INTRODUCTION

Hydrofluoric acid (HF) etching followed by silanization is the most widely accepted surface treatment for glass ceramics.¹ HF etching selectively dissolves the glassy matrix of the glass-ceramic surface, creating retentive micropores, pits, cracks, and grooves on the conditioned surface, favoring micromechanical retention.^2,3 A silane coupling agent is applied on the etched surface to promote a chemical bond between the inorganic phase of the ceramic (ie, silica) with the organic phase of the resin cement (ie, polymer).^4–6

However, this has become increasingly debatable, as some studies have also shown no need for HF etching for enhancing adhesion between resin cements and glass ceramics,^7–9 and it has also been demonstrated that HF etching may even lead to a ceramic weakening effect.^10,11 This deleterious impact is explained by the fact that HF etching also creates and increases the flaw population at the ceramic bonding interface. By that, it might increase the ceramic susceptibillity to slow crack growth of critical defects under the constant masticatory stresses, since the ceramic fracture strength is inversely proportional to the largest or critical flaw present in the loaded restoration, as described by Griffith's law.¹² From the clinical standpoint, the above-mentioned weakening effect is particularly relevant when considering the evidence showing that the glass-ceramic fails from flaws on the ceramic intaglio surface.^13,14

Thus, a relevant factor to be considered as a predictor of the restoration performance seems to be the capacity of the cement to completely fill in the introduced defects,¹⁵ which has been scarcely discussed in the literature. Resin cements are indicated for adhesive luting of feldspathic ceramic restorations. They provide better adhesion (higher bond strength) to both ceramic and restorative foundations/substrates as well as superior mechanical and optical properties (greater color selection possibilities and higher color stability), high resistance to hydrolysis, and great inherent tensile strength.¹⁶ The advent of self-adhesive resin cements in the attempt to provide simplification and easy handling techniques appeared as a promising alternative approach, and these cements have shown acceptable esthetic properties and bond strength results that are at least comparable with the conventional resin cements.¹

In addition to understanding the population of defects and the fill-up potential of different cements on feldspathic ceramic restorations, another important condition involved in such a scenario is the subsequent degradation of the bonding interface. According to Lu and others,¹⁷ aging the bonding interface in water promotes a reduced elastic modulus of the cements as well as an apparent reduction in the bond strength. This degradation of cement properties may be sufficient to redistribute stresses in the restorative set, thus reducing the ability of restorations to tolerate masticatory loads over time.

Few studies have evaluated the mechanical performance of feldspathic ceramics cemented to tooth substrates or analog materials under fatigue stimuli.^18,19 According to Strasser and others,²⁰ ceramic pretreatments (eg, HF etching and air abrasion) are fundamental to provide a better and more stable adhesion (ie, reduced susceptibility to interface degradation) and consequently better performance and long-term predictability for such restorations, being that its absence may increase the risk of premature failure.

Thus, considering the above assumptions, the lack of consensus on this issue, and the clinical appeal, this laboratory study investigated the following: does HF etching change the flaw population on the glass-ceramic intaglio surface and consequently affect the fatigue behavior of adhesively bonded simplified monolithic glass-ceramic restorations when subjected to aging?

The study aimed to evaluate the influence of HF etching and aging on the fatigue failure load of a computer-aided design/computer-aided manufacturing (CAD/CAM) monolithic feldspathic ceramic adhesively bonded to a dentin analog material using two resin cements (a self-adhesive and a conventional). The assumed hypotheses were the following: 1) HF etching will not promote different fatigue behavior compared with nonetching, 2) thermocycling (Tc) will reduce the fatigue failure load results, and 3) no significant difference will be found between the cements.

METHODS AND MATERIALS

Study Design

Restorative sets were fabricated through a simplified approach³ using feldspathic ceramic discs (Vita Mark II) cemented to glass fiber reinforced epoxy resin discs (G10 substrate, dentin analog) with conventional (MA; Multilink Automix, Ivoclar Vivadent, Schaan, Liechtenstein) and self-adhesive (U200; RelyX U200, 3M ESPE, Seefeld, Germany) resin cements. The diameter of the specimens was determined as 10.0 mm because it resembles the average diameter of the first molar occlusal surface,²¹ and the final thickness of the cemented set was 3.5 mm because it is equivalent to the average thickness between the occlusal surface and the pulp chamber roof of molar teeth.^22,23 The specimens were prepared and randomly (www.random.org) distributed considering the three factors under study: ceramic etching (ie, with vs without 10% HF etching plus silane application), resin cement (ie, self-adhesive [U200] or conventional [MA]), and Tc (ie, with vs without; 5-55°C/12,000 cycles; Table 1).

