adjacent to the interface. The nearby enamel was only slightly damaged, indicating that the ceramic inlay protected the residual tooth structure. When compared with the ceramic inlay, the biggest difference of the interfaces involving the resin composite was the extent of damage in the nearby enamel.
5. Conclusions Within the limitations of this study, the following conclusions were drawn: 1. In the early stage of sliding contact (N ≤ 5×102 cycles), wear evolved rapidly at the interface, and the wear of the enamel 8
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
exceeded that of the inlay, regardless of the material. However, with evolution of the sliding contact cycles (N = 5×104 cycles), the wear depth was highest in the inlay material, and lowest for the adjacent tooth enamel. 2. At the bonded interface area, the resin composite inlays exhibited significantly higher wear depth than the ceramic inlays in the early stage, while there was no significant difference in wear depth between the two materials at the later stage of the sliding contact history. 3. The restoration material is important to the stress distribution and extent of damage within enamel-inlay bonded interfaces. For ceramic inlays, the stress is primarily borne by the ceramic structure; cracks and chipping occur in the nearby ceramic, whereas damage in the nearby enamel is limited. Conversely, resin composite inlays shed contract stress at the interface to the tooth structure; the adjacent tooth enamel undergoes cracking and brittle fracture under cyclic sliding contact, which facilitates crack propagation into the inner enamel. Compared to resin composite inlay, the ceramic inlay appears to be more effective at protecting the residual dental tissue.
fabricated with milling, 3D printing, and conventional technologies. J. Prosthet. Dent 119, 783–790. Ishikiriama, et al., 2015. Surface roughness and wear of resin cements after toothbrush abrasion. Braz. Oral Res. 29, 1–5. Jung, et al., 2000. Lifetime-limiting strength degradation from contact fatigue in dental ceramics. J. Dent. Res. 79, 722–731. Kawai, et al., 1994. Effect of gap dimension on composite resin cement wear. Quintessence Int. 25, 53–58. Khvostenko, et al., 2015. Cyclic mechanical loading promotes bacterial penetration along composite restoration marginal gaps. Dent. Mater. 31, 702–710. Kraemer, Frankenberger, 2005. Clinical performance of bonded leucite-reinforced glass ceramic inlays and onlays after eight years. Dent. Mater. 21, 262–271. Krämer, et al., 2006. Ceramic inlays bonded with two adhesives after 4 years. Dent. Mater. 22, 13–21. Krämer, et al., 2015. Marginal quality and wear of extended posterior resin composite restorations: eight-year results in vivo. J. Mech. Behav. Biomed. Mater. 50, 13–22. Krämer, et al., 2000. Adhesive luting of indirect restorations. Am. J. Dent. 13, 60D–70D. Krämer, et al., 2008. Totally bonded ceramic inlays and onlays after eight years. J. Adhesive Dent. 10, 307–314. Krejci, et al., 1999. The influence of antagonist standardization on wear. J. Dent. Res. 78, 713–719. Kuijs, et al., 2006. A comparison of fatigue resistance of three materials for cusp-replacing adhesive restorations. J. Dent. 34, 19–25. Lange, Pfeiffer, 2009. Clinical evaluation of ceramic inlays compared to composite restorations. Oper. Dent. 34, 263–272. Lawson, et al., 2013. An analysis of the physiologic parameters of intraoral wear: a review. J. Phys. D Appl. Phys. 46, 632–637. Magne, 2007. Efficient 3D finite element analysis of dental restorative procedures using micro-CT data. Dent. Mater. 23, 539–548. Magne, Belser, 2003. Porcelain versus composite inlays/onlays: effects of mechanical loads on stress distribution, adhesion, and crown flexure. Int. J. Periodontics Restorative Dent. 23, 543–555. Manhart, et al., 2001. Three-year clinical evaluation of composite and ceramic inlays. Am. J. Dent. 14, 95–99. Mesquita, et al., 2006. Dynamic visco-elastic properties of dental composite resins. Dent. Mater. 22, 258–267. Morimoto, et al., 2016. Survival rate of resin and ceramic inlays, onlays, and overlays: a systematic review and meta-analysis. J. Dent. Res. 95, 985–994. Moszner, Klapdohr, 2004. Nanotechnology for dental composites. Int. J. Nanotechnol. 1, 130–156. Nishide, et al., 2011. Cracks formed by Vickers indentation adjacent to the interface in bonded dental ceramics with various marginal angles. Dent. Mater. J. 30, 308–314. Papadogiannis, et al., 2008. The effect of temperature on the viscoelastic properties of nano-hybrid composites. Dent. Mater. 24, 257–266. Pest, et al., 2006. Stress distribution in a post‐restored tooth using the three‐dimensional finite element method. J. Oral Rehabil. 33, 690–697. Peumans, et al., 2010. Eight-year clinical evaluation of a 2-step self-etch adhesive with and without selective enamel etching. Dent. Mater. 26, 1176–1184. Peumans, et al., 2013. Two-year clinical evaluation of a self-adhesive luting agent for ceramic inlays. Clin. Oral Investig. 17, 739–750. Pol, Kalk, 2011. A systematic review of ceramic inlays in posterior teeth: an update. Int. J. Prosthodont. 24, 566–575. Pott, et al., 2016. Marginal fit of indirect composite inlays using a new system for manual fabrication. Eur. J. Paediatr. Dent. 17, 223–226. Prakki, et al., 2007. Effect of toothbrushing abrasion on weight and surface roughness of pH-cycled resin cements and indirect restorative materials. Quintessence Int. 38, 544–554. Santos, et al., 2016. Clinical evaluation of ceramic inlays and onlays fabricated with two systems: 12-year follow-up. Clin. Oral Investig. 20, 1683–1690. Selma, et al., 2016. Biomechanics of cervical tooth region and noncarious cervical lesions of different morphology; three-dimensional finite element analysis. Eur. J. Dermatol. 10, 413–418. Sener-Yamaner, et al., 2016. Effect of material and fabrication technique on marginal fit and fracture resistance of adhesively luted inlays made of CAD/CAM ceramics and hybrid materials. J. Adhes. Sci. Technol. 31, 55–70. Shor, et al., 2003. Fatigue load of teeth restored with bonded direct composite and indirect ceramic inlays in MOD class II cavity preparations. Int. J. Prosthodont. 16, 64–69. Soares, et al., 2004. Fracture resistance of teeth restored with indirect-composite and ceramic inlay systems. Quintessence Int. 35, 281–286. St-Georges, et al., 2003. Fracture resistance of prepared teeth restored with bonded inlay restorations. J. Prosthet. Dent 89, 551–557. Takamizawa, et al., 2015. Simulated wear of self-adhesive resin cements. Oper. Dent. 41, 327–338. Takenaka, et al., 2015. Ultrasonic measurement of the effects of light irradiation and presence of water on the polymerization of self‐adhesive resin cement. Eur. J. Oral Sci. 123, 369–374. TE, 2013. Evaluation of fracture resistance and failure risks of posterior partial coverage restorations. J. Esthetic Restor. Dent. 25, 110–122. Thordrup, et al., 2001. A 5-year clinical study of indirect and direct resin composite and ceramic inlays. Quintessence Int. 32, 199–205. Tolidis, et al., 2012. Dynamic and static mechanical analysis of resin luting cements. J. Mech. Behav. Biomed. Mater. 6, 1–8. Torii, et al., 1999. Influence of filler content and gap dimension on wear resistance of resin composite luting cements around a CAD/CAM ceramic inlay restoration. Dent. Mater. J. 18, 453–461.
