Effects of Surface Treatments on the Bond Strength Between Resin Cement and a New Zirconia-reinforced Lithium Silicate Ceramic
This study evaluated the effects of surface treatments on the bond strength between the new zirconia-reinforced lithium silicate ceramic (ZLS) and resin cement. VITA Suprinity blocks were crystallized according to the manufacturer's instructions and randomly assigned to six groups (N=36; n=6), according to the surface treatment to be performed and aging conditions: HF20, 10% hydrofluoric acid for 20 seconds, baseline (control); HF20tc, 10% hydrofluoric acid for 20 seconds, aging; HF40, 10% hydrofluoric acid for 40 seconds, baseline; HF40tc, 10% hydrofluoric acid for 40 seconds, aging; CJ, CoJet sandblasting (25 seconds, 2.5 bar, 15-mm distance), baseline; and CJtc, CoJet sandblasting (25 seconds, 2.5 bar, 15-mm distance), aging. All specimens were silanized (Monobond S) and cemented with Panavia F to newly polymerized Z250 resin blocks. After specimens were immersed for 24 hours in distilled water at 37° C, 1-mm2 cross-section microbars were obtained by means of a cutting machine under constant cooling. Baseline groups were immediately tested, whereas “tc” groups were used to analyze the effect of aging on bond strength (10,000 thermal cycles, 5/55°C, 30-second bath). The microtensile bond strength test was performed with a universal testing machine (0.5 mm/min), and bond strength (MPa) was calculated when the load-to-failure (N) was divided by the adhesive area (mm2). We also evaluated the surface roughness (Sa, average roughness; Str, texture aspect ratio; Sdr, developed interfacial area ratio) and the contact angle resulting from the treatments. Data were statistically analyzed by one- or two-way analysis of variance and Tukey's test (all α=5%). The failure mode of each specimen was evaluated by stereomicroscopy, and representative specimens were analyzed by scanning electron microscopy. The microtensile bond strength was affected by the surface conditioning (p<0.0001), storage condition (p<0.0001), and the interaction between them (p=0.0012). The adhesion for HF etching was stable, whereas for CJ, aging significantly damaged the adhesion. Most failures were predominantly adhesive between ceramic and cement (52.6%). The roughness of the treated samples was higher compared with that of polished specimens for the three evaluated parameters (Sa, Str, and Sdr; all p<0.0001). Contact angle was also influenced by treatments (p<0.0001), with the CJ group showing values similar to those of control specimens. It can be concluded that the three surface treatment techniques present favorable immediate results, but silica coating was not effective in maintaining the bond strength over the long term.SUMMARY
INTRODUCTION
Glass ceramics have evolved over the years in their compositions and processing techniques. Leucite or lithium disilicate were the main crystal phase of the first generation of these materials, which were marketed in ingots for injection. Nowadays, the glass ceramics are mainly lithium disilicate–based pressable ingots or computer-aided design and computer-aided manufacturing (CAD-CAM) blocks. These restorations have satisfactorily served as monolithic restorations.1 Recently, a new material, zirconia-reinforced lithium silicate ceramic (ZLS), was launched under the argument that zirconia could act as a crystal phase that can reinforce the material; that is, avoid crack propagation.
The ZLS represents an attempt to unite resistance properties of polycrystalline ceramics with the esthetic excellence of the glass-ceramics in a monolithic restoration. Moreover, because a ceramic matrix is predominantly glass (from 8% to 12% zirconia), this material is considered acid sensitive and susceptible to hydrofluoric acid etching, unlike polycrystalline ceramics.2
The classification system proposed by Valandro and others2 was based on the existence of ceramic surface degradation by hydrofluoric acid (HF). Ceramics with high glass content in their composition, such as feldspar-, leucite-, and lithium disilicate–based ceramics, suffer as a result of the action of HF, resulting in a micromechanical retentive surface and thus are called acid sensitive. Ceramics based on glass infiltrated alumina or zirconia, densely sintered alumina, and yttria-tetragonal zirconia polycrystal (Y-TZP) do not degrade with HF, do not present micromechanical retention and undercuts,2 and are referred to as acid resistant.
