Shear bond strengths of various luting cements to zirconia ceramic: Surface chemical aspects

Shear bond strengths of various luting cements to zirconia ceramic: Surface chemical aspects

Journal of Dentistry, 2011-11-01, Volume 39, Issue 11, Pages 795-803, Copyright © 2011 Elsevier Ltd

Abstract

Objectives

To measure the shear bond strengths of various luting cements to a sandblasted zirconia ceramic and to determine the surface energy parameters of the luting cements.

Methods

Two conventional glass ionomer cements, two resin-modified glass ionomer cements, two compomer cements, and two adhesive resin cements were prepared and bonded to sandblasted zirconia (Lava). All bonded specimens were stored in water at 37 °C for 48 h and then half of them additionally thermocycled 10,000 times prior to the shear bond strength test ( n = 10). Surface roughness ( R a ) values and surface energy parameters of the eight luting cements and polished zirconia ceramic were evaluated using a profilometer and contact angle measurements, respectively ( n = 10). The bond strength and surface roughness data were statistically analysed using non-parametric and parametric procedures, respectively ( α = 0.05). Relationships between surface energy parameters and measured shear bond strengths were investigated using the Spearman rank correlation test.

Results

Panavia F 2.0 and Principle produced higher bond strengths than the other cements, with no significant changes before and after thermocycling. Fuji I, Ketac Cem Easymix, and Ionotite F yielded near-zero or zero values after thermocycling. All debonded specimens showed adhesive failure. Mean R a values ranged from 0.104 to 0.167 μm. We found the base (hydrogen bond accepting) components of the luting cements significantly affected the bond strengths both before and after thermocycling.

Conclusion

It is recommended that the surface energy parameters of luting cements be considered in evaluating their adhesive properties with zirconia ceramic.

Introduction

Modern all-ceramic dental restorations provide cosmetically good and metal-free treatment options. Amongst various ceramic systems, zirconium-oxide ceramics based on yttria-stabilized polycrystalline tetragonal zirconia (Y-TZP) have become favoured, especially where high functional demands must be met.

However, the surface characteristics of the zirconia ceramic make it difficult to establish a durable mechanical or chemical bond in zirconia-based restorations. To use the durable siloxane bond between silica and a silane coupling agent, tribochemical silica coating followed by silanization was introduced to improve bond strength with zirconia ceramic. As several studies using this method have yielded inconsistent bond strength data, the method seems unlikely to provide a reliable durable bond with zirconia ceramic. In contrast, it has been shown that a combination of sandblasting and resin luting agents containing organophosphate ester monomers, such as 10-methacryloyloxydecyl dihydrogenphosphate (10-MDP), may result in a relatively high and durable bond to zirconia ceramic. However, there are insufficient data on the actual bonding mechanism of the 10-MDP monomer to zirconia ceramic and the long-term in vivo performance of the 10-MDP-containing resin luting cements.

In clinical practice, full-coverage zirconia ceramic restorations and fixed partial dentures not requiring high retention may be luted using various luting cements, whether resin-based or non-resin-based. Nonetheless, a 3-year clinical follow-up study suggests that loss of retention of zirconia ceramic restorations is related to the cement used. Except for zinc phosphate cements, most of these luting cements are expected to adhere to zirconia ceramic due to the presence of functional polymers or monomers in their composition. As the luting cements containing functional polymers or monomers seem to have differing surface energies, there may arise different interfacial chemical interactions and bonding potentials with zirconia ceramic. Asmussen and Peutzfeldt demonstrated that the surface energy parameters of composite and adhesive-treated dentine influence bond strength. In their study, bond strength increased with the “base” component of the surface energy of the resin composites and dentine treated with a higher surface energy adhesive yielded higher bond strengths. However, these features have not been extensively researched in the field of adhesion to zirconia ceramic.

The purpose of this study was thus to evaluate the shear bond strengths of various luting cements to a sandblasted zirconia ceramic and to test their bonding durability through thermocycling. The chemical composition of the luting cements was analysed using Fourier transform infrared (FTIR) spectroscopy. The surface roughness of luting cements and zirconia ceramic was measured using a profilometer. Surface energy parameters were determined by measuring contact angles on the eight luting cements and zirconia ceramic surface. Relationships between surface energy parameters and measured shear bond strengths were investigated using the Spearman rank correlation test.

Materials and methods

Shear bond strength test

Two conventional glass ionomer cements (GICs) (Fuji I, FI; Ketac Cem Easymix, KC), two resin-modified glass ionomer cements (RMGICs) (Fuji Plus, FP; RelyX Luting, RL), two compomer cements (Principle, PR; Ionotite F, IT), one adhesive resin cement (Panavia F 2.0, PV), and one self-adhesive resin cement (RelyX Unicem, UC) were selected for this in vitro study. Their codes, manufacturers, and compositions are summarized in Table 1 .

Table 1
Code, brand, composition, and batch number of luting cements used.
Code Product (manufacturer) Main composition (manufacturer supplied) Batch number
FI Fuji I (GC Corp., Tokyo, Japan) Powder: aluminosilicate glass, polyacrylic acid Powder: 0809031
Liquid: polyacrylic acid, water Liquid: 0809031
KC Ketac Cem Easymix (3M ESPE, St. Paul, MN, USA) Powder: glass powder, acrylic acid/maleic acid copolymer Powder: 342394
Liquid: water, tartaric acid Liquid: 329044
FP Fuji Plus (GC Corp.) Powder: aluminosilicate glass Powder: 0810141
Liquid: polyacrylic acid, water, HEMA, UDMA Liquid: 0810131
RL RelyX Luting (3M ESPE) Powder: fluoroaluminosilicate glass, potassium persulfate, ascorbic acid Powder: 9UF
Liquid: methacrylated polycarboxylic acid, water, tartaric acid, HEMA Liquid: 9WL
PR Principle (Dentsply Caulk, Milford, DE, USA) Powder: strontium aluminium fluorosilicate glass, sodium fluoride Powder: 0510131
Liquid: polymerizeable carboxylic acid dimethacrylate, PENTA, Bis-GMA, TEGDMA, HEMA, water Liquid: 0507221
IT Ionotite F (Tokuyama Dental Corp., Tokyo, Japan) Power: fluoroaluminosilicate glass filler, amorphous-fused silica filler, benzoyl peroxide Powder: 002
Liquid: methacryloxyalkyl acid phosphate (phosphoric acid monomer), MTU-6, UDMA, HEMA Liquid: 149
PV Panavia F 2.0 (Kuraray Medical Inc., Okayama, Japan) Paste A: dimethacylates, 10-MDP, camphorquinone, catalysts, initiators, silanated silica filler, silanated colloidal silica Paste A: 00405A
Paste B: dimethacrylates, catalysts, accelerators, silanated barium glass filler, sodium fluoride Paste B: 00074A
UC RelyX Unicem (3M ESPE) Base: methacrylated phosphoric acid esters, TEGDMA, sodium presulfate, glass powder, silane treated silica Cliker (base/catalyst): 372372
Catalyst: substituted dimethacrylate, sodium p -toluenesulfinate, calcium hydroxide, glass powder, silane treated silica
Monomer abbreviations : HEMA, 2-hydroxyethyl methacrylate; UDMA, urethane dimethacrylate; PENTA, dipentaerythritol pentaacrylate monophosphate; Bis-GMA, bisphenol A diglycidyl methacrylate; TEGDMA, triethylene glycol dimethacrylate; MTU-6, 6-methacryloyloxyhexyl 2-thiouracil-5-carboxylate; 10-MDP, 10-methacryloyloxydecyl dihydrogenphosphate.

Thirty-two cylinder-shaped (20 mm diameter and 1.5 mm thickness) commercial zirconia ceramic discs (Lava, 3 M ESPE, Seefeld, Germany) were prepared according to the manufacturer’s instructions and embedded in round silicone rubber moulds (25.4 mm in inner diameter, 19.0 mm in height) using poly(methyl methacrylate) resin, ensuring that one surface of the zirconia disc remained uncovered for adhering to cement. The exposed surface of each specimen was sandblasted with 110 μm Al 2 O 3 from a distance of 10 mm perpendicular to the specimen surface at a pressure of 0.25 MPa for 13 s and ultrasonically cleaned in isopropyl alcohol for 3 min. The sandblasted and ultrasonically cleaned zirconia ceramic surfaces were examined using an optical microscope (MM-40, Nikon, Tokyo, Japan) at a magnification of 100× and 500×.

To prepare hollow cylinders for bonding, clear and flexible polyvinyl chloride tubing (CFT-02-C, Small Parts Inc., Miami Lakes, FL, USA) with an inner diameter of 3.175 mm and an outer diameter of 6.35 mm was used. The tubing was inserted inside a long, clear acyclic split mould (inner diameter: 6.5 mm) so that 1.0 mm height of the tubing protruded perpendicularly from one surface of the mould. The protruding portion was cut off by moving a sharp razor blade (ST-300, Dorco Co., Ltd., Seoul, Korea) along the surface. In this manner, precisely cut tubings (approximately 1.0 mm in height) were obtained. The luting cements were mixed following manufacturers’ instructions and carefully inserted into the tubing lumens, ensuring that the tubing did not move or slip on the surface by pressing the tubing with a custom-made wire loop tool during the procedure. The materials were allowed to self-cure. The PR, PV, and UC cements were additionally irradiated for 20 s by holding the tip of the light guide of a light-curing unit (Elipar TriLight, 3 M ESPE; standard mode, output intensity = 750 mW/cm 2 ) approximately 1 mm above the cement-filled tubing. All specimens were left undisturbed for 30 min in 100% humidity and the tubing around cement cylinders was removed by vertically cutting the tubing into two or more fragments using a sharp blade (KB, Olfa Corp., Osaka, Japan). Special care was taken not to apply any stress or damage to the boned material. Cement cylinders that showed any apparent defects under an optical microscope (SMZ800, Nikon Corp., Tokyo, Japan) were excluded from this study and replaced. In this manner, five bonded cement cylinders of each luting cement were arranged in a circle on one zirconia surface ( n = 10). Prior to debonding, all bonded specimens were stored in distilled water at 37 °C for 48 h. Half of them were additionally thermocycled 10,000 times between 5 °C and 55 °C waterbaths with a dwelling time of 30 s and an exchange time of 5 s between each bath ( n = 10).

The bonded specimens were mounted in a jig attached to a universal testing machine (3343, Instron Inc., Canton, MA, USA). A stainless steel wire loop (0.41 mm diameter; G&H Wire Company, Greenwood, IN, USA) was placed around the cement cylinder so that it made contact with the lower half-circle of the cylinder and touched the zirconia surface. Each cylinder was stressed one by one by turning the specimen at a crosshead speed of 1 mm/min and the maximum load at failure was recorded, then converted to MPa. Following debonding, all fractured interfaces were examined under an optical microscope (SMZ800) to determine the mode of fracture. These were classified into one of three types: A, adhesive failure at the ceramic–cement interface; C, cohesive failure within cement; and AC, combination of these failure modes.

FTIR spectra

The FTIR spectra of eight luting cements tested in this study were obtained using a FTIR spectrophotometer (IRPrestige-21, Shimadzu Corp., Kyoto, Japan) with an attenuated total reflectance (ATR) unit (MIRacle, Pike Technologies Inc., Madison, WI, USA). Prior to analysis, an adhesive tape (Rapi, Cemedine Co., Ltd., Tokyo, Japan) was placed around the ATR diamond surface to act a vertical spacer, ensuring standard specimen thickness (approximately 100 μm). A small amount of each freshly mixed material was pressed against the diamond surface and covered with a polyester strip. The absorbance spectrum was acquired by scanning the specimens 10 times over a 4000–700 cm −1 range at a resolution of 4 cm −1 . The spectra were analysed and their peaks were assigned based on references.