Declarations of interest None. Acknowledgments The authors gratefully acknowledge the financial support provided for this study by Sichuan Province Science and Technology Support Program (2017JY0236). The supporting organization had no role in design of the study, data collection, data analysis, decision to publish, or preparation of the manuscript. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jmbbm.2019.103430. References Ausiello, et al., 2017. CAD-FE modeling and analysis of class II restorations incorporating resin-composite, glass ionomer and glass ceramic materials. Dent. Mater. 33, 1456–1465. Ausiello, et al., 2004. Stress distributions in adhesively cemented ceramic and resincomposite Class II inlay restorations: a 3D-FEA study. Dent. Mater. 20, 862–872. Banditmahakun, et al., 2006. The effect of base materials with different elastic moduli on the fracture loads of machinable ceramic inlays. Oper. Dent. 31, 180–187. Belli, et al., 2005. Effect of hybrid layer on stress distribution in a premolar tooth restored with composite or ceramic inlay: an FEM study. J. Biomed. Mater. Res. B. 74B, 665–668. Belli, et al., 2009. In vitro wear gap formation of self-adhesive resin cements: a CLSM evaluation. J. Dent. 37, 984–993. Costa, et al., 2014. The influence of elastic modulus of inlay materials on stress distribution and fracture of premolars. Oper. Dent. 39, E160–E170. Da, et al., 2004. Fracture resistance of resin-based composite and ceramic inlays luted to sound human teeth. Am. J. Dent. 17, 404–406. Trindade, Flávia Zardo, et al., 2018. Elastic properties of lithium disilicate versus feldspathic inlays: effect on the bonding by 3D finite element analysis. J. Prosthodont. 27, 741–747. Frankenberger, et al., 2008. Leucite-reinforced glass ceramic inlays and onlays after 12 years. J. Adhesive Dent. 10, 393–398. Fron, et al., 2013. Clinical efficacy of composite versus ceramic inlays and onlays: a systematic review. Dent. Mater. 29, 1209–1218. Furuichi, et al., 2016. Mechanical properties and sliding-impact wear resistance of selfadhesive resin cements. Oper. Dent. 41, 83–92. Grivas, et al., 2014. Composite inlays: a systematic review. Eur. J. Prosthodont. Restor. Dent. 22, 117–124. Guess, et al., 2014. Marginal and internal fit of heat pressed versus CAD/CAM fabricated all-ceramic onlays after exposure to thermo-mechanical fatigue. J. Dent. 42, 199–209. Höland, et al., 2008. Ceramics as biomaterials for dental restoration. Expert Rev. Med. Devices 5, 729–745. Holberg, et al., 2013. Fracture risk of lithium-disilicate ceramic inlays: a finite element analysis. Dent. Mater. 29, 1244–1250. Homsy, et al., 2017. Marginal and internal fit of pressed lithium disilicate inlays
9
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al. Wendler, et al., 2017. Chairside CAD/CAM materials. Part 2: flexural strength testing. Dent. Mater. 33, 99–109. Wulfman, et al., 2018. Wear measurement of dental tissues and materials in clinical studies: a systematic review. Dent. Mater. 34, 825–850. Xu, et al., 2019. Wear and damage at the bonded interface between tooth enamel and resin composite. J. Dent. 83, 40–49. Xu, et al., 2017. A comparative study on the wear behavior of a polymer infiltrated
ceramic network (PICN) material and tooth enamel. Dent. Mater. 33, 1351–1361. Yamanel, et al., 2009. Effects of different ceramic and composite materials on stress distribution in inlay and onlay cavities: 3-D finite element analysis. Dent. Mater. J. 28, 661–670. Yap, et al., 2004. Wear behavior of new composite restoratives. Oper. Dent. 29, 269–274. Yu, et al., 2017. Effect of acidic agents on the wear behavior of a polymer infiltrated ceramic network (PICN) material. J. Mech. Behav. Biomed. Mater. 74, 154–163.
10
Contents lists available at ScienceDirect
Journal of the Mechanical Behavior of Biomedical Materials journal homepage: www.elsevier.com/locate/jmbbm
On the wear behavior and damage mechanism of bonded interface: Ceramic vs resin composite inlays
T
Ping Yua, Yuhuan Xionga, Peng Zhaoa, Zhou Xub, Haiyang Yua, Dwayne Arolac,d,e, Shanshan Gaoa,∗ a
State Key Laboratory of Oral Diseases, National Clinical Research Center for Oral Diseases, West China Hospital of Stomatology, Sichuan University, Chengdu, China Shanghai Putuo District Eye & Tooth Disease Control and Prevention Hospital, Shanghai, China c Department of Materials Science and Engineering, University of Washington, Seattle, WA, USA d Departments of Oral Health Sciences, School of Dentistry, University of Washington, Seattle, WA, USA e Departments of Restorative Dentistry, School of Dentistry, University of Washington, Seattle, WA, USA b
A R T I C LE I N FO
A B S T R A C T
Keywords: Inlays Bonded interface Damage mode Ceramic Resin composite Wear
Advances in adhesive technologies have increased indications for the use of inlays. Decrease in the bonded interface integrity due to wear has been cited as the main cause of its failure. However, this process of interface degradation and the influence of inlay material on damage mechanism appear to be poorly understood. Thus, we aimed to compare the wear behavior and interface damage between ceramic and resin composite inlays bonded to enamel under sliding contact and use the experimental findings to support recommendation of the appropriate inlay material. Bonded interface specimens involving tooth enamel and either ceramic or resin composite inlays were prepared and subjected to reciprocating wear tests up to 5×104 cycles. The wear track profiles and morphologies were characterized after increments of cyclic sliding contact using white light interferometry and scanning electron microscopy, respectively. Optical microscopy was used to evaluate sub-surface cracks and their propagation within the samples. A finite element analysis was used to analyze the stress distributions of the bonded interfaces. Composite inlays showed higher wear depth than the ceramic in the early stage (N ≤ 5×102 cycles), while no significant difference was found at the later stage. For ceramic inlay a greater portion of the contact load was concentrated in the ceramic structure, which facilitated cracks and chipping of the ceramic inlay, with rather minimal damage in the adjacent interface and enamel. In contrast, for the resin composite inlay there was larger stress concentrated in the adjacent enamel, which caused the development of cracks and their propagation to the inner enamel. The restoration material could contribute to the stress distribution and extent of damage within enamel-inlay bonded interfaces. A tough ceramic appears to be more effective at protecting the residual dental tissue.
