Bonding strengths to porcelain: An in vitro study of ultrasonic and conventional tooth preparation and etching.

OBJECTIVE: To compare the bond strength of porcelain to enamel and dentin preparations finished with either ultrasonic instruments or diamond burs, with or without acid etching.
MATERIALS AND METHODS: A total of 140 extracted bovine incisor teeth were divided into two groups, enamel and dentin. A split tooth model was employed: each tooth having a randomly allocated combination of diamond burs, ultrasonic instrument, etching, and no etching treatments on their labial surface. Lithium disilicate glass ceramic specimens were bonded to the prepared surfaces and tensile testing was carried out. Force was applied to failure, and linear mixed models were used to analyze the data.
RESULTS: Etching increased the bond strength of the dentin samples by 7.63 MPa (P < 0.001). When bonding to dentin, ultrasonic instrument preparation produced a 2.88 MPa greater bond strength compared to burs (P < 0.001).
CONCLUSION: Ultrasonic preparation produced greater bond strengths compared to rotary instruments. Etching improved the bond strength of all preparations, with enamel having consistently greater bonding values than dentin.

INTRODUCTION

For many years, high-speed dental handpieces have remained the instruments of choice for cavity preparations.[1] Unfortunately, this preparation technique utilizes rotational motion to cut dental tissues, resulting in a higher chance of iatrogenic damage to adjacent teeth or gingiva if meticulous care is not practiced.[2] It has been reported that 70%–95% of proximal cavities result in iatrogenic damage to the adjacent dentition.[23] Ultrasonic instruments were discovered in 1880 and operated based on magnetostrictive or piezoelectric principles. Since the 1950s, ultrasound has been recommended for the removal of small proximal carious lesions in both anterior and posterior teeth as it is possible to achieving a more conservative preparation.[4] Ultrasonic instruments do not cut tooth structure like a rotary instrument but use a frequency range from 20 to 40 kHz to abrade the tooth surface.[5]

When cutting tooth structure, a smear layer can be formed, reducing the bond strength of resin materials to tooth structure. Dentinal tubule openings can be occluded when a smear layer is present, and smear plugs also extend into the tubules.[6] Ultrasonic instruments are able to remove dentin with minimal smear layer and tubular obstruction.[23] Rotary instruments produce a smear layer, and if not removed, poor penetration of the resin into the tubules will result. Most literature supports the use of phosphoric acid with a concentration of 30%–40% to remove the smear layer and to provide sufficient depth of resin penetration.[789] However, etching contributes to the widening of dentinal tubules by demineralizing the intact dentinal matrix. During bonding procedures, some tubules may not be entirely sealed by adhesive resin, resulting in a microporous zone where hydrodynamic shifts in the tubules cause postoperative sensitivity. It is believed that ultrasonic instruments help to reduce the need for etching, bringing about a decrease in postoperative sensitivity.[23]

This research aims to investigate the bonding implications of preparations using ultrasonic rather than rotary instruments. The bond strength to enamel and dentin will be measured, and the effect of etching on the bond strength of porcelain to dentin will also be investigated. The first hypothesis was that the bonding strength of porcelain to enamel would be greater than to dentin preparations finished with the two types of instruments. The second hypothesis was that acid etching would have a beneficial effect on the bond strength of dentin to porcelain. Finally, it was hypothesized that ultrasonic preparation would increase bond strength independent of the use of acid etching.

MATERIALS AND METHODS

A total of 140 freshly extracted bovine permanent incisors were collected. Examination under a light microscope (×20) revealed no cracks, caries, or other defects. They were cleaned under running water and adherent tissues removed, then placed in a solution of 0.2% chlorhexidine for disinfection at 37°C for 1 week, and then stored in distilled water at room temperature. All surface preparations and tensile testing were carried out within 5 months in accordance with ISO 11405.

The teeth were embedded in acrylic (Castapress, Vertex-Dental B.V., Zeist, The Netherlands) [Table 1] with the labial side of the anatomical crown facing upward. The samples were divided into two groups: enamel and dentin with sample sizes of 60 and 80 teeth, respectively. The enamel group consisted of control (n = 20 teeth), ultrasonic instrument group (n = 20 teeth), and rotary instrument group (n = 20 teeth). The three dentin groups had been further divided into subgroups with each being etched or nonetched [Figure 1]. A split tooth model was employed: each tooth had two preparations prepared on its labial surface in accordance with the group in which they were randomly assigned.