Table 1 Study Experimental Design

View Fullscreen

RESULTS

Fatigue Failure Load Test: Staircase Method

Based on the three-way ANOVA, there were statistically significant influences of the factors cement (p=0.006; MA>U200), Tc (p=0.000; without Tc>with Tc), and HF etching (p=0.000; without HF>with HF) and for the interaction HF × Tc (p=0.000) on the fatigue load results. No statistically significant influence was detected for other associations.

Analyzing each cement separately, the one-way ANOVA and post hoc Bonferroni analyses showed that both cements behaved similarly to the different conditions, being that the fatigue resistance was statistically reduced only when the HF and Tc factors were applied together (Table 2; Figure 1).

Table 2 Results From the Monotonic Load-to-Failure Test (n=5)^a

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Comparing the same conditions between the cements through t-tests for independent samples, the MA cement yielded better results when the specimens were submitted to HF etching or Tc factors. There was no difference between the cements at baseline (without HF and without Tc) or for the association of factors (with HF and Tc; Table 2).

Contact Angle Measurements

The lowest contact angle was observed after HF etching plus silane application, indicating greater surface wettability. The highest contact angle (lowest wettability) was observed on the untreated surface (baseline group). Intermediate values were observed when HF etching only or silane only was used (Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

Topographic Analysis

Topographic and cross-sectioned SEM images of baseline (nonetched) and HF-etched (HF 10%; one minute) specimens are shown in Figure 3. Etched surfaces presented an irregular topography characterized by the presence of numerous micro irregularities, pits, grooves, and striations as a result of the glassy phase dissolution, known as a honeycomb-etched pattern (Figure 3A). The baseline images showed a smoother and more homogeneous surface without any relevant irregularity (Figure 3B).

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

Our findings show that the HF etching condition associated with Tc had a significant deleterious effect on the load for fatigue failure regardless of the cement used, and HF etching prior to silanization did not enhance the fatigue resistance of the simplified ceramic restorations.

HF etching promotes a surface dissolution of the ceramic glass matrix, creating micro retentions that contribute to the mechanical bonding with dental substrate when adhesively bonded.^9,10,29 However, this has become debatable, since fatigue properties of all-ceramic systems can be related to the flaw population (size, number, and distribution) of the material,¹² and some studies have demonstrated that HF etching may lead to a decrease in the feldspathic ceramics strength by introducing defects on the surface that may not be completely filled by the resin cement.^10,30,31 In our study, the first hypothesis was accepted since HF etching did not promote different fatigue failure load results compared with baseline. Regarding the Tc factor, the second hypothesis was partially accepted since it reduced the fatigue failure load only for the HF-etched specimens.

Other authors have obtained similar results, corroborating that HF etching improves the bond strength by creating micromechanical interlocking but may lead to a weakening effect on the ceramic, thereby compromising the clinical performance of the restoration.^10,31 Moreover, recent studies corroborate these findings, showing worse or equal results for bond strength,⁷ biaxial flexural strength,¹¹ and fatigue load¹⁹ when conditioning glass-ceramic with HF prior to bonding, proving that this subject still requires future evaluations and considerations. In addition, HF has potentially hazardous toxicity known from other applications and these risks should also be considered when applied in dentistry.³²

The topographic images in Figure 3C show the micro retentions, pits, and grooves created by the selective HF etching of the glass-ceramic vitreous matrix, which could lead to a decrease in the ceramic fatigue resistance, especially when the defects are not completely filled by the cement. Defects created by HF etching may also lead to stress concentration, which can result in ceramic premature fracture starting from the adhesive interface,^18,30 as observed in our study (Figure 4). This behavior is particularly important since sharp defects (as created by HF etching; Figure 3C) are more damaging than rounded defects. In this sense, it has to be emphasized that we used a high-concentration HF etchant (10% HF for one minute), which produces many more defects than in lower concentrations (for instance, 1% and 5%)¹⁹ and makes the material more susceptible to the presence of unfilled defects after bonding and, consequently, to crack initiation and propagation under fatigue loads. Clinical^33,34 and laboratory studies¹⁸ on failed glass-ceramic crowns have reported that the great majority of bulk fractures start from flaws on the ceramic-intaglio surface, where high tensile stress is concentrated.