1. Introduction The loss of substantial tooth structure most frequently occurs in posterior teeth as a result of primary caries, fractures and aging related degradation of existing restorations (Fron et al., 2013). With the development of improved adhesive technologies, inlays have been the preferred treatment option in such cases due to the greater emphasis on preservation of tooth structure (Holberg et al., 2013). Various types of restorative materials are available for use as inlays, yet ceramics and resin composites are presently the most widely used due to their excellent aesthetic performance. In comparing these two, dental ceramics exhibit superior mechanical properties and wear resistance, while resin
composite restorations are less brittle and more “user-friendly” (Höland et al., 2008; Moszner and Klapdohr, 2004; Papadogiannis et al., 2008; Yap et al., 2004). According to the modified US Public Health Service (USPHS) criteria, besides ideal anatomical form, good color match and smooth surface texture, a successful restoration should have perfect marginal integrity (Santos et al., 2016). While clinical evaluations have commented that an obvious deterioration in marginal adaptation was widely observed in both ceramic and resin composite inlays, especially in the occlusal margins (Kraemer and Frankenberger, 2005; Krämer et al., 2006, 2015; Peumans et al., 2010, 2013). For instance, Peumans et al. reported only 21.7% of the ceramic inlays exhibited an excellent
∗ Corresponding author. State Key Laboratory of Oral Diseases, National Clinical Research Center for Oral Diseases, West China Hospital of Stomatology, Sichuan University, Chengdu, 610041, China. E-mail address: [email protected] (S. Gao).
https://doi.org/10.1016/j.jmbbm.2019.103430 Received 25 July 2019; Received in revised form 9 September 2019; Accepted 12 September 2019 Available online 12 September 2019 1751-6161/ © 2019 Elsevier Ltd. All rights reserved.
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
were prepared with leucite-reinforced ceramic (IPS e.max CAD, Ivoclar Vivadent, Schaan, Liechtenstein) or resin composite (Tetric N Ceram Bulk Fill, Ivoclar Vivadent, Liechtenstein). The preparations involved use of 4.9% hydrofluoric acid gel (IPS ceramic etching gel, Ivoclar Vivadent, Schaan, Liechtenstein), 35% phosphoric acid gel (Eco etch, Ivoclar Vivadent, Liechtenstein) and self-adhesive resin cement (Rely X U200, 3M, America).
marginal adaptation compared to 70.7% at baseline after 24 months (Peumans et al., 2013). The percentage of marginal deficiencies increased to 98% for ceramic restorations and between 84% and 92% for composite restorations after only 8 years of clinical use (Kraemer and Frankenberger, 2005; Peumans et al., 2010). Degradation of the bonded interface integrity gradually increased with the extended time, which could cascade into various complications, such as marginal discoloration, secondary caries, dentin sensitivity, pulpitis, and even fracture of the restoration (Santos et al., 2016). Therefore, the interface integrity could be considered a foundational requirement for long-term effectiveness of inlay restorations. Conversely, an accelerated decrease in integrity is one of the primary contributions to restoration failures. Analyzing the bonded interface degradation process under simulated chewing forces could help to predict the longevity and clinical performance of restorations. Mechanical strength of inlay restorations is important for their durability in clinic applications. The differences in properties between ceramic and resin composite may result in different wear behavior of inlay-restored teeth as they would influence the stress concentration the bonded interface between the tooth and restoration (Flávia Zardo Trindade et al., 2018; Thordrup et al., 2001). To date, various investigations in comparing ceramic and resin composite inlays have been reported, including both clinical and laboratory studies. The results of clinical studies are inconsistent; there is no definite conclusions on which material performs better and has the best chance for long-term survival (Fron et al., 2013; Grivas et al., 2014; Lange and Pfeiffer, 2009; Morimoto et al., 2016; Pol and Kalk, 2011). In vitro tests have largely been focused on the effects of the restorative material on the stress distribution in tooth structures and fracture strength. Similar to a comparison of findings from the clinical studies, the results are sometimes contradicting. According to some authors, teeth restored with ceramic inlays achieve higher fracture resistance than those restored with composite inlays (Belli et al., 2005; Costa et al., 2014; Magne, 2007; Magne and Belser, 2003; Manhart et al., 2001; Mesquita et al., 2006; TE, 2013; Yamanel et al., 2009), while others have reported that glass-ceramic inlay materials elevates the risk of fracture (Ausiello et al., 2004; Pest et al., 2006; Soares et al., 2004; St-Georges et al., 2003). Yet, there are also reports that the fracture resistance of teeth restored with ceramic and resin composite inlays is similar (Da et al., 2004; Shor, 2003). The difference of mechanical strength may be related to the inlay material, bonding operation, test method, etc. However, fracture strength test could reflect the overall strength of restored teeth, rather than the resistance of inlay bonded interface to masticatory stress, which exhibit limited significance in the clinical prediction of inlay restorations. Hence, despite the significance of this issue, the wear behavior and damage mechanism of inlay bonded interface are still unclear, let alone the difference between ceramic and resin composite inlays. Therefore, the process of damage initiation and evolution degradation behavior and the damage mechanisms of enamel-inlay (ceramic and resin composite) bonded interfaces were investigated and understood in this study. Furthermore, stress propagation and potential crack growth inside adhesively restored teeth under occlusal loading have been simulated. The aim of the study was to develop an evidencebased understanding of the evolution in enamel-inlay bonded interface integrity that could help improve the inlay bonded interface integrity and support an informed choice of inlay materials.
2.2. Specimen preparation Thirty bonded-interface specimens were prepared for each inlay material according to the following procedures. The teeth were inspected to exclude those with caries, cracks and fluoric mottle. The crowns were separated from the roots and then cut into sections along the mesial-distal axis with a diamond-coated band saw (Struers Minitom; Struers, Copenhagen, Denmark) under continuous water coolant. A total of sixty tooth sections were prepared. The leucite-reinforced ceramic (IPS e.max CAD) was cut to sections of 6×6×4 mm3 and sintered according to the manufacturer's protocol. At the same time, the resin composite (Tetric N Ceram Bulk Fill) was applied incrementally to fill the mold cavity between two glass plates and lightcured for 60 s at a time with an LED-type light source (Bluephase, 800 mW/cm2, Ivoclar Vivadent). After 24 h, the resin composite stripes were released from the mold and cut to sections with cross section of 6×6 mm2 and length of approximately 4 mm. Thirty ceramic blocks and thirty composite blocks were fabricated in total. The sixty tooth blocks were randomly divided into two groups, and then bonded to ceramic or resin composite sections. The interior faces of the tooth sections and one of the 6×4 mm2 ceramic/composite surfaces were ground with #280 mesh silicon carbide (SiC) abrasive papers (Struers, Copenhagen, Denmark) under continuous water irrigation to prepare the surfaces for bonding. After cleaning with alcohol, the abraded surfaces of the tooth enamel were etched with the 35% phosphoric acid gel (30 s for enamel, 15 s for dentine), rinsed for 10 s and air dried. The abraded surfaces of ceramic blocks were etched with 4.9% hydrofluoric acid gel for 20 s, rinsed for 30 s with water spray and dried with oil-free air. The RelyX U200 resin cement was then mixed following the manufacturer's instructions and applied to the entire ceramic or composite surface. Subsequently, the inner surface of the tooth section was positioned on the resin cement (Fig. 1A), taking care that the occlusal surface of the tooth section was level with the face of the ceramic/ composite block. After removing excess adhesive, light curing was then performed for 60 s with the LED-type light source. To control the resin cement thickness, indicator papers with a thickness of approximately 50 μm were utilized. The sixty bonded specimens were stored in artificial saliva for 24 h at 37°C and then embedded in an epoxy resin (Struers, Copenhagen, Denmark) with the occlusal surface facing outward (Fig. 1A). After solidification, the exposed surface of each specimen was ground with SiC papers in a sequence of decreasing abrasive sizes (P280, P800, P1200, P2400, and P4000 mesh number) under continuous water irrigation. Then, all of the specimens were highly polished sequentially using DiaDuo-2 (Struers, Copenhagen, Denmark) with 3 μm particle size for 5 min, followed by OP-S NonDry (Struers, Copenhagen, Denmark) with particle size of 0.04 μm for 5 min. The polishing was performed on a dedicated instrument (Struers, Copenhagen, Denmark) with felt cloths (Dac, Struers, Denmark) under constant water lubrication. Finally, the specimens were ultrasonically cleaned for 5 min in deionized water (KQ-50B, Shumei, Kunshan, China) and inspected under an optical microscope (BX51RF, Olympus, Tokyo, Japan) to measure the interface width at magnification of 20×. The width of the interfaces was all between 30 μm and 80 μm (Fig. 1B).