Surface roughness

Prior to contact angle measurements, the surface roughness ( R a ) of the luting cements was determined. Cylindrical moulds (10 mm in diameter; 1 mm in height) were placed on a polyester strip (KerrHawe SA, Bioggio, Switzerland) over a glass slide. Each luting cement was filled into the mould, covered with another polyester strip and glass slide, and gently pressed to expel the excess material. The materials were then allowed to self-cure. The PR, PV, and UC cements were additionally irradiated for 20 s by placing the tip of the light guide of the light-curing unit against the upper glass slide. After 30 min, the strips were removed from the specimen, and the R a of each specimen was then measured five times using a previously calibrated profilometer (Surftest SV-400, Mitutoyo Corp., Kawasaki, Japan) at a stylus speed of 0.1 mm/s, a cutoff of 0.8 mm, and a range of 600 μm. The R a of each specimen was recorded as the average of the five readings. The R a values of the zirconia surfaces polished with silicon carbide paper up to #2000 and sandblasted were also measured after optical microscopic examination of the surfaces (MM-40) ( n = 10).

Surface energy parameters

To estimate the surface energy parameters of the eight luting cements and zirconia ceramic, the contact angle method was used. When a drop of liquid is placed on a solid that is assumed to be perfectly smooth and rigid, the angle of contact can be measured at the three phase point between the solid, the liquid, and the vapor. The interfacial tensions between the phases are expressed by the Young equation as follows :

γS=γSL+γLcosΘ γ S = γ S L + γ L cos Θ
where Θ is the contact angle and γ S , γ SL , and γ L are the surface tensions of the solid, solid–liquid and liquid surfaces, respectively. The Dupré equation states :
WA=γS+γLγSL W A = γ S + γ L γ S L
in which W A is the thermodynamic work of adhesion between a solid surface and a liquid. Substitution of the Young equation into the Dupré equation results in the Young–Dupré equation, which states :
WA=γL(1+cosΘ) W A = γ L ( 1 + cos Θ )
The surface energy parameters of the luting cements and zirconia ceramic can be calculated using the Young–Dupré equation in conjunction with Lifshitz-van der Waals/Lewis acid-base (van Oss and Good) theory :
γL(1+cosΘ)=2[(γdSγdL)1/2+(γ+SγL)1/2+(γSγ+L)1/2] γ L ( 1 + cos Θ ) = 2 [ ( γ S d γ L d ) 1 / 2 + ( γ S + γ L ) 1 / 2 + ( γ S γ L + ) 1 / 2 ]
where the superscripts d , +, and − refer to the dispersive (or Lifshitz-van der Waals), acid, and base components, respectively.

Cement specimens for contact angle measurement were prepared in the same way as for surface roughness measurement. The zirconia surface was polished with silicon carbide paper up to #2000. The contact angles were then determined by the sessile drop method at room temperature using a contact angle measurement apparatus (OCA 15 plus, DataPhysics Instrument GmbH, Filderstadt, Germany) on the surfaces of the luting cements and polished zirconia ceramic for each of four different test liquids with known surface energy parameters: water, glycerol, ethylene glycol, and 1-bromonaphthalene ( n = 10). By determination of Θ with the four reference liquids with known γ L , γdL,γL γ L d , γ L , and γ+L γ L + , four equations were set up for each of the surfaces from Eq. (4) . Based on the least square method, γdS,γ+S γ S d , γ S + , and γS γ S were calculated using computer software (SCA20, DataPhysics Instrument GmbH). The total surface energy γ S of the luting cements and zirconia ceramic was calculated from the following equation :

γS=γdS+2(γ+SγS)1/2 γ S = γ S d + 2 ( γ S + γ S ) 1 / 2

Statistical analysis

All the statistical analyses were carried out using SPSS 14.0 for Windows (SPSS Inc., Chicago, IL, USA) at a level of significance of α = 0.05. For the bond strength data, non-parametric statistical test procedures were used for group comparisons because Leven’s test showed inhomogeneity of variances between groups ( p < 0.001) and the data was found to be non-normally distributed. The Kruskal–Wallis test was employed between luting cements within each thermocycling condition, followed by the Mann–Whitney post hoc test, with adjustment of significance levels using the Sidak correction for multiple testing; between two thermocycling conditions within each luting cement, the Mann–Whitney test was used. Meanwhile, the surface roughness data for the luting cements and polished zirconia ceramic were analysed using one-way ANOVA and Tukey post hoc test. The Spearman rank correlation test was performed to evaluate the significance of the correlation between surface energy parameters and measured shear bond strengths.

Results

The shear bond strengths of the eight luting cements to sandblasted zirconia ceramic are summarized in Table 2 . Kruskall–Wallis test revealed statistically significant differences amongst the test luting cements both before and after thermocycling ( p < 0.001). FI and KC showed low initial bond strengths and, furthermore, significant drops in the value after thermocycling ( p = 0.001 and p < 0.001, respectively). Though FP showed the highest initial value amongst the four GICs ( p < 0.001), it exhibited a significant decrease in the value after thermocycling ( p < 0.001). PV exhibited the highest initial bond strength amongst all the luting cements tested, the value for PR being statistically similar ( p = 0.106). UC exhibited a significantly lower initial bond strength than PV and PR ( p < 0.001). After thermocycling, PV, PR, and UC exhibited no significant decreases in their respective initial bond strengths ( p > 0.05). IT yielded very low bond strengths both before and after thermocycling. In this study, all debonded specimens showed adhesive failure regardless of luting materials tested or thermocycling conditions.

Table 2
Shear bond strengths (MPa) of eight luting cements to sandblasted zirconia ceramic.
Luting cement 48-h immersion in water (thermocycling 0) 48-h immersion in water + additional 10,000 thermocycling
Median Mean S.D. Median Mean S.D.
FI 1.06 Da 0.94 0.36 0.07 Eb 0.06 0.05
KC 1.50 Ca 1.93 0.69 0.00 Eb 0.00 0.00
FP 8.18 Ba 7.87 0.86 3.91 Cb 3.84 0.62
RL 2.92 Ca 2.99 0.60 2.80 Da 2.77 0.45
PR 9.92 Aa 9.67 0.57 9.03 ABa 9.12 1.23
IT 0.13 Ea 0.14 0.12 0.00 Eb 0.00 0.00
PV 11.23 Aa 11.31 1.25 10.82 Aa 10.31 1.63
UC 6.75 Ba 6.83 0.50 7.13 Ba 7.12 1.33
FI, Fuji I; KC, Ketac Cem Easymix; FP: Fuji Plus; RL: RelyX Luting; PR: Principle; IT: Ionotite F; PV: Panavia F 2.0; UC: RelyX Unicem.
Within the same column, medians with the same uppercase letter are not statistically different according to the Kruskal–Wallis test followed by the Mann–Whitney post hoc test with adjustment of significance levels using the Sidak correction for multiple testing ( p > 0.05). Within the same row, medians with the same lowercase letter are not statistically different according to Mann–Whitney test ( p > 0.05).

Fig. 1 shows the representative FTIR spectra of the freshly mixed luting cements. For the conventional GICs and RMGICs, broad absorption bands indicating hydrogen-bonded O–H stretching vibrations were noted at 3500–3250 and 2700–2500 cm −1 ( Fig. 1 a–d). For the RMGICs (FP and RL), the peak indicating C C stretching vibration (∼1640 cm −1 ) was observed. In contrast, the compomer cements (PR and IT) and resin cements (PV and UC) showed no prominent hydrogen-bonded O–H stretching vibrations ( Fig. 1 e–h). In addition, the latter four luting cements showed phosphorus–oxygen double bond (P O) and phosphorus–oxygen–alkyl (P–O–R) stretching vibrations at ∼1250 and ∼1150 cm −1 , respectively ( Fig. 1 e–h).

FITR spectra of luting cements used in this study: (a) Fuji I, (b) Ketac Cem Easymix, (c) Fuji Plus, (d) RelyX Luting, (e) Principle, (f) Ionotite F, (g) Panavia F 2.0, and (h) RelyX Unicem. References for peak assignment: O–H: Eliades et al. 21 ; C C: Kakaboura et al. 19 ; P O and P–O–R: Verdelis et al. 20
Fig. 1
FITR spectra of luting cements used in this study: (a) Fuji I, (b) Ketac Cem Easymix, (c) Fuji Plus, (d) RelyX Luting, (e) Principle, (f) Ionotite F, (g) Panavia F 2.0, and (h) RelyX Unicem. References for peak assignment: O–H: Eliades et al. ; C C: Kakaboura et al. ; P O and P–O–R: Verdelis et al.

Surface roughness values of the luting cements and zirconia ceramic are presented in Table 3 . The mean R a values for the eight luting cements and polished zirconia ceramic surface ranged from 0.104 to 0.167 μm, with significant differences amongst them (one-way ANOVA, p = 0.003). Sandblasting substantially increased the surface roughness of polished zirconia ceramic.

Table 3
Mean R a surface roughness (μm) (S.D.) of luting cements and zirconia ceramic.
FI KC FP RL PR IT PV UC Polished zirconia Sandblasted zirconia
0.104 (0.025) A 0.121 (0.030) AB 0.167 (0.045) B 0.115 (0.032) A 0.129 (0.036) AB 0.130 (0.043) AB 0.135 (0.031) AB 0.132 (0.034) AB 0.107 (0.013) A 0.784 (0.081)
FI, Fuji I; KC, Ketac Cem Easymix; FP: Fuji Plus; RL: RelyX Luting; PR: Principle; IT: Ionotite F; PV: Panavia F 2.0; UC: RelyX Unicem.
The same uppercase letter indicates lack of statistically difference amongst the luting cements and polished zirconia based on one-way ANOVA and Tukey post hoc tests ( p > 0.05).

Tables 4 and 5 summarize the contact angles, total surface energies, and surface energy parameters of the four liquids measured on luting cements and polished zirconia ceramic. The calculated mean total surface energy of the zirconia ceramic was 36.64 mJ/m 2 ; the mean values for the eight luting cements varied from 27.10 to 38.79 mJ/m 2 . Examining a correlation between individual surface energy parameters for the luting cements and the shear bond strengths, we found significant effects of the base component on the bond strengths both before and after thermocycling ( r s = 0.857, p = 0.007 and r s = 0.826, p = 0.011, respectively) ( Fig. 2 ).