Adhesive cementation significantly increases the restorative material's fracture loads.³⁵ Using finite element analysis and fatigue testing for the lithium disilicate glass-ceramic, de Kok and others³⁶ proved that proper adhesion can better distribute stress during loading, increasing the material's resistance. In our study, the group treated with HF (which provides a surface with many more defects; Figure 3) had a statistically similar fatigue resistance to the nonetched surface (baseline; Table 2), corroborating the results found by de Kok and others³⁶ of the protective role of bonding in overcoming the ceramic's internal roughness effect.

Regarding the resin cement filling capacity, cements with low viscosity are more prone to penetrate the ceramic than cements with high viscosity.³⁷ According to Gamal and others,³⁸ self-adhesive resin cements have high viscosity and consequently are less capable of spreading onto the ceramic and substrate surfaces, which may compromise its wettability. The application of HF etching plus silane has been recommended to promote chemical bonding between inorganic molecules of the ceramic with organic molecules of these resin cements.^39,40 This sequence is already well established in the literature in terms of bond strength, as these procedures increase the surface energy of the ceramic and the wettability of the resin cement, improving adhesion.^3,31 This greater wettability can be evidenced by the lower contact angle of the treated surfaces, as shown in Figure 2A. In our work, in terms of fatigue resistance, we can note that this procedure may not summarily be necessary, since the groups that were not submitted to HF etching had the same or even better results than the etched ones. However, it must be considered that this assertion is a finding only with respect to fatigue resistance. From this viewpoint, there is a need to develop novel surface-conditioning methods to address the problem related to the bond durability and defect fill-in potential⁶ and to try to eliminate the use of HF to produce a less technique-sensitive and safer (ie, using a less hazardous material) bonding system.

As the conventional cement performed better than the self-adhesive when the HF etching and Tc factors were applied separately, the third hypothesis was partially accepted. According to Gamal and others,³⁸ that result could be explained by the high viscosity of the self-adhesive resin cement and less spreadability onto the substrate-ceramic surface, reducing its ability to infiltrate into surface irregularities.

When the materials are free to deform, they will expand or contract due to fluctuations in temperatures. By that, the temperature change during Tc is a deleterious factor for the adhesion between ceramic and dental substrate.⁴¹ Because of the different coefficients of linear thermal expansion between the materials in the adhesive interface (ceramic/cement/substrate), they have different degrees of contraction and expansion during Tc, leading to micromechanical fatigue stresses in the adhesive interface, breaking adhesive bonds, and finally reducing the adhesion quality.⁴¹ In the present study, aging (Tc 5-55°C/12,000 times) significantly affected the fatigue resistance for both cements only when the ceramic was previously etched with HF. This corroborates the findings of Venturini and others,³¹ who hypothesized that when the micro retentions are not completely filled by the cement, this empty and unfilled space allows faster water absorption at the interface of the restoration with its consequent hydrolysis and degradation and finally a decrease in the material fatigue resistance.

The present study implemented a fatigue test under a wet environment, where constant loads with ranging intensity were applied until the failure of the specimens. This method mimics the oral environment and more closely simulates the masticatory stresses when compared with the static test.³⁵ However, the applied test setup (axial load) may not fully simulate all the forces to which the material is subjected in the oral environment, especially to lateral loads (sliding motion) that generate compressive, tensile, and shear stresses on the ceramic surface leading to the subsurface crack formation and propagation.⁴² Another limitation may be the simplified restorative set (disc-shaped specimens), which does not completely simulate the anatomy of a molar crown.

In our study, the loads at initial radial cracks (overall mean equal to 534.83 N) exceeded the maximum bite forces during mastication (148.73 to 354.01 N)⁴³ but were far below the maximum bite forces reached in sleep associated bruxism (±800 N) and maximum voluntary bite forces during daytime (±1000 N).⁴⁴ Considering the subjects discussed above, we emphasize that our results should be carefully analyzed, and more in vitro and clinical findings can corroborate our results.

CONCLUSIONS

• The HF etching associated with a silane coupling agent increased the ceramic surface free energy (lower contact angle and consequently higher wettability), but it did not provide better results in terms of fatigue resistance compared with silane agent application only.

• When the feldspathic ceramics were HF etched, bonded (regardless of the cement used), and then subjected to aging, the fatigue failure load of the restorations was significantly reduced.

• HF etching and Tc factors applied alone did not lead to degradation of the ceramic fatigue resistance, regardless of the resin cement used (conventional [MA] or self-adhesive [U200]), being that the conventional cement performed better in both cases (only HF and only Tc).