2. Materials and methods 2.1. Materials The bonded interface specimens were prepared using freshly extracted human third molars (donor age: 18 years to 25 years). Both the collection and the experimental procedures involving the teeth were approved by the Research Ethics Committee. Model inlay preparations
2.3. Wear tests Reciprocating wear tests were conducted on occlusal surface of the 2
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 1. Details of the specimen configuration. A: schematic diagram of the specimen preparation (section view), B: optical microscope view of the bonded interface specimen surface. The d in figure B indicates the width of bonded interface. The red two-way arrow in figure B shows the position of the reciprocating wear test.
and OP-S NonDry for 5 min as previously described under constant lubrication (Xu et al., 2019; Yu et al., 2017). Finally, the specimens were ultrasonically cleaned for 5 min in deionized water. Then all the specimens were examined by optical microscopy to observe subsurface damage and cracks extending beneath the worn surface.
bonded interface specimens (Fig. 1B) using a commercial machine (Rtec instruments inc, MFT5000, USA), with a ball-on-flat contact mode. Silicon nitride balls with a diameter of 6.35 mm were used as the antagonist (Jung et al., 2000; Krejci et al., 1999). All tests were performed at room temperature under simulated artificial saliva lubrication. The testing parameters included an imposed contact load of 20 N, a reciprocating amplitude of 2 mm (Lawson et al., 2013) and a frequency of 1 Hz. The displacement midpoint of the articulations was located at the bonded interface. A total of 5×104 cycles were performed and three intervals of analysis were employed over the whole cycle in order to analyze the evolution of interface degradation, including after 5×102, 5×103 and 5×104 cycles (Xu et al., 2019; Yu et al., 2017). Ten samples were used to evaluate the wear characteristics for each interval (3x10 = 30 specimens per group). The wear scars were scanned using a white light interferometer operating in the vertical scanning mode equipped with the tribometer. From each scan obtained, the wear depths of the bonded interface area were calculated, including adhesive, nearby enamel and inlay material. The locations of measurement for the enamel and inlay material were 30 μm in distance from the adhesive border. Then, three-dimensional topography maps of the wear tracks were reconstructed using the commercial software (Gwyddion 2.30, Department of Nanometrology, Czech Metrology Institute), which accompanies the interferometer.
2.7. Finite element analysis In order to analyze the differences in stress distribution caused by the different inlay materials, two-dimensional finite element models consisting of tooth enamel, adhesive and inlay (ceramic and resin composite) were developed using a stress analysis software ANSYS 10.0 (ANSYS, Inc., Canonsburg, USA). All of the materials were treated as isotropic and homogeneous. The elastic modulus and Poisson's ratio of the meshed models are listed in Table 1 (Ausiello et al., 2017; Selma et al., 2016; Takenaka et al., 2015; Tolidis et al., 2012; Wendler et al., 2017). The interfaces between different parts of the models were defined as perfectly bonded and nodal displacements at the bottom of the models were constrained in all directions. The width of the adhesive interface was set as 50 μm. A vertical load of 20 N was applied at different locations of the occlusal surface of the bonded interface model through a rigid cambered contact surface. This contact surface was defined with radius of curvature equivalent to that of the Silicon nitride antagonist ball. The placement of the contact load was varied as a function of distance from the bonded interface, and the principal stress was plotted and analyzed.
2.4. Statistical analysis Statistical analyses were performed using the software SPSS 18.0. A one-way ANOVA and LSD Fisher test were performed to assess the difference in wear depths between the tooth enamel, inlay material and adhesive. The difference between the two groups of inlay materials was analyzed using a Student's t-test. A p-value of less than 0.05 was considered statistically significant.
3. Results 3.1. Wear behavior A comparison of typical 3D topography and vertical profiles of wear tracks is presented in Fig. 2 after the three different periods sliding contact. Profiles of the wear tracks were obtained from near the center of the tracks and perpendicular to the bonded interfaces along the path of articulation (Fig. 2A). As evident from this figure, the wear depth and wear volume increased with number of testing cycles. The wear track
2.5. Surface morphology In order to analyze the overall progression of wear and damage mechanisms, representative specimens from each group were ultrasonically cleaned and examined using a scanning electron microscope (SEM, INSPECTE, Czech, Republic) after 5×102, 5×103, and 5×104 cycles of testing.
Table 1 Elastic modulus and Poisson's ratio that were used in the finite element simulations.
2.6. Subsurface evaluation After completion of the wear test, three specimens from each group were ground with SiC abrasive papers perpendicular to the bonded interface to reveal the subsurface region beneath the center of the wear scars. When reaching near the center of the tracks, the surfaces were ground with SiC papers in a sequence of decreasing abrasive sizes (P280, P800, P1200, P2400, and P4000 mesh number) and polished sequentially using DiaDuo-2 (Struers, Copenhagen, Denmark) for 5 min 3
Elastic modulus (GPa)
Poisson's ratio
References
Enamel Adhesive (U200)
84.1 10.76
0.30 0.30
Ceramic (emax) Resin composite (TNB)
102.7 12.1
0.215 0.30
Selma et al. (2016) (Takenaka et al., 2015; Tolidis et al., 2012) Wendler et al. (2017) Ausiello et al. (2017)
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 2. A comparison of typical 3D topography and vertical profiles for the ceramic and composite groups after different testing increments. A: 3D topography of wear tracks after 5×102 cycles; B-D: vertical profiles of wear tracks after 5×102, 5×103 and 5×104 cycles, respectively. The black and red solid lines in figure A indicate where the vertical profile is obtained; the red dotted lines in figure B-D indicate the bonded interface.