Table 4
Mean contact angles (°) (S.D.) of four liquids measured on luting cements and polished zirconia ceramic.
Surface Water Glycerol Ethylene glycol 1-Bromonaphthalene
FI 72.2 (2.3) 83.8 (4.4) 76.7 (3.2) 52.6 (3.0)
KC 86.7 (7.2) 84.9 (7.2) 66.1 (6.4) 57.1 (5.1)
FP 63.3 (1.5) 85.7 (3.9) 64.1 (4.2) 43.4 (1.9)
RL 73.7 (4.2) 84.0 (4.4) 71.7 (7.7) 44.5 (4.2)
PR 66.5 (3.2) 80.9 (4.3) 61.9 (2.6) 50.6 (2.0)
IT 88.8 (6.8) 87.3 (8.5) 76.0 (1.8) 46.5 (5.7)
PV 67.3 (5.5) 85.8 (4.3) 59.7 (2.4) 47.1 (3.1)
UC 71.1 (4.0) 88.5 (2.2) 64.5 (3.3) 44.0 (2.1)
Zirconia 52.7 (3.4) 78.4 (3.4) 60.5 (2.1) 45.9 (1.0)
FI, Fuji I; KC, Ketac Cem Easymix; FP: Fuji Plus; RL: RelyX Luting; PR: Principle; IT: Ionotite F; PV: Panavia F 2.0; UC: RelyX Unicem.
Table 5
Total surface energy and surface energy parameters (mJ/m 2 ) of luting cements and zirconia ceramic [mean (S.D.)].
Surface γ S (total surface energy) γdS γ S d (dispersive component) γ+S γ S + (acid component) γS γ S (base component)
FI 32.10 (4.90) 28.06 (1.45) 0.18 (0.29) 23.08 (1.58)
KC 27.10 (8.54) 26.69 (2.54) 0.00 (0.14) 8.80 (1.74)
FP 38.79 (3.66) 33.05 (0.87) 0.26 (0.24) 31.72 (1.64)
RL 35.34 (6.93) 32.41 (1.94) 0.12 (0.40) 17.95 (1.62)
PR 31.18 (5.47) 29.64 (0.97) 0.02 (0.13) 26.78 (1.50)
IT 33.63 (7.63) 31.21 (2.63) 0.25 (0.98) 5.85 (1.28)
PV 34.80 (8.84) 31.58 (1.46) 0.09 (0.41) 28.76 (1.79)
UC 37.78 (5.83) 32.85 (0.97) 0.26 (0.51) 23.11 (0.96)
Zirconia 36.64 (5.57) 31.86 (0.47) 0.14 (0.28) 42.28 (1.46)
FI, Fuji I; KC, Ketac Cem Easymix; FP: Fuji Plus; RL: RelyX Luting; PR: Principle; IT: Ionotite F; PV: Panavia F 2.0; UC: RelyX Unicem.
Scatter plot of Spearman correlation analysis results: (a) 48-h immersion in water (thermocycling 0) and (b) 48-h immersion in water + additional 10,000 thermocycling.
Fig. 2
Scatter plot of Spearman correlation analysis results: (a) 48-h immersion in water (thermocycling 0) and (b) 48-h immersion in water + additional 10,000 thermocycling.

Discussion

A passive film of zirconium oxide easily covers the zirconium surface. As chemical characteristics of zirconia ceramic surface are similar to those of metal surface, hydroxyl (O–H) groups may be present on the zirconia ceramic surface. It is also likely that a luting cement containing any polymers or monomers with polar functional groups is capable of bonding chemically with polar hydroxyl groups on the zirconia ceramic surface. Therefore, all the luting cements tested in this study were expected to have the potential to bond chemically to zirconia ceramic due to their respective functional polymers (e.g. polyalkenoic acid copolymer) or monomers [e.g. 2-hydroxyethyl methacrylate (HEMA), carboxylic acid monomer, and an organophosphate ester monomer such as dipentaerythritol pentaacrylate monophosphate (PENTA) and 10-MDP] ( Table 1 ).

Assuming a minimal bond strength requirement of approximately 10–13 MPa for acceptable clinical bonding, the shear bond strengths for all the luting cements to “sandblasted” zirconia were not high enough to ensure favourable clinical service ( Table 2 ). In addition, regardless of the luting cements tested or the thermocycling conditions, all debonded specimens showed only adhesive failure at the cement–zirconia interface. These findings suggest that the bond formed at the interface might be limited to weak intermolecular attractions such as hydrogen bonding between the polar functional groups of the polymers or monomers in the cement and the polar hydroxyl groups on the zirconia, in tandem with micromechanical retention of the roughened zirconia ceramic surface by sandblasting.

To estimate the bonding potential of the luting cements to zirconia ceramic, we calculated surface energy parameters of the materials using a contact angle method. The surface energy parameters were determined on cured cements in this study, though uncured material is applied to the zirconia surface clinically. It can be assumed that monomer or polymer surface energy parameters of the materials are similar in value to those of the cured materials because they contain the same molecular groups and segments, with the exception of the vinyl group. In contact angle methods, moreover, the solid is assumed to be perfectly smooth and rigid. However, changes in surface roughness can influence the contact angle, thereby changing surface energy characteristics. Therefore, we investigated the surface roughness of the luting cements and the polished zirconia ceramic prior to contact angle measurements of the materials. Despite significant differences in the R a amongst the materials, there were only small variations in the value ( Table 3 ). Busscher et al. reported that changes in solid surface R a below 0.1 μm have no effect on contact angle. Given that the range of mean R a values of the luting cements and polished zirconia ceramic were below 0.1 μm in the present study, R a may be assumed to have no significant effect on the contact angle used to calculate the surface energy parameters.

The Lifshitz-van der Waals/Lewis acid–base approach entails decomposing the surface energy into Lifshitz-van der Waals interactions and Lewis acid–base interactions. In the present study, the base component (γS) ( γ S ) of the polished zirconia ceramic surface was larger than the acid component (γ+S) ( γ S + ) ( Table 5 ). When the Lewis acid component (γ+S) ( γ S + ) and the Lewis base component (γS) ( γ S ) are considered in the context of a hydrogen bond, the former and the latter may be viewed as the hydrogen bond donating (HBD) component and the hydrogen bond accepting (HBA) component, respectively. The zirconia ceramic surface thus appears to have a predominantly HBA character. Similarly, all luting cements tested in this study also showed large base components and very small acid components, indicating a primarily HBA nature. The Spearman rank correlation test between surface energy parameters and measured shear bond strengths showed significant influence of the base component on bond strengths both before and after thermocycling as opposed to the acid component, dispersive component, and total surface energy ( Fig. 2 ). This suggests that the HBA (base) component of the luting cements interacts with the HBD (acid) component of the zirconia surface. Although the HBA component of the zirconia surface is great, some of the surface appears to bear an HBD character. As shown in the correlation test, increasing the HBA character of the luting cement results in increased interaction with the HBD zirconia ceramic surface.

GICs contain a large amount of water, as shown in the FTIR spectra (3500–3250 cm −1 , Fig. 1 ). Polyalkenoic acid copolymers that remain unionized in water may bear two sites potentially available for hydrogen boding with zirconia: oxygen atoms of carbonyl moieties and hydrogen atoms of hydroxyl groups. The surface energy analysis results suggest the role of the former might be predominant as a hydrogen bond acceptor rather than the latter. It is also possible that HEMA also participates in hydrogen bonding with zirconia due to the presence of a polar head (hydroxyl group). A small amount of dimethacrylate monomer incorporated into RMGICs form crosslinked poly-HEMA during polymerization. As for polyalkenoic acid copolymers, poly-HEMA is hydrophilic and tends to absorb water, leading to plasticity and hygroscopic expansion. Although not recommended for cementation of most all-ceramic restorations, mainly due their suspected role in crack initiation, RMGICs can be indicated for luting of all-zirconia strengthened core ceramic restorations. A significantly higher bond strength for the FP compared to the other three GICs could be partly attributed to the incorporation of the less hydrophilic dimethacrylat monomer urethane dimethacrylate (UDMA) ( Table 1 ). However, UDMA also tends to absorb water due to the presence of hydrophilic urethane groups in its structure. Thus, RMGICs as well as conventional GICs do not seem to impart high bonding durability to zirconia though some of them share a high HBA character ( Tables 2 and 5 ).

In this study, PV produced the highest initial bond strength ( Table 2 ), in accordance with the results of previous studies. Panavia products contain a well-known organophosphate ester monomer, 10-MDP, which is structurally a phosphate monoester having one P O and two hydroxyl groups in one molecule. As PV contains little water, most of the 10-MDP seems to remain undissociated notwithstanding its acidic nature. Similar to the carboxyl groups of the polyalkenoic acid copolymers, the hydroxyl groups and P O oxygen are likely to participate as hydrogen bonding acceptors with the zirconia surface. Unlike the polyalkenoic acid copolymers and HEMA in GIC materials, the apolar spacer group consisting of a long saturated carbon chain in 10-MDP might keep water molecules at a distance from the hydrogen bond formed at the cement–zirconia interface. UC, which contains another organophosphate ester monomer, produced a significantly lower initial bond strength than PV. This may be attributable to a difference in the HBA component between the two materials ( Table 5 ). It seems that hydrogen bonding by the phosphate ester monomers may also depend on the chemical structure of the monomer.

The two compomer cements produced greatly differing initial bond strengths from each other: PR showed a bond strength statistically comparable to PV, whereas IT exhibited a very poor value. As represented in the FTIR spectra ( Fig. 1 ), the two materials might fall into the category of resin luting cements, although they are sometimes classified as RMGICs. PR showed a considerably larger HBA component than IT ( Table 5 ). The high bond strength of PR before and after thermocycling might be partly attributable to the combined effect of the two types of functional monomers (PENTA and carboxylic acid monomer, Table 1 ), which could heighten the HBA capability.

It must be noted that the shear bond strengths of the luting cements do not indicate genuine values of chemical bonding because the materials were bonded to a sandblasted zirconia ceramic surface. In this study, bond strengths were determined using sandblasted zirconia ceramic surfaces, whereas surface energy analysis was performed on polished ones. Preliminary bonding tests performed on polished zirconia ceramic surfaces showed frequent failure during bonded specimen preparation, making differentiation amongst the groups impossible. Although there is still some controversy about the effect, sandblasting followed by ultrasonic cleansing is believed not only to increase surface roughness and bonding surface area but also to remove any contaminants from the ceramic surface. The surface area of sandblasted zirconia available for chemical bonding with the luting cements was relatively constant ( R a = 0.784 ± 0.081 μm, Table 3 ), possibly allowing a correlation between bond strengths and surface energy parameters. It should also be noted that only one sandblasting condition (with 110 μm Al 2 O 3 at 0.25 MPa) was tested in the current study. Kern et al. suggested that low-pressure (0.05 MPa) sandblasting with 50 μm Al 2 O 3 prior to priming was effective in obtaining durable resin bonding to zirconia ceramics (Cercon, DeguDent), reduction in surface roughness notwithstanding. In contrast, Ntala et al. reported that sandblasting with 50 μm Al 2 O 3 at 0.25 MPa did not produce sufficient micromechanical retention on zirconia ceramics (Kavo Everest, Kavo) for adhesive bonding. Therefore, various sandblasting conditions of zirconia ceramic should be further tested in conjunction with variations in Y-TZP composition to determine the optimal surface for bonding with luting cements.

The present study provides laboratory data on the performance of a zirconia-based restoration luted with various cements. At present, there seems to be no definite thermocycling condition for predicting clinical performance of luting cements to zirconia ceramic. Previous zirconia ceramic studies show great variability in number of thermal cycles (for example, 5000 cycles, 10,000 cycles, and 20,000 cycles ). In Kern and Wegner and Kern et al. bonded zirconia ceramic specimens were stored in water 150 days and then thermal cycled 37,500 times to stress the bond interface. Gale and Darvell advise that the poor discriminatory power of excessive stresses may lead to unreasonable characterisation of materials which could in fact serve well in clinical practice. We used 10,000 thermal cycles to statistically distinguish the bonding performance of the luting cements, although this condition may not directly translate into clinical results.