The extension of the etching time indicated by the manufacturer for an acid-sensitive ceramic was analyzed by Zogheib and others,3 who concluded that lithium disilicate ceramic requires more than 60 seconds of HF etching for the creation of effective microretention.3 Menees and others4 found that HF etching for 20 seconds in concentrations varying from 5% and 9.5% is enough to remove the glass matrix. Despite extensive removal for 120 seconds, it was clear that the resulting etch pattern for these conditions was uniform and was not enough to affect the bending strength of the lithium disilicate ceramic.4
However, the use of HF requires careful attention due to its potential risk for the degradation of organic matter,5-7 and for this reason, other options have been investigated for ceramic surface treatment, including air abrasion with silica-coated aluminum oxide particles. During silica coating, the high energy of the shock resulting from the aluminum oxide particles is responsible for the fusion of these silica particles to the ceramic surface, making it chemically reactive to the resin cement through the silane agent and also increasing the bond strength to ceramics.8,9
Conversely, silica coating treatment is controversial, because some authors reported a decrease in mechanical strength of the material10-12 and the induction of crack propagation,13 whereas others have shown no deleterious effect on long-term mechanical behavior.11-18 There are no data in the literature on the effects of silica coating on lithium silicate reinforced by zirconia ceramic. We believe this type of surface treatment can be particularly important if the zirconia content present in in ZLS makes it less sensitive to acid conditioning.
The maintenance of bond strength in the long term is a factor that must be considered, because studies have shown that thermocycling decreases bond strength between the resin cement and glass-ceramic.9
The aim of this study was to evaluate the effects of surface treatments on the bond strength of a new ceramic lithium silicate reinforced by zirconia and resin cement. The hypothesis that extended etching time and silica coating result in stronger and longer-lasting bond strength was tested.
METHODS AND MATERIALS
The ZLS ceramic (Vita Suprinity, Vita Zahnfabrik, Bad Säckingen, Germany) was cut into blocks (12 × 8 × 7 mm) and crystallized according to the manufacturer's instructions. Blocks were embedded in acrylic resin bases, and the top face of each ceramic was polished with 800-grit sandpaper granulation under constant cooling in a polishing machine.
Ceramic Surface Treatment
The blocks (N=36) were randomly divided into six groups (n=6), and these were subjected to six experimental conditions established by the two factors under study: treatments (at three levels: HF20, HF40, and Cojet) and aging (at two levels: absence and presence).
For this, ZLS blocks were treated according to surface treatments and aging conditions: HF20, etching with 10% hydrofluoric acid (Condacporcelana, FGM, Joinville, Brazil) for 20 seconds, followed by rinsing for the same time (control protocol indicated by the manufacturer19); HF20tc, HF20 treatment + aging; HF40, etching with 10% hydrofluoric acid (Condacporcelana, FGM) for 40 seconds, followed by washing for the same time; HF40tc, HF40 treatment + aging; CJ, air-abrasion with 30 micron silica-coated alumina particles (CoJet Sand, 3M ESPE, Seefeld, Germany; 25 seconds, 2.5 bar, 15-mm distance); and CJtc, CJ treatment + aging. All specimens were silanized (Monobond S, Ivoclar Vivadent, Schaan, Liechtenstein) and cemented, one at a time, with a resin cement (Panavia F, Kuraray, Okayama, Japan) to composite resin (Filtek Z250, 3M ESPE, Irvine, CA, USA) blocks that had been made and polymerized just before cementation. Thus, six repetitions per group were obtained. Figure 1 presents a chart of specimen preparation.
Citation: Operative Dentistry 41, 3; 10.2341/14-357-L
Microtensile Bond Test
After 24 hours in distilled water at 37°C, 1-mm2 cross-section microbars composed of ceramic/cement/resin were obtained by means of a cutting machine (ISOMET 1000, Buehler, Lake Bluff, IL, USA) under constant cooling. The ends of the blocks were demarcated, and microbars from this area were excluded. The microtensile bond strength of all microbars from the same block was averaged to obtain the mean value for that block. Twenty microbars were randomly selected from each block and included in the analysis.
Microbars obtained from blocks of the baseline groups were immediately tested. Microbars from the “tc” groups were used to analyze the effect of aging on bond strength. For this, they were subjected to 10,000 thermal cycles in water at a temperature ranging between 5°C and 55°C, with 30 seconds of immersion and five seconds of transition (bath for the 521-6D cycle test, Ethik Technology, Vargem Grande, São Paulo, Brazil).
After specimen dimensions were measured with a digital caliper, specimens were attached to the testing device (OG01, Odeme, Lucerne, Brazil) with cyanoacrylate (superglue, Loctite, Lucerne, Brazil). The microtensile bond strength test was performed in a universal testing machine (DL-1000, EMIC, São José dos Pinhais, Brazil; 0.5 mm/min), and bond strength (MPa) was calculated by dividing the load-to-failure (N) by the adhesive area (mm2). A mean value for each block was calculated and used in the data analysis.