differences in the wear track morphologies between the two groups that reflected the mechanisms of degradation. After 5×102 cycles of sliding contact (Fig. 4), the wear tracks were wider in the enamel side than the inlay material, and the enamel was covered by wear debris (triangles). Besides some similarity in features of the groups restored with ceramic and resin composite, there were also some distinct differences. For example, cracks were evident in the ceramic surface (Fig. 4A2), especially in the immediate vicinity of the bonded interface (arrows), which partly extended to the boundary. While the edge of the ceramic inlay adjacent to the bonded interface was damaged and exhibited chipping (circles), the enamel-adhesive interface appeared intact. For the resin composite inlay shown in Fig. 4B, the width of the wear scar at the bonded interface was wider than nearby in the enamel or inlay, suggesting a scuffing motion at the interface. Characteristics of the tooth enamel adjacent to the bonded interface were different from the surrounding area in the wear track. Specifically, chipping and cracks (arrows) were apparent in the adjacent tooth enamel, and the adhesive bond integrity between the tooth enamel and adhesive appears degraded. In comparison, abrasive wear of the resin composite was evident, but the bonded interface appeared largely intact with no signs of damage (Fig. 4B1). With an increase in to 5×103 cycles, the wear scar on the side of the inlay material broadened and exceeded that on the enamel side (Fig. 5). For the ceramic inlay, the wear tracks exhibited shallow plough marks on the surface and a small amount of wear debris was evident on the enamel surface (Fig. 5A). In comparison to features evident after 5×102 cycles, the bonded interface appeared to be intact, with minimal evidence of damage except for a few cracks concentrated in the ceramic just adjacent to the interface (Fig. 5A2). In the resin composite group, a mass of wear debris covered the majority of the enamel surface and obscured details of the microstructure. The surface of the resin composite showed evidence of exfoliated particles (Fig. 5B). The extent of damage in the enamel also appeared substantially lower than at 5×102 cycles, and the bonded interface exhibited higher integrity overall (Fig. 5B2). As the number of cycles increased to 5×104 cycles, the wear scars evolved in size with growth preferentially on the side of the inlay material (Fig. 6). In fact, the wear scar width on the inlay side exceeded that on the enamel side, regardless of the inlay material. For the ceramic group, deep and dense plough marks were noted running
profiles for the two material groups exhibited unique features at specific stages of the testing evaluated. At the initial stage of contact wear (5×102 cycles), wear of the enamel exceeded that of the inlay, regardless of the material (Fig. 2B). The vertical profiles at this stage of the wear history exhibited a discontinuity in the vicinity of the bonded interface. As evident from the comparison in Fig. 2B, wear in the resin composite inlay appeared more extensive than in the ceramic throughout the wear scar, but also in the adjacent tooth enamel. With continuation of the sliding contact to the next interval (5×103 cycles), the wear tracks appeared relatively smooth, with no abrupt discontinuities, implying the wear response was approaching steady state (Fig. 2C). With further progression of the sliding contact, wear of the inlay materials increased rapidly. In the tertiary period of the loading history (5×104 cycles), the wear track profiles showed that the wear depth of the inlay material exceeded that of the tooth enamel (Fig. 2D). A comparison of the wear track depths at the three intervals of the wear testing is shown in Fig. 3. Specifically, the depth of wear in the enamel, adhesive and inlay material are compared after 5×102 cycles, 5×103 cycles and 5×104 cycles respectively. In the early stage of sliding contact (5×102 cycles), the wear depth was location dependent. In the ceramic group the depth was ordered in decreasing magnitude from the adhesive > adjacent enamel > adjacent ceramic. In the resin composite group, the depth was ordered in decreasing magnitude from the enamel > adhesive > nearby resin composite. The wear was significantly greater in the enamel, adhesive and inlay material for the tooth specimens restored with resin composite inlay. However, with evolution in the cyclic loading process, there was no significant difference in the extent of wear between two groups or among different testing locations in each group. With further progression in cyclic contact the inlay materials adjacent to the interface exhibited the highest wear depth, followed by the adhesive and nearby enamel. Interestingly, after 5×104 cycles there was no significant difference in the extent of wear in the bonded interface between the groups restored with ceramic and resin composite inlays. 3.2. Wear morphology Micrographs that document the morphology of typical wear tracks after the three different periods of sliding contact are shown in Figs. 4–6 after the three intervals of sliding contact. Overall, there were 4
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 3. A comparison of wear depths for the ceramic and composite groups at the three intervals of the wear testing. A: 5×102 cycles; B: 5×103 cycles; C: 5×104 cycles. *,**,*** The same symbol denotes that there is no significant difference in wear depth between tooth enamel, adhesive and inlay material for each material group; # indicates there are significant difference in wear depth between the two material groups.
are shown in Fig. 7. As evident from these images, the enamel-adhesive and inlay-adhesive interfaces appeared intact in both groups. For the ceramic inlays, a few microcracks were found within the ceramic adjacent to the bonded interface (Fig. 7A). For the composite group, cracks were clearly observed within the enamel and extended adjacent to the bonded interface (Fig. 7B).
parallel to the sliding direction in both the enamel and ceramic surface (Fig. 6A). Furthermore, the enamel-adhesive and adhesive-ceramic interfaces appeared intact, although obvious cracks were identified (arrows) in the ceramic near the bonded interface (Fig. 6A1). In the composite group, the surface characteristics within the adhesive layer and nearby resin composite appeared consistent with those from earlier observations in the wear history; exfoliated particles were identified in the resin composite. At the bonded interface, the extent of damage in the enamel varied, with some obvious cracks (arrows) noted in the enamel adjacent to the bonded interface (Fig. 6B1). A few cracks extended to the bonded interface. Nevertheless, the interface between the adhesive and enamel/resin composite still appeared to remain intact (Fig. 6B2).
3.4. FEA results Results of the FEM were evaluated in terms of the location and the homogeneity of the stress distributions. Stress fields for the ceramic and resin composite are shown in Fig. 8A and B, respectively. Within the range of contact from the bonded interface area, it was noted that the stress distributions were highly dependent on the location of contact load. Clearly the stresses were lowest when contact was far from the bonded interface (not shown). When the contact load approached the bonded interface, a non-uniform stress distribution develops due to the
3.3. Crack propagation Subsurface damage and cracks extending beneath the wear tracks
Fig. 4. Representative SEM micrographs of the worn surfaces after 5×102 testing cycles. A: ceramic group, B: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. The white dotted square outlines the bonded interface in the wear scar and is shown in the right image; red triangles indicate the wear debris; red arrows indicate cracks; red circle indicates the chipping.
5
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 5. Representative SEM micrographs of the worn surfaces after 5×103 testing cycles. A: ceramic group, B: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. The white dotted square outlines the bonded interface in the wear scar and is shown in the right image; red triangles indicate the wear debris; red arrows indicate cracks.
4. Discussion
modulus mismatch, and a stress concentration is posed at the corner of the material with highest elastic modulus (Fig. 8A2 and 8B2). In comparing the stress distributions at the bonded interface area, there are distinct differences between the ceramic and resin composite inlay materials. For the ceramic inlay, a majority of the contact stress is transmitted to the inlay, which causes shielding of the adjacent tooth structure (Fig. 8A1). The point of maximum principal stress is at the edge of the inlay near the bonded interface (Fig. 8A2). In contrast, for the resin composite inlay a larger portion of the contact stress is transmitted to the enamel due to the comparatively lower elastic modulus of the resin composite. As a result, the stress in the adjacent enamel exceeds that in the resin composite (Fig. 8B1 and 8B3). The maximum principal stress in the bonded interfaces involving resin composite develops in the enamel structure adjacent to the bonded interface (Fig. 8B2).