This in vitro study suggests that surface characteristics of luting cements may be key determinants of their bonding potential with zirconia ceramic, as shown in Fig. 2 . The Spearman rank correlation test reveals a strong, though not perfect, linear relationship between the HBA (base) component and bond strength ( r s = 0.857 and r s = 0.826 before and after thermocycling, respectively), so that other factors may also, to a small extent, affect bonding properties. In addition, more precise surface energy parameter approaches and experimental set-ups may better predict the adhesive potential of luting cements. Although this study only tested adhesive resin cements, commercially available non-adhesive resin cements lack potential for chemical interaction with zirconia ceramic and thus require surface pre-treatment to achieve strong and durable bonding. Recent approaches to improving bonding with zirconia include the use of primers, including one containing organophosphate and carboxylic acid monomers. When these primers are applied to a zirconia ceramic surface, eliminating hydrophilic constituents such as water can improve the bond strength of composite resin to the ceramic. Such approaches might also benefit from utilising surface energy analyses to predict bonding behaviour of primers to zirconia ceramic based on application method. In particular, these analyses should consider surface roughness in the context of methods that alter the surface topography of zirconia ceramic surface (e.g., selective infiltration etching, the application of hot chemical etching solution, and the application of a glass–ceramic/glaze material containing a major lithium disilicate phase ).

Along with bonding properties, mechanical properties of luting cements should be considered in designing zirconia-based restorations with good clinical performance during service. In addition, optical properties of various luting cements have to be considered for clinically optimal aesthetic results. Further research is plainly required to establish guidelines on conventional and adhesive cementation in specific clinical situations.

Conclusions

The resin cement Panavia F 2.0 and the compomer cement Principle bonded to sandblasted zirconia gave clinically acceptable shear bond strengths both before and after thermocycling, whereas bond strength values of four GICs were too low for acceptable clinical bonding. Measured bond strengths both before and after thermocycling showed significant positive correlations with the base (hydrogen bond accepting) components of the luting materials.

Acknowledgement

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology ( 2010-0029494 ).

References

  • 1. Derand T., Molin M., Kvam K.: Bond strength of composite luting cement to zirconia ceramic surfaces. Dental Materials 2005; 21: pp. 1158-1162.
  • 2. Kern M., Wegner S.M.: Bonding to zirconia ceramic: adhesion methods and their durability. Dental Materials 1998; 14: pp. 64-71.
  • 3. Yoshida K., Yamashita M., Atsuta M.: Zirconate coupling agent for bonding resin luting cement to pure zirconium. American Journal of Dentistry 2004; 17: pp. 249-252.
  • 4. Piwowarczyk A., Lauer H.C., Sorensen J.A.: The shear bond strength between luting cements and zirconia ceramics after two pre-treatments. Operative Dentistry 2005; 30: pp. 382-388.
  • 5. Matinlinna J.P., Heikkinen T., Özcan M., Lassila L.V., Vallittu P.K.: Evaluation of resin adhesion to zirconia ceramic using some organosilanes. Dental Materials 2006; 22: pp. 824-831.
  • 6. Wegner S.M., Kern M.: Long-term resin bond strength to zirconia ceramic. Journal of Adhesive Dentistry 2000; 2: pp. 139-147.
  • 7. Aboushelib M.N., Matinlinna J.P., Salameh Z., Ounsi H.: Innovations in bonding to zirconia-based materials: Part I. Dental Materials 2008; 24: pp. 1268-1272.
  • 8. Roediger M., Gersdorff N., Huels A., Rinke S.: Prospective evaluation of zirconia posterior fixed partial dentures: four-year clinical results. International Journal of Prosthodontics 2010; 23: pp. 141-148.
  • 9. Ntala P., Chen X., Niggli J., Cattell M.: Development and testing of multi-phase glazes for adhesive bonding to zirconia substrates. Journal of Dentistry 2010; 38: pp. 773-781.
  • 10. Örtorp A., Kihl M.L., Carlsson G.E.: A 3-year retrospective and clinical follow-up study of zirconia single crowns performed in a private practice. Journal of Dentistry 2009; 37: pp. 731-736.
  • 11. Rosentritt M., Steiger D., Behr M., Handel G., Kolbeck C.: Influence of substructure design and spacer settings on the in vitro performance of molar zirconia crowns. Journal of Dentistry 2009; 37: pp. 978-983.
  • 12. Hill E.E.: Dental cements for definitive luting: a review and practical clinical considerations. Dental Clinics of North America 2007; 51: pp. 643-658. vi
  • 13. Asmussen E., Peutzfeldt A.: Resin composites: strength of the bond to dentin versus surface energy parameters. Dental Materials 2005; 21: pp. 1039-1043.
  • 14. Phark J.H., Duarte S., Blatz M., Sadan A.: An in vitro evaluation of the long-term resin bond to a new densely sintered high-purity zirconium-oxide ceramic surface. Journal of Prosthetic Dentistry 2009; 101: pp. 29-38.
  • 15. Glasspoole E.A., Erickson R.L., Davidson C.L.: Effect of surface treatments on the bond strength of glass ionomers to enamel. Dental Materials 2002; 18: pp. 454-462.
  • 16. Lüthy H., Loeffel O., Hammerle C.H.: Effect of thermocycling on bond strength of luting cements to zirconia ceramic. Dental Materials 2006; 22: pp. 195-200.
  • 17. Sadr A., Ghasemi A., Shimada Y., Tagami J.: Effects of storage time and temperature on the properties of two self-etching systems. Journal of Dentistry 2007; 35: pp. 218-225.
  • 18. Arrais C.A., Giannini M., Rueggeberg F.A.: Effect of sodium sulfinate salts on the polymerization characteristics of dual-cured resin cement systems exposed to attenuated light-activation. Journal of Dentistry 2009; 37: pp. 219-227.
  • 19. Kakaboura A., Eliades G., Palaghias G.: An FTIR study on the setting mechanism of resin-modified glass ionomer restoratives. Dental Materials 1996; 12: pp. 173-178.
  • 20. Verdelis K., Eliades G., Oviir T., Margelos J.: Effect of chelating agents on the molecular composition and extent of decalcification at cervical, middle and apical root dentin locations. Endodontics and Dental Traumatology 1999; 15: pp. 164-170.
  • 21. Eliades T., Eliades G., Silikas N., Watts D.C.: Orthodontic bonding to wet enamel with water-insensitive and water-activated orthodontic adhesive resins.Eliades G.Watts D.C.Eliades T.Dental hard tissues and bonding: interfacial phenomena and related properties.2005.SpringerBerlin/Heidelberg:pp. 71-85.
  • 22. Schrader M.E.: Work of adhesion of a sessile drop to a clean surface. Journal of Colloid and Interface Science 1999; 213: pp. 602-605.
  • 23. Pocius A.V.: Adhesion and adhesive technology: an introduction.2nd ed.2002.Hanser Gardner Publications Inc.Cincinnati p. 132–63
  • 24. Ratner B.D.: Surface structure and properties.Williams D.Cahn R.W.Bever M.B.Concise encyclopedia of medical & dental materials.1990.PergamonOxford:pp. 337-346.
  • 25. Asmussen E., Attal J.P., Degrange M.: Factors affecting the adherence energy of experimental resin cements bonded to a nickel–chromium alloy. Journal of Dental Research 1995; 74: pp. 715-720.
  • 26. Lee L.H.: Adhesion and surface-hydrogen-bond components for polymers and biomaterials. Journal of Adhesion 1998; 67: pp. 1-18.
  • 27. Chen H., Li L., Jones P.M., Hsia Y.T., Jhon M.S.: Lifshitz-van der Waals and Lewis acid–base approach for analyzing surface energy of molecularly thin lubricant films. IEEE Transactions on Magnetics 2007; 43: pp. 2226-2228.
  • 28. Yoshida K., Tsuo Y., Atsuta M.: Bonding of dual-cured resin cement to zirconia ceramic using phosphate acid ester monomer and zirconate coupler. Journal of Biomedical Materials Research Part B: Applied Biomaterials 2006; 77: pp. 28-33.
  • 29. Yun J.Y., Ha S.R., Lee J.B., Kim S.H.: Effect of sandblasting and various metal primers on the shear bond strength of resin cement to Y-TZP ceramic. Dental Materials 2010; 26: pp. 650-658.
  • 30. Kern M., Barloi A., Yang B.: Surface conditioning influences zirconia ceramic bonding. Journal of Dental Research 2009; 88: pp. 817-822.
  • 31. Vaidyanathan T.K., Vaidyanathan J.: Recent advances in the theory and mechanism of adhesive resin bonding to dentin: a critical review. Journal of Biomedical Materials Research Part B: Applied Biomaterials 2009; 88: pp. 558-578.
  • 32. Fabianelli A., Pollington S., Papacchini F., Goracci C., Cantoro A., Ferrari M., et. al.: The effect of different surface treatments on bond strength between leucite reinforced feldspathic ceramic and composite resin. Journal of Dentistry 2010; 38: pp. 39-43.
  • 33. Quirynen M., Bollen C.M.: The influence of surface roughness and surface-free energy on supra- and subgingival plaque formation in man. A review of the literature. Journal of Clinical Periodontology 1995; 22: pp. 1-14.
  • 34. Bollen C.M., Papaioanno W., Van Eldere J., Schepers E., Quirynen M., van Steenberghe D.: The influence of abutment surface roughness on plaque accumulation and peri-implant mucositis. Clinical Oral Implants Research 1996; 7: pp. 201-211.
  • 35. Ahn H.B., Ahn S.J., Lee S.J., Kim T.W., Nahm D.S.: Analysis of surface roughness and surface free energy characteristics of various orthodontic materials. American Journal of Orthodontics and Dentofacial Orthopedics 2009; 136: pp. 668-674.
  • 36. Busscher H.J., van Pelt A.W.J., de Boer P., de Jong H.P., Arends J.: The effect of surface roughening of polymers on measured contact angles of liquids. Colloids and Surfaces 1984; 9: pp. 319-331.
  • 37. Steed J.W., Turner D.R., Wallace K.J.: Core concepts in supramolecular chemistry and nanochemistry.1st ed.2007.John Wiley & Sons, Ltd.Chichester p. 1–27
  • 38. McCabe J.F.: Resin-modified glass-ionomers. Biomaterials 1998; 19: pp. 521-527.
  • 39. Smith D.C.: Development of glass-ionomer cement systems. Biomaterials 1998; 19: pp. 467-478.
  • 40. Van Landuyt K.L., Snauwaert J., De Munck J., Peumans M., Yoshida Y., Poitevin A., et. al.: Systematic review of the chemical composition of contemporary dental adhesives. Biomaterials 2007; 28: pp. 3757-3785.
  • 41. Good M.L., Mitchell C.A., Pintado M.R., Douglas W.H.: Quantification of all-ceramic crown margin surface profile from try-in to 1-week post-cementation. Journal of Dentistry 2009; 37: pp. 65-75.
  • 42. Van Landuyt K.L., Yoshida Y., Hirata I., Snauwaert J., De Munck J., Okazaki M., et. al.: Influence of the chemical structure of functional monomers on their adhesive performance. Journal of Dental Research 2008; 87: pp. 757-761.
  • 43. Taher N.M., Ateyah N.Z.: Shear bond strength of resin modified glass ionomer cement bonded to different tooth-colored restorative materials. Journal of Contemporary Dental Practice 2007; 8: pp. 25-34.
  • 44. Wolfart M., Lehmann F., Wolfart S., Kern M.: Durability of the resin bond strength to zirconia ceramic after using different surface conditioning methods. Dental Materials 2007; 23: pp. 45-50.
  • 45. Gale M.S., Darvell B.W.: Thermal cycling procedures for laboratory testing of dental restorations. Journal of Dentistry 1999; 27: pp. 89-99.
  • 46. Casucci A., Osorio E., Osorio R., Monticelli F., Toledano M., Mazzitelli C., et. al.: Influence of different surface treatments on surface zirconia frameworks. Journal of Dentistry 2009; 37: pp. 891-897.
  • 47. Magne P., Paranhos M.P., Burnett L.H.: New zirconia primer improves bond strength of resin-based cements. Dental Materials 2010; 26: pp. 345-352.
  • 48. Chang J., Da Silva J.D., Sakai M., Kristiansen J., Ishikawa-Nagai S.: The optical effect of composite luting cement on all ceramic crowns. Journal of Dentistry 2009; 37: pp. 937-943.