Failure Analysis
The fractured specimens were examined by stereomicroscopy (Stereo Discovery V20, Zeiss, Göttingen, Germany), and failure mode was classified as resin or resin cement cohesive, predominantly adhesive between resin cement and ceramic, mixed, or ceramic cohesive.
Roughness and Topographic Analysis
For additional analysis of the effects of surface treatments on ZLS ceramic, we obtained 2 × 1 × 1-mm specimens following the same procedures as used in the ceramic blocks for microtensile testing (cutting, polishing, and surface treatments). One sample was used as a control group and received only polishing.
The roughness of the specimens was evaluated in five samples of each group by means of a digital optical profiler (Wyko NT 1100, Veeco, Plainview, NY, USA), connected to the Wyko Vision 32 (Wyko, Veeco) at 20× magnification and 301.3 × 229.2 μm of analysis area. We obtained values (nm) for height (Sa, average roughness), spatial (Str, texture aspect ratio), and hybrid (Sdr, developed interfacial area ratio) parameters.
The contact angles of treated surfaces were also evaluated in six samples from each group before and after silanization by means of an optical tensiometer (TL 1000, Theta Lite Attention, Lichfield, Staffordshire, United Kingdom) by the sessile drop technique. For this, a syringe (#1001 Gastight Syringes, 1 mL, Hamilton, Reno, NV, USA) deposited a drop of distilled water on the sample surface. After 10 seconds of waiting for the drop to settle, a series of 30 images per second was recorded by the equipment for 20 seconds. OneAttension (Biolin Scientific, Lichfield, Staffordshire, United Kingdom) software was used for calculation of the average values of contact angle for each sample from the images obtained.
Finally, the morphology of the surface was analyzed by scanning electron microscopy. For this, the sample surface was coated with a thin layer of gold in low atmospheric pressure by means of an ion sputter-coater (SC7620 'Mini' Sputter Coater/Glow Discharge System, Emitech, East Sussex, United Kingdom), and the topography was analyzed and photographed with high-vacuum equipment (Inspect S 50, FEI Company, Brno, Czech Republic) operating at 20-25 kV, 5.0 spot, and magnifications of 500× and 5000×.
Data Analysis
Data were tabulated, and the blocks were used as experimental units for statistical analysis of the microtensile bond strength data. The results of bond strength were analyzed by two-way analysis of variance (ANOVA) (treatment and aging) and Tukey's test (both α=5%), and roughness and contact angle values were analyzed by one-way ANOVA and Tukey's test (both α=5%).
RESULTS
The microtensile bond strength results are presented in Table 1. ANOVA showed that interaction of surface and aging influenced the bond strength (p=0.0012) (Table 2). The adhesion for the HF groups was stable, whereas for CJ, aging significantly damaged the adhesion. Samples in the CJtc group also showed a large number of pretest failures after thermal cycling (Table 3).
Failure was predominantly adhesive between the ceramic and resin cement (52.6%) (Figure 2).
Citation: Operative Dentistry 41, 3; 10.2341/14-357-L
Surface roughness was also influenced by surface treatments for all parameters evaluated: Sa (p<0.0001), Str (p<0.0001), and Sdr (p<0.0001) (Figure 3). All treatments increased roughness parameters compared with those of control samples (polished, without treatment).
Citation: Operative Dentistry 41, 3; 10.2341/14-357-L
The contact angle was also influenced by surface treatments for both silanized (p<0.0001) and unsilanized samples (p<0.0001). In both cases, the group conditioned with hydrofluoric acid for 40 seconds (HF40) showed the highest wettability values (Figures 4 and 5).
Citation: Operative Dentistry 41, 3; 10.2341/14-357-L
Figure 6 shows representative profilometry images and scanning electron microscope micrographs of the surface patterns for tested groups.
Citation: Operative Dentistry 41, 3; 10.2341/14-357-L
DISCUSSION
The baseline assessment showed that the three treatments tested (HF20, HF40, and CJ) had similar baseline bond strength values.
In the context of surface treatment by extension of the etching time suggested by the manufacturer, studies have found no deleterious effect of hydrofluoric acid on the resistance of glass-ceramic,20,21 because the excessive etching reduced existing surface faults in size and depth, besides removing or stabilizing surface defects.22,23 Thus, increasing the etching time can enhance the materials strength24 but does not lead to improved bond strengths, according to our findings.