Advances in adhesive technologies have increased indications for the use of inlay restorations in dentistry. With the escalation in aesthetic demands, the most commonly used materials for inlay restoration are ceramics and resin composites. Long-term clinical studies found that the failure of inlay restorations was primarily associated with a decrease in the bonded interface integrity (Frankenberger et al., 2008; Krämer et al., 2008). Therefore, the in vitro study was performed to compare the wear behavior and damage mechanism of bonded interfaces involving ceramic and composite inlay restorations using two clinically representative materials. The film thickness of cement is one of the critical factors essential to the clinical success of fixed prostheses and an adequate internal adaptation of a restoration is considered as a decisive factor for longevity. Threshold values for internal fit dimensions have not been determined (Guess et al., 2014). The International Standard for Water-based cement (ISO9917-1, 2007) defines the critical thickness of the cement layer is
Fig. 6. Representative SEM micrographs of the worn surfaces after 5×104 testing cycles. A: ceramic group, B: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. The white dotted square outlines the bonded interface in the wear scar and is shown in the right image; red arrows indicate cracks. 6
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Fig. 7. A vertical cross-section of the specimens beneath the wear scars after 5×104 cycles of sliding articulation. A: ceramic group, B: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. Red triangles indicate the cracks.
25 μm. It is difficult to achieve a cement thickness less than 25 μm in clinical practice due to the limitations in machining accuracy of inlay restorations. In fact, measured cement thickness values for inlay restorations varied substantially between various studies, with a range from 20 to 150 μm (Homsy et al., 2017; Pott et al., 2016; SenerYamaner et al., 2016). Some scholars have advocated that the clinically acceptable interface width limit for reducing the effect of shrinkage stress was 100 μm (Krämer et al., 2000). The interface width of all specimens in the present study were 30–80 μm, which were all within the clinically acceptable range. According to the experimental results, distinct differences in the depth and morphology of wear and damage were evident at the bonded interface area between the ceramic and resin composite inlay groups. The bonded interface typically consists of three primary parts, namely the tooth structure, adhesive and the restoration. The differences in mechanical properties between these materials influences the stress distribution of the bonded interface as well as both the type and extent of damage that develops (Xu et al., 2017). For the group restored with ceramic, the elastic modulus of the inlay was higher than the enamel and adhesive, which supports the tooth structure. At the early stages of sliding contact, the contact area between the restored tooth sample and the antagonist ball was essentially
a ball-on-flat configuration. As a consequence, the contact stress was very high. And due to its brittle nature, the ceramic possesses only moderate resistance to localized tensile and shear stresses (Kuijs et al., 2006). Thus, under the high contact stress, numerous fatigue cracks (Fig. 4A) occurred on the ceramic surface. Within the immediate vicinity of the bonded interface, the stress is not equally distributed between the enamel and ceramic. Due to its higher modulus, the ceramic inlay bears a larger portion of the contact load, and the stress in the enamel structure is lower (Fig. 8A1). When contact occurs at the bonded interface the maximum principal stress is concentrated in the ceramic (Fig. 8A2), promoting damage there preferentially. Consistent with the FEA results, the results of the experiments showed dense and long cracks in the ceramic adjacent to the bonded interface. Damage within the nearby enamel surface was limited to characteristics of abrasive wear (Fig. 4A). At the bonded interface proper, the adhesive underwent more severe wear than the nearby enamel and ceramic due to its inferior wear-resistance (Furuichi et al., 2016; Ishikiriama et al., 2015; Prakki et al., 2007). Meanwhile, without the support of the adhesive, the edge of the ceramic was exposed and exhibited chipping as a result of sliding contact (Fig. 4A2) (Nishide et al., 2011). At 5×103 cycles, the bonded interface area exhibited less damage (Fig. 5A), which could be attributed to two reasons. First, the contact
Fig. 8. The stress distributions at different distances from the enamel-inlay bonded interface. A1-A3: ceramic group, B1-B3: composite group; e: enamel, a: adhesive, c: ceramic, r: resin composite. The width of the adhesive layer is 50 μm. 7
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
Specifically, the resin composite inlay shed stress to the nearby enamel, which resulted in cracks and fracture. With progression of sliding contact, the dominant cracks propagated to the inner enamel region, parallel to the bonded interface (Fig. 7B). Cracks and related damage that forms under cyclic loading could leave the interface more susceptible to biofilm degradation by enabling bacteria to access the underlying dental tissue, and possibly secondary caries and tooth discoloration (Khvostenko et al., 2015). Cracks that extend more deeply in enamel could cause postoperative sensitivity and pulpitis [52], or even lead to catastrophic fracture of the tooth under higher cyclic stress via cyclic crack growth. Conventional thoughts are that postoperative sensitivity and secondary caries is attributed to microleakage at the margins of resin composites, and caused by polymerization shrinkage or the different thermal expansion coefficients between the resin composite and tooth structure (Morimoto et al., 2016). However, the results of this study suggest that secondary caries at the margins are not necessarily all associated with marginal gaps. Alternatively, cracks may develop in the enamel near the bonded interface and along the enameladhesive interface as initiated by early sliding contact and progressive growth. That is an additional contributing mechanism for the occurrence of postoperative complications. Another interesting finding was that wear of the adhesive exceeded that of the enamel and the restorations within the first 5×102 cycles. However, wear of the adhesive was intermediate (between that of the enamel and inlay) at the later stage of the evaluation. This response is quite different from previous studies, which reported that the cement undergoes the most wear in the bonded interface area (Belli et al., 2009; Takamizawa et al., 2015; Torii et al., 1999). The difference in findings could be associated with the adhesive layer thickness (Kawai et al., 1994). In this study, the adhesive thickness ranged from 30 to 80 μm. The antagonist ball was in direct contact with the adhesive at the early stage and caused rapid wear due in the adhesive to its inferior wear resistance. But with evolution in the sliding contact, the antagonist ball was prevented from causing further abrasion of the adhesive by the adjacent enamel and inlay material. In previous studies the interface gap was as much as 300 μm (Belli et al., 2009; Takamizawa et al., 2015; Torii et al., 1999). In that scenario, the adhesive could undergo more substantial contact with the antagonist ball and undergo preferential wear more continuously. According to these results, a close marginal fit serves to diminish the wear of luting cements in clinical function. This study investigated and compared the sliding contact damage and wear behavior of enamel-inlay bonded interface within two types of inlay material (ceramic and resin composite). A finite element analysis enabled an evaluation of the stress distribution at the bonded interface area under contact, and more comprehensive understanding of the experimental results. The combined findings of experiments and simulations suggest that the ceramic inlay is more effective at protecting the residual dental tissue, although no significant difference was found in wear depth between the two materials at the later stage. Of course, a real oral environment that includes temperature changes, pH variations, bacterial biofilm and other challenges may also affect the wear behavior of bonded interface. While, due to the complexity and uncontrollability of oral environment, in vitro studies are easily performed and cheaper than clinical studies. Laboratory tests under very controlled conditions remain an important predictive tool for clinical performances despite their limitations (Wulfman et al., 2018). Also, aiming at the limitation, further in vivo studies are currently underway to explore the importance of these additional factors.