Was this article helpful?

Shear bond strengths of various luting cements to zirconia ceramic: Surface chemical aspects Min-Jeong Kim , Young Kyung Kim , Kyo-Han Kim and Tae-Yub Kwon Journal of Dentistry, 2011-11-01, Volume 39, Issue 11, Pages 795-803, Copyright © 2011 Elsevier Ltd Abstract Objectives To measure the shear bond strengths of various luting cements to a sandblasted zirconia ceramic and to determine the surface energy parameters of the luting cements. Methods Two conventional glass ionomer cements, two resin-modified glass ionomer cements, two compomer cements, and two adhesive resin cements were prepared and bonded to sandblasted zirconia (Lava). All bonded specimens were stored in water at 37 °C for 48 h and then half of them additionally thermocycled 10,000 times prior to the shear bond strength test ( n = 10). Surface roughness ( R a ) values and surface energy parameters of the eight luting cements and polished zirconia ceramic were evaluated using a profilometer and contact angle measurements, respectively ( n = 10). The bond strength and surface roughness data were statistically analysed using non-parametric and parametric procedures, respectively ( α = 0.05). Relationships between surface energy parameters and measured shear bond strengths were investigated using the Spearman rank correlation test. Results Panavia F 2.0 and Principle produced higher bond strengths than the other cements, with no significant changes before and after thermocycling. Fuji I, Ketac Cem Easymix, and Ionotite F yielded near-zero or zero values after thermocycling. All debonded specimens showed adhesive failure. Mean R a values ranged from 0.104 to 0.167 μm. We found the base (hydrogen bond accepting) components of the luting cements significantly affected the bond strengths both before and after thermocycling. Conclusion It is recommended that the surface energy parameters of luting cements be considered in evaluating their adhesive properties with zirconia ceramic. 1 Introduction Modern all-ceramic dental restorations provide cosmetically good and metal-free treatment options. Amongst various ceramic systems, zirconium-oxide ceramics based on yttria-stabilized polycrystalline tetragonal zirconia (Y-TZP) have become favoured, especially where high functional demands must be met. However, the surface characteristics of the zirconia ceramic make it difficult to establish a durable mechanical or chemical bond in zirconia-based restorations. To use the durable siloxane bond between silica and a silane coupling agent, tribochemical silica coating followed by silanization was introduced to improve bond strength with zirconia ceramic. As several studies using this method have yielded inconsistent bond strength data, the method seems unlikely to provide a reliable durable bond with zirconia ceramic. In contrast, it has been shown that a combination of sandblasting and resin luting agents containing organophosphate ester monomers, such as 10-methacryloyloxydecyl dihydrogenphosphate (10-MDP), may result in a relatively high and durable bond to zirconia ceramic. However, there are insufficient data on the actual bonding mechanism of the 10-MDP monomer to zirconia ceramic and the long-term in vivo performance of the 10-MDP-containing resin luting cements. In clinical practice, full-coverage zirconia ceramic restorations and fixed partial dentures not requiring high retention may be luted using various luting cements, whether resin-based or non-resin-based. Nonetheless, a 3-year clinical follow-up study suggests that loss of retention of zirconia ceramic restorations is related to the cement used. Except for zinc phosphate cements, most of these luting cements are expected to adhere to zirconia ceramic due to the presence of functional polymers or monomers in their composition. As the luting cements containing functional polymers or monomers seem to have differing surface energies, there may arise different interfacial chemical interactions and bonding potentials with zirconia ceramic. Asmussen and Peutzfeldt demonstrated that the surface energy parameters of composite and adhesive-treated dentine influence bond strength. In their study, bond strength increased with the “base” component of the surface energy of the resin composites and dentine treated with a higher surface energy adhesive yielded higher bond strengths. However, these features have not been extensively researched in the field of adhesion to zirconia ceramic. The purpose of this study was thus to evaluate the shear bond strengths of various luting cements to a sandblasted zirconia ceramic and to test their bonding durability through thermocycling. The chemical composition of the luting cements was analysed using Fourier transform infrared (FTIR) spectroscopy. The surface roughness of luting cements and zirconia ceramic was measured using a profilometer. Surface energy parameters were determined by measuring contact angles on the eight luting cements and zirconia ceramic surface. Relationships between surface energy parameters and measured shear bond strengths were investigated using the Spearman rank correlation test. 2 Materials and methods 2.1 Shear bond strength test Two conventional glass ionomer cements (GICs) (Fuji I, FI; Ketac Cem Easymix, KC), two resin-modified glass ionomer cements (RMGICs) (Fuji Plus, FP; RelyX Luting, RL), two compomer cements (Principle, PR; Ionotite F, IT), one adhesive resin cement (Panavia F 2.0, PV), and one self-adhesive resin cement (RelyX Unicem, UC) were selected for this in vitro study. Their codes, manufacturers, and compositions are summarized in Table 1 . Table 1 Code, brand, composition, and batch number of luting cements used. Code Product (manufacturer) Main composition (manufacturer supplied) Batch number FI Fuji I (GC Corp., Tokyo, Japan) Powder: aluminosilicate glass, polyacrylic acid Powder: 0809031 Liquid: polyacrylic acid, water Liquid: 0809031 KC Ketac Cem Easymix (3M ESPE, St. Paul, MN, USA) Powder: glass powder, acrylic acid/maleic acid copolymer Powder: 342394 Liquid: water, tartaric acid Liquid: 329044 FP Fuji Plus (GC Corp.) Powder: aluminosilicate glass Powder: 0810141 Liquid: polyacrylic acid, water, HEMA, UDMA Liquid: 0810131 RL RelyX Luting (3M ESPE) Powder: fluoroaluminosilicate glass, potassium persulfate, ascorbic acid Powder: 9UF Liquid: methacrylated polycarboxylic acid, water, tartaric acid, HEMA Liquid: 9WL PR Principle (Dentsply Caulk, Milford, DE, USA) Powder: strontium aluminium fluorosilicate glass, sodium fluoride Powder: 0510131 Liquid: polymerizeable carboxylic acid dimethacrylate, PENTA, Bis-GMA, TEGDMA, HEMA, water Liquid: 0507221 IT Ionotite F (Tokuyama Dental Corp., Tokyo, Japan) Power: fluoroaluminosilicate glass filler, amorphous-fused silica filler, benzoyl peroxide Powder: 002 Liquid: methacryloxyalkyl acid phosphate (phosphoric acid monomer), MTU-6, UDMA, HEMA Liquid: 149 PV Panavia F 2.0 (Kuraray Medical Inc., Okayama, Japan) Paste A: dimethacylates, 10-MDP, camphorquinone, catalysts, initiators, silanated silica filler, silanated colloidal silica Paste A: 00405A Paste B: dimethacrylates, catalysts, accelerators, silanated barium glass filler, sodium fluoride Paste B: 00074A UC RelyX Unicem (3M ESPE) Base: methacrylated phosphoric acid esters, TEGDMA, sodium presulfate, glass powder, silane treated silica Cliker (base/catalyst): 372372 Catalyst: substituted dimethacrylate, sodium p -toluenesulfinate, calcium hydroxide, glass powder, silane treated silica Monomer abbreviations : HEMA, 2-hydroxyethyl methacrylate; UDMA, urethane dimethacrylate; PENTA, dipentaerythritol pentaacrylate monophosphate; Bis-GMA, bisphenol A diglycidyl methacrylate; TEGDMA, triethylene glycol dimethacrylate; MTU-6, 6-methacryloyloxyhexyl 2-thiouracil-5-carboxylate; 10-MDP, 10-methacryloyloxydecyl dihydrogenphosphate. Thirty-two cylinder-shaped (20 mm diameter and 1.5 mm thickness) commercial zirconia ceramic discs (Lava, 3 M ESPE, Seefeld, Germany) were prepared according to the manufacturer's instructions and embedded in round silicone rubber moulds (25.4 mm in inner diameter, 19.0 mm in height) using poly(methyl methacrylate) resin, ensuring that one surface of the zirconia disc remained uncovered for adhering to cement. The exposed surface of each specimen was sandblasted with 110 μm Al 2 O 3 from a distance of 10 mm perpendicular to the specimen surface at a pressure of 0.25 MPa for 13 s and ultrasonically cleaned in isopropyl alcohol for 3 min. The sandblasted and ultrasonically cleaned zirconia ceramic surfaces were examined using an optical microscope (MM-40, Nikon, Tokyo, Japan) at a magnification of 100× and 500×. To prepare hollow cylinders for bonding, clear and flexible polyvinyl chloride tubing (CFT-02-C, Small Parts Inc., Miami Lakes, FL, USA) with an inner diameter of 3.175 mm and an outer diameter of 6.35 mm was used. The tubing was inserted inside a long, clear acyclic split mould (inner diameter: 6.5 mm) so that 1.0 mm height of the tubing protruded perpendicularly from one surface of the mould. The protruding portion was cut off by moving a sharp razor blade (ST-300, Dorco Co., Ltd., Seoul, Korea) along the surface. In this manner, precisely cut tubings (approximately 1.0 mm in height) were obtained. The luting cements were mixed following manufacturers’ instructions and carefully inserted into the tubing lumens, ensuring that the tubing did not move or slip on the surface by pressing the tubing with a custom-made wire loop tool during the procedure. The materials were allowed to self-cure. The PR, PV, and UC cements were additionally irradiated for 20 s by holding the tip of the light guide of a light-curing unit (Elipar TriLight, 3 M ESPE; standard mode, output intensity = 750 mW/cm 2 ) approximately 1 mm above the cement-filled tubing. All specimens were left undisturbed for 30 min in 100% humidity and the tubing around cement cylinders was removed by vertically cutting the tubing into two or more fragments using a sharp blade (KB, Olfa Corp., Osaka, Japan). Special care was taken not to apply any stress or damage to the boned material. Cement cylinders that showed any apparent defects under an optical microscope (SMZ800, Nikon Corp., Tokyo, Japan) were excluded from this study and replaced. In this manner, five bonded cement cylinders of each luting cement were arranged in a circle on one zirconia surface ( n = 10). Prior to debonding, all bonded specimens were stored in distilled water at 37 °C for 48 h. Half of them were additionally thermocycled 10,000 times between 5 °C and 55 °C waterbaths with a dwelling time of 30 s and an exchange time of 5 s between each bath ( n = 10). The bonded specimens were mounted in a jig attached to a universal testing machine (3343, Instron Inc., Canton, MA, USA). A stainless steel wire loop (0.41 mm diameter; G&H Wire Company, Greenwood, IN, USA) was placed around the cement cylinder so that it made contact with the lower half-circle of the cylinder and touched the zirconia surface. Each cylinder was stressed one by one by turning the specimen at a crosshead speed of 1 mm/min and the maximum load at failure was recorded, then converted to MPa. Following debonding, all fractured interfaces were examined under an optical microscope (SMZ800) to determine the mode of fracture. These were classified into one of three types: A, adhesive failure at the ceramic–cement interface; C, cohesive failure within cement; and AC, combination of these failure modes. 