We found that the thermocycling had a negative effect on the bond strength, and samples in the silica-coated group (CJtc) showed an almost 80% decrease in the microtensile bond strength to resin cement. Most specimens in this group did not resist the thermal cycling, as well as the large percentage of predominantly adhesive failures between resin cement and ceramic showed that the surface modification by the addition of silica did not guarantee a stable bonding between ZLS and resin cement in the long term. According to Kern and Thompson,25 this is attributable to the fact that the irregularities created by the sandblasting are devoid of microretention. Corroborating this, Menees and others4 argued that for the ceramic lithium disilicate ceramic, hydrofluoric acid etching introduces more uniform and better-distributed surface changes.
In fact, silica-coated topography in our study confirms the existence of small irregularities, although the surface roughness of the three groups were not much different. In this same context, Kern and Wegner12 stated that, because of this limited ceramic roughness, chemical and physical Bis-GMA (Bisphenol A-glycidyl methacrylate)–based resin bonds are not water resistant and end up suffering detachment.
For roughness parameters, the three treatments were effective in changing the surface pattern, generating similar results. When a glass-ceramic surface is exposed to HF etching, a selective removal of its vitreous matrix occurs, exposing the crystalline structure6,26 and resulting in a topography with a total contact area greater than that of a smooth surface.27 Silica deposition generates an irregular rough surface with increased surface area that improves ceramic wettability. However, sandblasting can also induce excessive gaps in the surface of the material or even a significant loss of material.25,28
This conformation of the material's surface topography is closely linked to the characteristics that it presents in terms of contact angle and consequent wettability.29
The contact angle measurement of the dispersion liquid (often water of high purity) on a substrate is used as an indicator of total surface energy and wettability of the substrate.27,30 The smallest contact angle resulting from silica coating represents better surface wettability. However, the silica coating showed a reduction in bond strength after aging, suggesting that, although this technique ensured good initial bond strength to resin cement, the pattern generated by silica coating is not favorable for the long-term maintenance of this resistance.29,31 The lower bond strength obtained with acid etching for 40 seconds is probably due to the removal of a greater quantity of glass matrix and exposure of lithium silicate crystals and particles of zirconia creating a surface with lower wettability.
A few microbars were lost during cutting. Most microbars showed predominantly adhesive failure between ceramic and resin cement. Thus, despite the decline in bond strength of the HF40 group, the CJ group was most vulnerable to temperature changes in the water, with great loss of microbars during thermocycling.
Therefore, considering the results of microtensile bond strength, surface roughness, contact angle, and morphology obtained for surface treatments and considering the potential risks with the use of HF, it is suggested that one should not extend the time suggested by the manufacturer. The use of silica coating is also not advantageous for long-term maintenance of the bond.
CONCLUSION
On the basis of our results, it can be concluded that the silica coating was not efficient in maintaining the bond strength after aging, and etching with hydrofluoric acid for 20 or 40 seconds was equally effective in producing stable resin bonding to a ZLS ceramic.
Sketch of the sample fabrication process for microtensile bond strength test. Orange = ZLS ceramic; blue = resin cement; beige = Composite. (A): Surface treatments (HF20, HF40, or CJ) were performed on ZLS blocks that were then cemented with a resin cement to newly polymerized composite resin blocks (B). Twenty-four hours after polymerization, blocks were cut into x (C) and y (D) axes. One half of the specimens of each treatment were thermo cycled (HF20tc, HF40tc, and CJtc; n=6), whereas the other half were immediately tested (HF20, HF40, and CJ; n=6). In total, 120 specimens per group were tested using the microtensile bond strength test.
Bar graph of failure type distribution. ADHES cem/cer = predominantly adhesive failure between resin cement and ceramic; ADHES comp/cem = predominantly adhesive failure between composite resin and cement; COHES cer = cohesive failure of ceramic; COHES comp = cohesive failure of composite/resin cement; MIXED = cohesive and adhesive failures.
Bar graph of roughness values for tested groups. For each roughness parameter, means with different superscript letters are significantly different.
Bar graph of the mean values of contact angle (°) for tested groups (not silanized). Means with different superscript letters are significantly different.
Figure 5. Bar graph of the mean values of contact angle (°) for silanized specimens. Means with different superscript letters are significantly different.
Representative images of surface patterns of the tested groups. Left column: profilometry images. Center column: micrographs, 500×. Right column: micrographs, 5000×. Line A: control samples (only polished). Line B: HF20. Line C: HF40. Line D: CJ.
Contributor Notes