area between the wear track and the antagonist ball increased, which resulted in a decrease in the contact stress. And secondly, the wear tracks became relatively smooth with no abrupt discontinuities (Fig. 2B1), which made the stress distribute more uniformly at the bonded interface. With additional contact cycles, at the bonded interface, the concentrated stress in the ceramic facilitated accelerated wear relative to the nearby tooth enamel in the following stage. The difference in wear depth between the enamel and ceramic caused aggravated accumulation of stress in the ceramic structure near the bonded interface. Cracks were apparent in the ceramic immediately adjacent to the bonded interface (Fig. 6A1), which could lead to ceramic restoration fracture (Kuijs et al., 2006), and its prominence in the failure of porcelain inlays (Banditmahakun et al., 2006). In clinic practice, monitoring the bonded interface and eliminating the wear differences on both sides of the interface could help to reduce the risk of ceramic inlay restoration fractures. Through a comparison of the wear tracks morphologies after 5×103 cycles and 5×104 cycles, it appears that damage of the bonded interface is largely associated with the dissimilar abrasion resistance between tooth enamel and ceramic. Therefore, matching the wear resistance of the tooth enamel and ceramic restorative is an important consideration when choosing restorative materials. It may be a more appropriate goal in the development of “next generation” dental ceramics, rather than maximizing the toughness and wear resistance. In contrast to the ceramic inlay, the elastic modulus of the resin composite is similar to the adhesive, and significantly lower than that of tooth enamel. As a result of the low elastic modulus, a greater portion of the contact load was concentrated in the enamel structure, especially at its boundary with the bonded interface (Fig. 8B2). Consistent with the FEA results, the wear track in the resin composite at 5×102 cycles was flared at the immediate interface, due to its lower modulus and hardness (Fig. 4B). The enamel adjacent to the bonded interface was severely damaged, exhibiting cracks and brittle fracture, while the surrounding area of the wear track in the resin composite only showed exfoliated particles (Fig. 4B2). These particles and lower elastic modulus of the resin composite are responsible for the larger wear depth in the tooth enamel near the bonded interface. Owing to the inferior wearresistance, the resin composite inlays showed higher wear depth than the ceramic inlays in the early stage of sliding contact. But with the increase in sliding contact cycles to 5×103, damage of the bonded interface became less prominent and more consistent with that in the ceramic inlays (Fig. 5B1). With further increase in cycles, the wear depth of the resin composite inlays exceeded that of the enamel due to its inferior wear-resistance. In addition, cracks appeared again within the adjacent enamel structures due to the stress concentration (Fig. 6B1). It was unexpected to find that there was no difference in the wear depth of the two inlay materials after the initial wear (N > 5×102 cycles) despite the greater wear resistance of the ceramic relative to the resin composite. This could be attributed to the difference in stress distribution at the bonded interface. In the ceramic inlay, the stress accommodated by the stiff ceramic accelerates damage and wear in the nearby ceramic. That resulted in similar wear depth with that of the resin composite and greater depth of wear than in the nearby tooth enamel at the later stage of evaluation. Hence, wear of the bonded interface is not necessarily decided by the wear resistance of the inlay material, but by the difference or “mismatch” in mechanical properties of the tooth structure and restorative material. Apart from the wear responses of the two inlay groups, their damage and cracking behavior were very different during the sliding contact history. For the ceramic inlay, degradation of the bonded interface occurred mostly in the ceramic adjacent to the interface. The nearby enamel was only slightly damaged, indicating that the ceramic inlay protected the residual tooth structure. When compared with the ceramic inlay, the biggest difference of the interfaces involving the resin composite was the extent of damage in the nearby enamel.
5. Conclusions Within the limitations of this study, the following conclusions were drawn: 1. In the early stage of sliding contact (N ≤ 5×102 cycles), wear evolved rapidly at the interface, and the wear of the enamel 8
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al.
exceeded that of the inlay, regardless of the material. However, with evolution of the sliding contact cycles (N = 5×104 cycles), the wear depth was highest in the inlay material, and lowest for the adjacent tooth enamel. 2. At the bonded interface area, the resin composite inlays exhibited significantly higher wear depth than the ceramic inlays in the early stage, while there was no significant difference in wear depth between the two materials at the later stage of the sliding contact history. 3. The restoration material is important to the stress distribution and extent of damage within enamel-inlay bonded interfaces. For ceramic inlays, the stress is primarily borne by the ceramic structure; cracks and chipping occur in the nearby ceramic, whereas damage in the nearby enamel is limited. Conversely, resin composite inlays shed contract stress at the interface to the tooth structure; the adjacent tooth enamel undergoes cracking and brittle fracture under cyclic sliding contact, which facilitates crack propagation into the inner enamel. Compared to resin composite inlay, the ceramic inlay appears to be more effective at protecting the residual dental tissue.
fabricated with milling, 3D printing, and conventional technologies. J. Prosthet. Dent 119, 783–790. Ishikiriama, et al., 2015. Surface roughness and wear of resin cements after toothbrush abrasion. Braz. Oral Res. 29, 1–5. Jung, et al., 2000. Lifetime-limiting strength degradation from contact fatigue in dental ceramics. J. Dent. Res. 79, 722–731. Kawai, et al., 1994. Effect of gap dimension on composite resin cement wear. Quintessence Int. 25, 53–58. Khvostenko, et al., 2015. Cyclic mechanical loading promotes bacterial penetration along composite restoration marginal gaps. Dent. Mater. 31, 702–710. Kraemer, Frankenberger, 2005. Clinical performance of bonded leucite-reinforced glass ceramic inlays and onlays after eight years. Dent. Mater. 21, 262–271. Krämer, et al., 2006. Ceramic inlays bonded with two adhesives after 4 years. Dent. Mater. 22, 13–21. Krämer, et al., 2015. Marginal quality and wear of extended posterior resin composite restorations: eight-year results in vivo. J. Mech. Behav. Biomed. Mater. 50, 13–22. Krämer, et al., 2000. Adhesive luting of indirect restorations. Am. J. Dent. 13, 60D–70D. Krämer, et al., 2008. Totally bonded ceramic inlays and onlays after eight years. J. Adhesive Dent. 10, 307–314. Krejci, et al., 1999. The influence of antagonist standardization on wear. J. Dent. Res. 78, 713–719. Kuijs, et al., 2006. A comparison of fatigue resistance of three materials for cusp-replacing adhesive restorations. J. Dent. 34, 19–25. Lange, Pfeiffer, 2009. Clinical evaluation of ceramic inlays compared to composite restorations. Oper. Dent. 34, 263–272. Lawson, et al., 2013. An analysis of the physiologic parameters of intraoral wear: a review. J. Phys. D Appl. Phys. 46, 632–637. Magne, 2007. Efficient 3D finite element analysis of dental restorative procedures using micro-CT data. Dent. Mater. 