2.2 FTIR spectra The FTIR spectra of eight luting cements tested in this study were obtained using a FTIR spectrophotometer (IRPrestige-21, Shimadzu Corp., Kyoto, Japan) with an attenuated total reflectance (ATR) unit (MIRacle, Pike Technologies Inc., Madison, WI, USA). Prior to analysis, an adhesive tape (Rapi, Cemedine Co., Ltd., Tokyo, Japan) was placed around the ATR diamond surface to act a vertical spacer, ensuring standard specimen thickness (approximately 100 μm). A small amount of each freshly mixed material was pressed against the diamond surface and covered with a polyester strip. The absorbance spectrum was acquired by scanning the specimens 10 times over a 4000–700 cm −1 range at a resolution of 4 cm −1 . The spectra were analysed and their peaks were assigned based on references. 2.3 Surface roughness Prior to contact angle measurements, the surface roughness ( R a ) of the luting cements was determined. Cylindrical moulds (10 mm in diameter; 1 mm in height) were placed on a polyester strip (KerrHawe SA, Bioggio, Switzerland) over a glass slide. Each luting cement was filled into the mould, covered with another polyester strip and glass slide, and gently pressed to expel the excess material. The materials were then allowed to self-cure. The PR, PV, and UC cements were additionally irradiated for 20 s by placing the tip of the light guide of the light-curing unit against the upper glass slide. After 30 min, the strips were removed from the specimen, and the R a of each specimen was then measured five times using a previously calibrated profilometer (Surftest SV-400, Mitutoyo Corp., Kawasaki, Japan) at a stylus speed of 0.1 mm/s, a cutoff of 0.8 mm, and a range of 600 μm. The R a of each specimen was recorded as the average of the five readings. The R a values of the zirconia surfaces polished with silicon carbide paper up to #2000 and sandblasted were also measured after optical microscopic examination of the surfaces (MM-40) ( n = 10). 2.4 Surface energy parameters To estimate the surface energy parameters of the eight luting cements and zirconia ceramic, the contact angle method was used. When a drop of liquid is placed on a solid that is assumed to be perfectly smooth and rigid, the angle of contact can be measured at the three phase point between the solid, the liquid, and the vapor. The interfacial tensions between the phases are expressed by the Young equation as follows : γS=γSL+γLcosΘ γ S = γ S L + γ L cos Θ where Θ is the contact angle and γ S , γ SL , and γ L are the surface tensions of the solid, solid–liquid and liquid surfaces, respectively. The Dupré equation states : WA=γS+γL−γSL W A = γ S + γ L − γ S L in which W A is the thermodynamic work of adhesion between a solid surface and a liquid. Substitution of the Young equation into the Dupré equation results in the Young–Dupré equation, which states : WA=γL(1+cosΘ) W A = γ L ( 1 + cos Θ ) The surface energy parameters of the luting cements and zirconia ceramic can be calculated using the Young–Dupré equation in conjunction with Lifshitz-van der Waals/Lewis acid-base (van Oss and Good) theory : γL(1+cosΘ)=2[(γdSγdL)1/2+(γ+Sγ−L)1/2+(γ−Sγ+L)1/2] γ L ( 1 + cos Θ ) = 2 [ ( γ S d γ L d ) 1 / 2 + ( γ S + γ L − ) 1 / 2 + ( γ S − γ L + ) 1 / 2 ] where the superscripts d , +, and − refer to the dispersive (or Lifshitz-van der Waals), acid, and base components, respectively. Cement specimens for contact angle measurement were prepared in the same way as for surface roughness measurement. The zirconia surface was polished with silicon carbide paper up to #2000. The contact angles were then determined by the sessile drop method at room temperature using a contact angle measurement apparatus (OCA 15 plus, DataPhysics Instrument GmbH, Filderstadt, Germany) on the surfaces of the luting cements and polished zirconia ceramic for each of four different test liquids with known surface energy parameters: water, glycerol, ethylene glycol, and 1-bromonaphthalene ( n = 10). By determination of Θ with the four reference liquids with known γ L , γdL,γ−L γ L d , γ L − , and γ+L γ L + , four equations were set up for each of the surfaces from Eq. (4) . Based on the least square method, γdS,γ+S γ S d , γ S + , and γ−S γ S − were calculated using computer software (SCA20, DataPhysics Instrument GmbH). The total surface energy γ S of the luting cements and zirconia ceramic was calculated from the following equation : γS=γdS+2(γ+Sγ−S)1/2 γ S = γ S d + 2 ( γ S + γ S − ) 1 / 2 2.5 Statistical analysis All the statistical analyses were carried out using SPSS 14.0 for Windows (SPSS Inc., Chicago, IL, USA) at a level of significance of α = 0.05. For the bond strength data, non-parametric statistical test procedures were used for group comparisons because Leven's test showed inhomogeneity of variances between groups ( p < 0.001) and the data was found to be non-normally distributed. The Kruskal–Wallis test was employed between luting cements within each thermocycling condition, followed by the Mann–Whitney post hoc test, with adjustment of significance levels using the Sidak correction for multiple testing; between two thermocycling conditions within each luting cement, the Mann–Whitney test was used. Meanwhile, the surface roughness data for the luting cements and polished zirconia ceramic were analysed using one-way ANOVA and Tukey post hoc test. The Spearman rank correlation test was performed to evaluate the significance of the correlation between surface energy parameters and measured shear bond strengths. 3 Results The shear bond strengths of the eight luting cements to sandblasted zirconia ceramic are summarized in Table 2 . Kruskall–Wallis test revealed statistically significant differences amongst the test luting cements both before and after thermocycling ( p < 0.001). FI and KC showed low initial bond strengths and, furthermore, significant drops in the value after thermocycling ( p = 0.001 and p < 0.001, respectively). Though FP showed the highest initial value amongst the four GICs ( p < 0.001), it exhibited a significant decrease in the value after thermocycling ( p < 0.001). PV exhibited the highest initial bond strength amongst all the luting cements tested, the value for PR being statistically similar ( p = 0.106). UC exhibited a significantly lower initial bond strength than PV and PR ( p < 0.001). After thermocycling, PV, PR, and UC exhibited no significant decreases in their respective initial bond strengths ( p > 0.05). IT yielded very low bond strengths both before and after thermocycling. In this study, all debonded specimens showed adhesive failure regardless of luting materials tested or thermocycling conditions. Table 2 Shear bond strengths (MPa) of eight luting cements to sandblasted zirconia ceramic. Luting cement 48-h immersion in water (thermocycling 0) 48-h immersion in water + additional 10,000 thermocycling Median Mean S.D. Median Mean S.D. FI 1.06 Da 0.94 0.36 0.07 Eb 0.06 0.05 KC 1.50 Ca 1.93 0.69 0.00 Eb 0.00 0.00 FP 8.18 Ba 7.87 0.86 3.91 Cb 3.84 0.62 RL 2.92 Ca 2.99 0.60 2.80 Da 2.77 0.45 PR 9.92 Aa 9.67 0.57 9.03 ABa 9.12 1.23 IT 0.13 Ea 0.14 0.12 0.00 Eb 0.00 0.00 PV 11.23 Aa 11.31 1.25 10.82 Aa 10.31 1.63 UC 6.75 Ba 6.83 0.50 7.13 Ba 7.12 1.33 FI, Fuji I; KC, Ketac Cem Easymix; FP: Fuji Plus; RL: RelyX Luting; PR: Principle; IT: Ionotite F; PV: Panavia F 2.0; UC: RelyX Unicem. Within the same column, medians with the same uppercase letter are not statistically different according to the Kruskal–Wallis test followed by the Mann–Whitney post hoc test with adjustment of significance levels using the Sidak correction for multiple testing ( p > 0.05). Within the same row, medians with the same lowercase letter are not statistically different according to Mann–Whitney test ( p > 0.05). Fig. 1 shows the representative FTIR spectra of the freshly mixed luting cements. For the conventional GICs and RMGICs, broad absorption bands indicating hydrogen-bonded O–H stretching vibrations were noted at 3500–3250 and 2700–2500 cm −1 ( Fig. 1 a–d). For the RMGICs (FP and RL), the peak indicating C C stretching vibration (∼1640 cm −1 ) was observed. In contrast, the compomer cements (PR and IT) and resin cements (PV and UC) showed no prominent hydrogen-bonded O–H stretching vibrations ( Fig. 1 e–h). In addition, the latter four luting cements showed phosphorus–oxygen double bond (P O) and phosphorus–oxygen–alkyl (P–O–R) stretching vibrations at ∼1250 and ∼1150 cm −1 , respectively ( Fig. 1 e–h). Fig. 1 FITR spectra of luting cements used in this study: (a) Fuji I, (b) Ketac Cem Easymix, (c) Fuji Plus, (d) RelyX Luting, (e) Principle, (f) Ionotite F, (g) Panavia F 2.0, and (h) RelyX Unicem. References for peak assignment: O–H: Eliades et al. ; C C: Kakaboura et al. ; P O and P–O–R: Verdelis et al. Surface roughness values of the luting cements and zirconia ceramic are presented in Table 3 . The mean R a values for the eight luting cements and polished zirconia ceramic surface ranged from 0.104 to 0.167 μm, with significant differences amongst them (one-way ANOVA, p = 0.003). Sandblasting substantially increased the surface roughness of polished zirconia ceramic. Table 3 Mean R a surface roughness (μm) (S.D.) of luting cements and zirconia ceramic. FI KC FP RL PR IT PV UC Polished zirconia Sandblasted zirconia 0.104 (0.025) A 0.121 (0.030) AB 0.167 (0.045) B 0.115 (0.032) A 0.129 (0.036) AB 0.130 (0.043) AB 0.135 (0.031) AB 0.132 (0.034) AB 0.107 (0.013) A 0.784 (0.081) FI, Fuji I; KC, Ketac Cem Easymix; FP: Fuji Plus; RL: RelyX Luting; PR: Principle; IT: Ionotite F; PV: Panavia F 2.0; UC: RelyX Unicem. The same uppercase letter indicates lack of statistically difference amongst the luting cements and polished zirconia based on one-way ANOVA and Tukey post hoc tests ( p > 0.05). Tables 4 and 5 summarize the contact angles, total surface energies, and surface energy parameters of the four liquids measured on luting cements and polished zirconia ceramic. The calculated mean total surface energy of the zirconia ceramic was 36.64 mJ/m 2 ; the mean values for the eight luting cements varied from 27.10 to 38.79 mJ/m 2 . Examining a correlation between individual surface energy parameters for the luting cements and the shear bond strengths, we found significant effects of the base component on the bond strengths both before and after thermocycling ( r s = 0.857, p = 0.007 and r s = 0.826, p = 0.011, respectively) ( Fig. 2 ). Table 4 Mean contact angles (°) (S.D.) of four liquids measured on luting cements and polished zirconia ceramic. Surface Water Glycerol Ethylene glycol 1-Bromonaphthalene FI 72.2 (2.3) 83.8 (4.4) 76.7 (3.2) 52.6 (3.0) KC 86.7 (7.2) 84.9 (7.2) 66.1 (6.4) 57.1 (5.1) FP 63.3 (1.5) 85.7 (3.9) 64.1 (4.2) 43.4 (1.9) RL 73.7 (4.2) 84.0 (4.4) 71.7 (7.7) 44.5 (4.2) PR 66.5 (3.2) 80.9 (4.3) 61.9 (2.6) 50.6 (2.0) IT 88.8 (6.8) 87.3 (8.5) 76.0 (1.8) 46.5 (5.7) PV 67.3 (5.5) 85.8 (4.3) 59.7 (2.4) 47.1 (3.1) UC 71.1 (4.0) 88.5 (2.2) 64.5 (3.3) 44.0 (2.1) Zirconia 52.7 (3.4) 78.4 (3.4) 60.5 (2.1) 45.9 (1.0) FI, Fuji I; KC, Ketac Cem Easymix; FP: Fuji Plus; RL: RelyX Luting; PR: Principle; IT: Ionotite F; PV: Panavia F 2.0; UC: RelyX Unicem. Table 5 Total surface energy and surface energy parameters (mJ/m 2 ) of luting cements and zirconia ceramic [mean (S.D.)]