23, 539–548. Magne, Belser, 2003. Porcelain versus composite inlays/onlays: effects of mechanical loads on stress distribution, adhesion, and crown flexure. Int. J. Periodontics Restorative Dent. 23, 543–555. Manhart, et al., 2001. Three-year clinical evaluation of composite and ceramic inlays. Am. J. Dent. 14, 95–99. Mesquita, et al., 2006. Dynamic visco-elastic properties of dental composite resins. Dent. Mater. 22, 258–267. Morimoto, et al., 2016. Survival rate of resin and ceramic inlays, onlays, and overlays: a systematic review and meta-analysis. J. Dent. Res. 95, 985–994. Moszner, Klapdohr, 2004. Nanotechnology for dental composites. Int. J. Nanotechnol. 1, 130–156. Nishide, et al., 2011. Cracks formed by Vickers indentation adjacent to the interface in bonded dental ceramics with various marginal angles. Dent. Mater. J. 30, 308–314. Papadogiannis, et al., 2008. The effect of temperature on the viscoelastic properties of nano-hybrid composites. Dent. Mater. 24, 257–266. Pest, et al., 2006. Stress distribution in a post‐restored tooth using the three‐dimensional finite element method. J. Oral Rehabil. 33, 690–697. Peumans, et al., 2010. Eight-year clinical evaluation of a 2-step self-etch adhesive with and without selective enamel etching. Dent. Mater. 26, 1176–1184. Peumans, et al., 2013. Two-year clinical evaluation of a self-adhesive luting agent for ceramic inlays. Clin. Oral Investig. 17, 739–750. Pol, Kalk, 2011. A systematic review of ceramic inlays in posterior teeth: an update. Int. J. Prosthodont. 24, 566–575. Pott, et al., 2016. Marginal fit of indirect composite inlays using a new system for manual fabrication. Eur. J. Paediatr. Dent. 17, 223–226. Prakki, et al., 2007. Effect of toothbrushing abrasion on weight and surface roughness of pH-cycled resin cements and indirect restorative materials. Quintessence Int. 38, 544–554. Santos, et al., 2016. Clinical evaluation of ceramic inlays and onlays fabricated with two systems: 12-year follow-up. Clin. Oral Investig. 20, 1683–1690. Selma, et al., 2016. Biomechanics of cervical tooth region and noncarious cervical lesions of different morphology; three-dimensional finite element analysis. Eur. J. Dermatol. 10, 413–418. Sener-Yamaner, et al., 2016. Effect of material and fabrication technique on marginal fit and fracture resistance of adhesively luted inlays made of CAD/CAM ceramics and hybrid materials. J. Adhes. Sci. Technol. 31, 55–70. Shor, et al., 2003. Fatigue load of teeth restored with bonded direct composite and indirect ceramic inlays in MOD class II cavity preparations. Int. J. Prosthodont. 16, 64–69. Soares, et al., 2004. Fracture resistance of teeth restored with indirect-composite and ceramic inlay systems. Quintessence Int. 35, 281–286. St-Georges, et al., 2003. Fracture resistance of prepared teeth restored with bonded inlay restorations. J. Prosthet. Dent 89, 551–557. Takamizawa, et al., 2015. Simulated wear of self-adhesive resin cements. Oper. Dent. 41, 327–338. Takenaka, et al., 2015. Ultrasonic measurement of the effects of light irradiation and presence of water on the polymerization of self‐adhesive resin cement. Eur. J. Oral Sci. 123, 369–374. TE, 2013. Evaluation of fracture resistance and failure risks of posterior partial coverage restorations. J. Esthetic Restor. Dent. 25, 110–122. Thordrup, et al., 2001. A 5-year clinical study of indirect and direct resin composite and ceramic inlays. Quintessence Int. 32, 199–205. Tolidis, et al., 2012. Dynamic and static mechanical analysis of resin luting cements. J. Mech. Behav. Biomed. Mater. 6, 1–8. Torii, et al., 1999. Influence of filler content and gap dimension on wear resistance of resin composite luting cements around a CAD/CAM ceramic inlay restoration. Dent. Mater. J. 18, 453–461.
Declarations of interest None. Acknowledgments The authors gratefully acknowledge the financial support provided for this study by Sichuan Province Science and Technology Support Program (2017JY0236). The supporting organization had no role in design of the study, data collection, data analysis, decision to publish, or preparation of the manuscript. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jmbbm.2019.103430. References Ausiello, et al., 2017. CAD-FE modeling and analysis of class II restorations incorporating resin-composite, glass ionomer and glass ceramic materials. Dent. Mater. 33, 1456–1465. Ausiello, et al., 2004. Stress distributions in adhesively cemented ceramic and resincomposite Class II inlay restorations: a 3D-FEA study. Dent. Mater. 20, 862–872. Banditmahakun, et al., 2006. The effect of base materials with different elastic moduli on the fracture loads of machinable ceramic inlays. Oper. Dent. 31, 180–187. Belli, et al., 2005. Effect of hybrid layer on stress distribution in a premolar tooth restored with composite or ceramic inlay: an FEM study. J. Biomed. Mater. Res. B. 74B, 665–668. Belli, et al., 2009. In vitro wear gap formation of self-adhesive resin cements: a CLSM evaluation. J. Dent. 37, 984–993. Costa, et al., 2014. The influence of elastic modulus of inlay materials on stress distribution and fracture of premolars. Oper. Dent. 39, E160–E170. Da, et al., 2004. Fracture resistance of resin-based composite and ceramic inlays luted to sound human teeth. Am. J. Dent. 17, 404–406. Trindade, Flávia Zardo, et al., 2018. Elastic properties of lithium disilicate versus feldspathic inlays: effect on the bonding by 3D finite element analysis. J. Prosthodont. 27, 741–747. Frankenberger, et al., 2008. Leucite-reinforced glass ceramic inlays and onlays after 12 years. J. Adhesive Dent. 10, 393–398. Fron, et al., 2013. Clinical efficacy of composite versus ceramic inlays and onlays: a systematic review. Dent. Mater. 29, 1209–1218. Furuichi, et al., 2016. Mechanical properties and sliding-impact wear resistance of selfadhesive resin cements. Oper. Dent. 41, 83–92. Grivas, et al., 2014. Composite inlays: a systematic review. Eur. J. Prosthodont. Restor. Dent. 22, 117–124. Guess, et al., 2014. Marginal and internal fit of heat pressed versus CAD/CAM fabricated all-ceramic onlays after exposure to thermo-mechanical fatigue. J. Dent. 42, 199–209. Höland, et al., 2008. Ceramics as biomaterials for dental restoration. Expert Rev. Med. Devices 5, 729–745. Holberg, et al., 2013. Fracture risk of lithium-disilicate ceramic inlays: a finite element analysis. Dent. Mater. 29, 1244–1250. Homsy, et al., 2017. Marginal and internal fit of pressed lithium disilicate inlays
9
Journal of the Mechanical Behavior of Biomedical Materials 101 (2020) 103430
P. Yu, et al. Wendler, et al., 2017. Chairside CAD/CAM materials. Part 2: flexural strength testing. Dent. Mater. 33, 99–109. Wulfman, et al., 2018. Wear measurement of dental tissues and materials in clinical studies: a systematic review. Dent. Mater. 34, 825–850. Xu, et al., 2019. Wear and damage at the bonded interface between tooth enamel and resin composite. J. Dent. 83, 40–49. Xu, et al., 2017. A comparative study on the wear behavior of a polymer infiltrated
ceramic network (PICN) material and tooth enamel. Dent. Mater. 33, 1351–1361. Yamanel, et al., 2009. Effects of different ceramic and composite materials on stress distribution in inlay and onlay cavities: 3-D finite element analysis. Dent. Mater. J. 28, 661–670. Yap, et al., 2004. Wear behavior of new composite restoratives. Oper. Dent. 29, 269–274. Yu, et al., 2017. Effect of acidic agents on the wear behavior of a polymer infiltrated ceramic network (PICN) material. J. Mech. Behav. Biomed. Mater. 74, 154–163.
10