. Surface γ S (total surface energy) γdS γ S d (dispersive component) γ+S γ S + (acid component) γ−S γ S − (base component) FI 32.10 (4.90) 28.06 (1.45) 0.18 (0.29) 23.08 (1.58) KC 27.10 (8.54) 26.69 (2.54) 0.00 (0.14) 8.80 (1.74) FP 38.79 (3.66) 33.05 (0.87) 0.26 (0.24) 31.72 (1.64) RL 35.34 (6.93) 32.41 (1.94) 0.12 (0.40) 17.95 (1.62) PR 31.18 (5.47) 29.64 (0.97) 0.02 (0.13) 26.78 (1.50) IT 33.63 (7.63) 31.21 (2.63) 0.25 (0.98) 5.85 (1.28) PV 34.80 (8.84) 31.58 (1.46) 0.09 (0.41) 28.76 (1.79) UC 37.78 (5.83) 32.85 (0.97) 0.26 (0.51) 23.11 (0.96) Zirconia 36.64 (5.57) 31.86 (0.47) 0.14 (0.28) 42.28 (1.46) FI, Fuji I; KC, Ketac Cem Easymix; FP: Fuji Plus; RL: RelyX Luting; PR: Principle; IT: Ionotite F; PV: Panavia F 2.0; UC: RelyX Unicem. Fig. 2 Scatter plot of Spearman correlation analysis results: (a) 48-h immersion in water (thermocycling 0) and (b) 48-h immersion in water + additional 10,000 thermocycling. 4 Discussion A passive film of zirconium oxide easily covers the zirconium surface. As chemical characteristics of zirconia ceramic surface are similar to those of metal surface, hydroxyl (O–H) groups may be present on the zirconia ceramic surface. It is also likely that a luting cement containing any polymers or monomers with polar functional groups is capable of bonding chemically with polar hydroxyl groups on the zirconia ceramic surface. Therefore, all the luting cements tested in this study were expected to have the potential to bond chemically to zirconia ceramic due to their respective functional polymers (e.g. polyalkenoic acid copolymer) or monomers [e.g. 2-hydroxyethyl methacrylate (HEMA), carboxylic acid monomer, and an organophosphate ester monomer such as dipentaerythritol pentaacrylate monophosphate (PENTA) and 10-MDP] ( Table 1 ). Assuming a minimal bond strength requirement of approximately 10–13 MPa for acceptable clinical bonding, the shear bond strengths for all the luting cements to “sandblasted” zirconia were not high enough to ensure favourable clinical service ( Table 2 ). In addition, regardless of the luting cements tested or the thermocycling conditions, all debonded specimens showed only adhesive failure at the cement–zirconia interface. These findings suggest that the bond formed at the interface might be limited to weak intermolecular attractions such as hydrogen bonding between the polar functional groups of the polymers or monomers in the cement and the polar hydroxyl groups on the zirconia, in tandem with micromechanical retention of the roughened zirconia ceramic surface by sandblasting. To estimate the bonding potential of the luting cements to zirconia ceramic, we calculated surface energy parameters of the materials using a contact angle method. The surface energy parameters were determined on cured cements in this study, though uncured material is applied to the zirconia surface clinically. It can be assumed that monomer or polymer surface energy parameters of the materials are similar in value to those of the cured materials because they contain the same molecular groups and segments, with the exception of the vinyl group. In contact angle methods, moreover, the solid is assumed to be perfectly smooth and rigid. However, changes in surface roughness can influence the contact angle, thereby changing surface energy characteristics. Therefore, we investigated the surface roughness of the luting cements and the polished zirconia ceramic prior to contact angle measurements of the materials. Despite significant differences in the R a amongst the materials, there were only small variations in the value ( Table 3 ). Busscher et al. reported that changes in solid surface R a below 0.1 μm have no effect on contact angle. Given that the range of mean R a values of the luting cements and polished zirconia ceramic were below 0.1 μm in the present study, R a may be assumed to have no significant effect on the contact angle used to calculate the surface energy parameters. The Lifshitz-van der Waals/Lewis acid–base approach entails decomposing the surface energy into Lifshitz-van der Waals interactions and Lewis acid–base interactions. In the present study, the base component (γ−S) ( γ S − ) of the polished zirconia ceramic surface was larger than the acid component (γ+S) ( γ S + ) ( Table 5 ). When the Lewis acid component (γ+S) ( γ S + ) and the Lewis base component (γ−S) ( γ S − ) are considered in the context of a hydrogen bond, the former and the latter may be viewed as the hydrogen bond donating (HBD) component and the hydrogen bond accepting (HBA) component, respectively. The zirconia ceramic surface thus appears to have a predominantly HBA character. Similarly, all luting cements tested in this study also showed large base components and very small acid components, indicating a primarily HBA nature. The Spearman rank correlation test between surface energy parameters and measured shear bond strengths showed significant influence of the base component on bond strengths both before and after thermocycling as opposed to the acid component, dispersive component, and total surface energy ( Fig. 2 ). This suggests that the HBA (base) component of the luting cements interacts with the HBD (acid) component of the zirconia surface. Although the HBA component of the zirconia surface is great, some of the surface appears to bear an HBD character. As shown in the correlation test, increasing the HBA character of the luting cement results in increased interaction with the HBD zirconia ceramic surface. GICs contain a large amount of water, as shown in the FTIR spectra (3500–3250 cm −1 , Fig. 1 ). Polyalkenoic acid copolymers that remain unionized in water may bear two sites potentially available for hydrogen boding with zirconia: oxygen atoms of carbonyl moieties and hydrogen atoms of hydroxyl groups. The surface energy analysis results suggest the role of the former might be predominant as a hydrogen bond acceptor rather than the latter. It is also possible that HEMA also participates in hydrogen bonding with zirconia due to the presence of a polar head (hydroxyl group). A small amount of dimethacrylate monomer incorporated into RMGICs form crosslinked poly-HEMA during polymerization. As for polyalkenoic acid copolymers, poly-HEMA is hydrophilic and tends to absorb water, leading to plasticity and hygroscopic expansion. Although not recommended for cementation of most all-ceramic restorations, mainly due their suspected role in crack initiation, RMGICs can be indicated for luting of all-zirconia strengthened core ceramic restorations. A significantly higher bond strength for the FP compared to the other three GICs could be partly attributed to the incorporation of the less hydrophilic dimethacrylat monomer urethane dimethacrylate (UDMA) ( Table 1 ). However, UDMA also tends to absorb water due to the presence of hydrophilic urethane groups in its structure. Thus, RMGICs as well as conventional GICs do not seem to impart high bonding durability to zirconia though some of them share a high HBA character ( Tables 2 and 5 ). In this study, PV produced the highest initial bond strength ( Table 2 ), in accordance with the results of previous studies. Panavia products contain a well-known organophosphate ester monomer, 10-MDP, which is structurally a phosphate monoester having one P O and two hydroxyl groups in one molecule. As PV contains little water, most of the 10-MDP seems to remain undissociated notwithstanding its acidic nature. Similar to the carboxyl groups of the polyalkenoic acid copolymers, the hydroxyl groups and P O oxygen are likely to participate as hydrogen bonding acceptors with the zirconia surface. Unlike the polyalkenoic acid copolymers and HEMA in GIC materials, the apolar spacer group consisting of a long saturated carbon chain in 10-MDP might keep water molecules at a distance from the hydrogen bond formed at the cement–zirconia interface. UC, which contains another organophosphate ester monomer, produced a significantly lower initial bond strength than PV. This may be attributable to a difference in the HBA component between the two materials ( Table 5 ). It seems that hydrogen bonding by the phosphate ester monomers may also depend on the chemical structure of the monomer. The two compomer cements produced greatly differing initial bond strengths from each other: PR showed a bond strength statistically comparable to PV, whereas IT exhibited a very poor value. As represented in the FTIR spectra ( Fig. 1 ), the two materials might fall into the category of resin luting cements, although they are sometimes classified as RMGICs. PR showed a considerably larger HBA component than IT ( Table 5 ). The high bond strength of PR before and after thermocycling might be partly attributable to the combined effect of the two types of functional monomers (PENTA and carboxylic acid monomer, Table 1 ), which could heighten the HBA capability. It must be noted that the shear bond strengths of the luting cements do not indicate genuine values of chemical bonding because the materials were bonded to a sandblasted zirconia ceramic surface. In this study, bond strengths were determined using sandblasted zirconia ceramic surfaces, whereas surface energy analysis was performed on polished ones. Preliminary bonding tests performed on polished zirconia ceramic surfaces showed frequent failure during bonded specimen preparation, making differentiation amongst the groups impossible. Although there is still some controversy about the effect, sandblasting followed by ultrasonic cleansing is believed not only to increase surface roughness and bonding surface area but also to remove any contaminants from the ceramic surface. The surface area of sandblasted zirconia available for chemical bonding with the luting cements was relatively constant ( R a = 0.784 ± 0.081 μm, Table 3 ), possibly allowing a correlation between bond strengths and surface energy parameters. It should also be noted that only one sandblasting condition (with 110 μm Al 2 O 3 at 0.25 MPa) was tested in the current study. Kern et al. suggested that low-pressure (0.05 MPa) sandblasting with 50 μm Al 2 O 3 prior to priming was effective in obtaining durable resin bonding to zirconia ceramics (Cercon, DeguDent), reduction in surface roughness notwithstanding. In contrast, Ntala et al. reported that sandblasting with 50 μm Al 2 O 3 at 0.25 MPa did not produce sufficient micromechanical retention on zirconia ceramics (Kavo Everest, Kavo) for adhesive bonding. Therefore, various sandblasting conditions of zirconia ceramic should be further tested in conjunction with variations in Y-TZP composition to determine the optimal surface for bonding with luting cements. The present study provides laboratory data on the performance of a zirconia-based restoration luted with various cements. At present, there seems to be no definite thermocycling condition for predicting clinical performance of luting cements to zirconia ceramic. Previous zirconia ceramic studies show great variability in number of thermal cycles (for example, 5000 cycles, 10,000 cycles, and 20,000 cycles ). In Kern and Wegner and Kern et al. bonded zirconia ceramic specimens were stored in water 150 days and then thermal cycled 37,500 times to stress the bond interface. Gale and Darvell advise that the poor discriminatory power of excessive stresses may lead to unreasonable characterisation of materials which could in fact serve well in clinical practice. We used 10,000 thermal cycles to statistically distinguish the bonding performance of the luting cements, although this condition may not directly translate into clinical results. This in vitro study suggests that surface characteristics of luting cements may be key determinants of their bonding potential with zirconia ceramic, as shown in Fig. 2 . The Spearman rank correlation test reveals a strong, though not perfect, linear relationship between the HBA (base) component and bond strength ( r s = 0.857 and r s = 0.826 before and after thermocycling, respectively), so that other factors may also, to a small extent, affect bonding properties. In addition, more precise surface energy parameter approaches and experimental set-ups may better predict the adhesive potential of luting cements. Although this study only tested adhesive resin cements, commercially available non-adhesive resin cements lack potential for chemical interaction with zirconia ceramic and thus require surface pre-treatment to achieve strong and durable bonding. Recent approaches to improving bonding with zirconia include the use of primers, including one containing organophosphate and carboxylic acid monomers. When these primers are applied to a zirconia ceramic surface, eliminating hydrophilic constituents such as water can improve the bond strength of composite resin to the ceramic. Such approaches might also benefit from utilising surface energy analyses to predict bonding behaviour of primers to zirconia ceramic based on application method. In particular, these analyses should consider surface roughness in the context of methods that alter the surface topography of zirconia ceramic surface (e.g., selective infiltration etching, the application of hot chemical etching solution, and the application of a glass–ceramic/glaze material containing a major lithium disilicate phase ). Along with bonding properties, mechanical properties of luting cements should be considered in designing zirconia-based restorations with good clinical performance during service. In addition, optical properties of various luting cements have to be considered for clinically optimal aesthetic results. Further research is plainly required to establish guidelines on conventional and adhesive cementation in specific clinical situations. 5 Conclusions The resin cement Panavia F 2.0 and the compomer cement Principle bonded to sandblasted zirconia gave clinically acceptable shear bond strengths both before and after thermocycling, whereas bond strength values of four GICs were too low for acceptable clinical bonding. Measured bond strengths both before and after thermocycling showed significant positive correlations with the base (hydrogen bond accepting) components of the luting materials. Acknowledgement This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology ( 2010-0029494 ). References 1. Derand T., Molin M., Kvam K.: Bond strength of composite luting cement to zirconia ceramic surfaces. Dental Materials 2005; 21: pp. 1158-1162. 2. Kern M., Wegner S.M.: Bonding to zirconia ceramic: adhesion methods and their durability. Dental Materials 1998; 14: pp. 64-71. 3. Yoshida K., Yamashita M., Atsuta M.: Zirconate coupling agent for bonding resin luting cement to pure zirconium. American Journal of Dentistry 2004; 17: pp. 249-252. 4. Piwowarczyk A., Lauer H.C., Sorensen J.A.: The shear bond strength between luting cements and zirconia ceramics after two pre-treatments. Operative Dentistry 2005; 30: pp. 382-388. 5. Matinlinna J.P., Heikkinen T., Özcan M., Lassila L.V., Vallittu P.K.: Evaluation of resin adhesion to zirconia ceramic using some organosilanes. Dental Materials 2006; 22: pp. 824-831. 6. Wegner S.M., Kern M.: Long-term resin bond strength to zirconia ceramic. Journal of Adhesive Dentistry 2000; 2: pp. 139-147. 7. Aboushelib M.N., Matinlinna J.P., Salameh Z., Ounsi H.: Innovations in bonding to zirconia-based materials: Part I. Dental Materials 2008; 24: pp. 1268-1272. 8. Roediger M., Gersdorff N., Huels A., Rinke S.: Prospective evaluation of zirconia posterior fixed partial dentures: four-year clinical results. International Journal of Prosthodontics 2010; 23: pp. 141-148. 9. Ntala P., Chen X., Niggli J., Cattell M.: Development and testing of multi-phase glazes for adhesive bonding to zirconia substrates. Journal of Dentistry 2010; 38: pp. 773-781. 10. Örtorp A., Kihl M.L., Carlsson G.E.: A 3-year retrospective and clinical follow-up study of zirconia single crowns performed in a private practice. Journal of Dentistry 2009; 37: pp. 731-736. 11. Rosentritt M., Steiger D., Behr M., Handel G., Kolbeck C.: Influence of substructure design and spacer settings on the in vitro performance of molar zirconia crowns. Journal of Dentistry 2009; 37: pp. 978-983. 12. Hill E.E.: Dental cements for definitive luting: a review and practical clinical considerations. Dental Clinics of North America 2007; 51: pp. 643-658. vi 13. Asmussen E., Peutzfeldt A.: Resin composites: strength of the bond to dentin versus surface energy parameters. Dental Materials 2005; 21: pp. 1039-1043. 14. Phark J.H., Duarte S., Blatz M., Sadan A.: An in vitro evaluation of the long-term resin bond to a new densely sintered high-purity zirconium-oxide ceramic surface. Journal of Prosthetic Dentistry 2009; 101: pp. 29-38. 15. Glasspoole E.A., Erickson R.L., Davidson C.L.: Effect of surface treatments on the bond strength of glass ionomers to enamel. Dental Materials 2002; 18: pp. 454-462. 16. Lüthy H., Loeffel O., Hammerle C.H.: Effect of thermocycling on bond strength of luting cements to zirconia ceramic. Dental Materials 2006; 22: pp. 195-200. 17. Sadr A., Ghasemi A., Shimada Y., Tagami J.: Effects of storage time and temperature on the properties of two self-etching systems. Journal of Dentistry 2007; 35: pp. 218-225. 18. Arrais C.A., Giannini M., Rueggeberg F.A.: Effect of sodium sulfinate salts on the polymerization characteristics of dual-cured resin cement systems exposed to attenuated light-activation. Journal of Dentistry 2009; 37: pp. 219-227. 19. Kakaboura A., Eliades G., Palaghias G.: An FTIR study on the setting mechanism of resin-modified glass ionomer restoratives. Dental Materials 1996; 12: pp. 173-178. 20. Verdelis K., Eliades G., Oviir T., Margelos J.: Effect of chelating agents on the molecular composition and extent of decalcification at cervical, middle and apical root dentin locations. Endodontics and Dental Traumatology 1999; 15: pp. 164-170. 21. Eliades T., Eliades G., Silikas N., Watts D.C.: Orthodontic bonding to wet enamel with water-insensitive and water-activated orthodontic adhesive resins.Eliades G.Watts D.C.Eliades T.Dental hard tissues and bonding: interfacial phenomena and related properties.2005.SpringerBerlin/Heidelberg:pp. 71-85. 22. Schrader M.E.: Work of adhesion of a sessile drop to a clean surface. Journal of Colloid and Interface Science 1999; 213: pp. 602-605. 23. Pocius A.V.: Adhesion and adhesive technology: an introduction.2nd ed.2002.Hanser Gardner Publications Inc.Cincinnati p. 132–63 24. Ratner B.D.: Surface structure and properties.Williams D.Cahn R.W.Bever M.B.Concise encyclopedia of medical & dental materials.1990.PergamonOxford:pp. 337-346. 25. Asmussen E., Attal J.P., Degrange M.: Factors affecting the adherence energy of experimental resin cements bonded to a nickel–chromium alloy. Journal of Dental Research 1995; 74: pp. 715-720. 26. Lee L.H.: Adhesion and surface-hydrogen-bond components for polymers and biomaterials. Journal of Adhesion 1998; 67: pp. 1-18. 27. Chen H., Li L., Jones P.M., Hsia Y.T., Jhon M.S.: Lifshitz-van der Waals and Lewis acid–base approach for analyzing surface energy of molecularly thin lubricant films. IEEE Transactions on Magnetics 2007; 43: pp. 2226-2228. 28. Yoshida K., Tsuo Y., Atsuta M.: Bonding of dual-cured resin cement to zirconia ceramic using phosphate acid ester monomer and zirconate coupler. Journal of Biomedical Materials Research Part B: Applied Biomaterials 2006; 77: pp. 28-33. 29. Yun J.Y., Ha S.R., Lee J.B., Kim S.H.: Effect of sandblasting and various metal primers on the shear bond strength of resin cement to Y-TZP ceramic. Dental Materials 2010; 26: pp. 650-658. 30. Kern M., Barloi A., Yang B.: Surface conditioning influences zirconia ceramic bonding. Journal of Dental Research 2009; 88: pp. 817-822. 31. Vaidyanathan T.K., Vaidyanathan J.: Recent advances in the theory and mechanism of adhesive resin bonding to dentin: a critical review. Journal of Biomedical Materials Research Part B: Applied Biomaterials 2009; 88: pp. 558-578. 32. Fabianelli A., Pollington S., Papacchini F., Goracci C., Cantoro A., Ferrari M., et. al.: The effect of different surface treatments on bond strength between leucite reinforced feldspathic ceramic and composite resin. Journal of Dentistry 2010; 38: pp. 39-43. 33. Quirynen M., Bollen C.M.: The influence of surface roughness and surface-free energy on supra- and subgingival plaque formation in man. A review of the literature. Journal of Clinical Periodontology 1995; 22: pp. 1-14. 34. Bollen C.M., Papaioanno W., Van Eldere J., Schepers E., Quirynen M., van Steenberghe D.: The influence of abutment surface roughness on plaque accumulation and peri-implant mucositis. Clinical Oral Implants Research 1996; 7: pp. 201-211. 35. Ahn H.B., Ahn S.J., Lee S.J., Kim T.W., Nahm D.S.: Analysis of surface roughness and surface free energy characteristics of various orthodontic materials. American Journal of Orthodontics and Dentofacial Orthopedics 2009; 136: pp. 668-674. 36. Busscher H.J., van Pelt A.W.J., de Boer P., de Jong H.P., Arends J.: The effect of surface roughening of polymers on measured contact angles of liquids. Colloids and Surfaces 1984; 9: pp. 319-331. 37. Steed J.W., Turner D.R., Wallace K.J.: Core concepts in supramolecular chemistry and nanochemistry.1st ed.2007.John Wiley & Sons, Ltd.Chichester p. 1–27 38. McCabe J.F.: Resin-modified glass-ionomers. Biomaterials 1998; 19: pp. 521-527. 39. Smith D.C.: Development of glass-ionomer cement systems. Biomaterials 1998; 19: pp. 467-478. 40. Van Landuyt K.L., Snauwaert J., De Munck J., Peumans M., Yoshida Y., Poitevin A., et. al.: Systematic review of the chemical composition of contemporary dental adhesives. Biomaterials 2007; 28: pp. 3757-3785. 41. Good M.L., Mitchell C.A., Pintado M.R., Douglas W.H.: Quantification of all-ceramic crown margin surface profile from try-in to 1-week post-cementation. Journal of Dentistry 2009; 37: pp. 65-75. 42. Van Landuyt K.L., Yoshida Y., Hirata I., Snauwaert J., De Munck J., Okazaki M., et. al.: Influence of the chemical structure of functional monomers on their adhesive performance. Journal of Dental Research 2008; 87: pp. 757-761. 43. Taher N.M., Ateyah N.Z.: Shear bond strength of resin modified glass ionomer cement bonded to different tooth-colored restorative materials. Journal of Contemporary Dental Practice 2007; 8: pp. 25-34. 44. Wolfart M., Lehmann F., Wolfart S., Kern M.: Durability of the resin bond strength to zirconia ceramic after using different surface conditioning methods. Dental Materials 2007; 23: pp. 45-50. 45. Gale M.S., Darvell B.W.: Thermal cycling procedures for laboratory testing of dental restorations. Journal of Dentistry 1999; 27: pp. 89-99. 46. Casucci A., Osorio E., Osorio R., Monticelli F., Toledano M., Mazzitelli C., et. al.: Influence of different surface treatments on surface zirconia frameworks. Journal of Dentistry 2009; 37: pp. 891-897. 47. Magne P., Paranhos M.P., Burnett L.H.: New zirconia primer improves bond strength of resin-based cements. Dental Materials 2010; 26: pp. 345-352. 48. Chang J., Da Silva J.D., Sakai M., Kristiansen J., Ishikawa-Nagai S.: The optical effect of composite luting cement on all ceramic crowns. Journal of Dentistry 2009; 37: pp. 937-943.

Related Articles

Leave A Comment?

You must be logged in to post a comment.