Table 1

Materials, product name, and manufacturer information

Materials	Product name	Manufacturer
Acrylic	Castapress	Vertex-Dental B.V., Zeist, The Netherlands
End cutting burs	Tissue Guard End-Cutting bur (fine grit 60 µm)	Premier Two Striper, OH, PA, USA
	Tissue Guard End-Cutting bur (very fine grit 45 µm)	Premier Two Striper, OH, PA, USA
Glass ionomer cement	GC Fuji IX GP	GC Corporation, Tokyo, Japan
Shoulder kit	Satelec Perfect Margin PMS 1 (76 µm)	Satelec, Merignac, France
	Satelec Perfect Margin PMS 2 (46 µm)	Satelec, Merignac, France
	Satelec Perfect Margin PMS 3 (no grit)	Satelec, Merignac, France
Bonding system	Syntac System	Ivoclar Vivadent AG, Schaan, Liechtenstein
Phosphoric acid (37%)	Total etch	Ivoclar Vivadent AG, Schaan, Liechtenstein
Lithium disilicate glass ceramic	IPS e.max CAD	Ivoclar Vivadent AG, Schaan, Liechtenstein
Silane coupling agent	Monobond	Ivoclar Vivadent AG, Schaan, Liechtenstein
Resin luting agent	Variolink Veneer Esthetic	Ivoclar Vivadent AG, Schaan, Liechtenstein
Light cure	Bluephase style	Ivoclar Vivadent AG, Schaan, Liechtenstein

PMS: Perfect Margin Shoulder,CAD: Computer Aid Design, IPS: Ivoclar Porcelain System

Figure 1

Samples distribution

The enamel group size of 40 preparations for each method (none, rotary, and ultrasonic) provided 80% power to detect differences of 0.7 standard deviation (SD) (“large” effects) between any pair of treatments using a two-sided test at the 0.05 level. For the dentin groups for the rotary and ultrasonic methods, 32 preparations each would allow detecting differences of around 0.5 SD (“medium” effects) between each in the main effects model. For the etch and nonetch groups, the detectable effect size in a main effects model would also be 0.5 SD.

The surface enamel was made uniform with silicone carbide abrasive paper (grit size 600). This was monitored under a light microscope (×20) to ensure the preparations remained within enamel. The enamel rotary instrument preparation group was then finished using an end-cutting bur (tissue guard end-cutting [TGE] bur, fine grit [60 μm], Premier Two Striper, Plymouth Meeting, PA, USA) in a high-speed handpiece with water spray for 30 s, followed by a TGE very fine grit (45 μm) bur [Figure 2] (TGE bur, very fine grit [45 μm], Premier Two Striper, OH, PA, USA) for 60 s. The second enamel preparation was finished using a Satelec Perfect Margin Shoulder (PMS) Kit with a factory calibrated ultrasonic generator (P5XS Newtron, Satelec, Merignac, France) with water spray. The PMS 1 tip (76 μm grit) was used for 30 s, followed by the PMS2 tip (46 μm grit) for 60 s and then the PMS3 tip (no grit) for 120 s, each at the recommended power setting [Figure 3]. A jig of 200 g was used during all finishing procedures to ensure the cutting instrument was always perpendicular to the tooth surface and to ensure constant pressure.

Figure 2

Scanning electron microscope image of rotary instrument tissue guard end-cutting (bur) (×65). Only the tip is uniformly coated with diamond particles. The extremity of the shaft is also beveled inward to avoid the production of any undercuts during placement of finishing lines. (a) Scanning electron microscope image of tissue guard end-cutting bur, grit 60 μm, (b) scanning electron microscope image of tissue guard end-cutting bur, grit 45 μm

Figure 3

Scanning electron microscope image of ultrasonic instrument (×50). For perfect margin shoulder 1 and 2 the entire surface of the instrument is uniformly coated with diamond particles. Perfect margin shoulder 3 is uncoated. (a) Scanning electron microscope image of perfect margin shoulder 1, 76 μm grit. (b) Scanning electron microscope image of perfect margin shoulder 2, 46 μm grit. (c) Scanning electron microscope image of perfect margin shoulder 3, no grit

A 200 grit silicone carbide paper was used to remove the enamel from each tooth, while a 600 grit silicone carbide paper was used to expose the superficial dentin and create a standardized smear layer characteristic. Glass ionomer cement (GC Fuji IX GP, GC Corporation, Tokyo, Japan) was placed as a barrier to isolate the two preparation sites on each tooth to avoid contamination by etchant. Finishing was carried out with either ultrasonic instruments or end cutting burs following random assignment. The sequence of surface finishing with ultrasonic instruments and rotary instruments in the dentin group was duplicated as before.

The enamel group was then fully dried, and the preparations underwent acid etching before priming and bonding. The control group (no finishing intervention) was etched followed by priming and bonding. A three-step etch, prime, and bond technique was carried out as per the manufacturer's recommendation (Syntac System, Ivoclar Vivadent AG, Schaan, Liechtenstein). The dentin group was rewetted with a moist cotton pellet to avoid desiccation. For samples not undergoing etching, primer was applied directly to the buccal surface of moist dentin. Subsequent steps of priming and bonding were carried out according to the manufacturer's recommendations (Syntac System, Ivoclar Vivadent).

Two hundred and eighty 4 mm × 4 mm lithium disilicate glass ceramic segments (IPS e.max CAD, Ivoclar Vivadent) were prepared with a slow-speed diamond-coated saw (LECO VC-50 Model 801–900, LECO Corporation, St. Joseph, MI, USA). The specimens were etched (IPS Ceramic Etching Gel, Ivoclar), silanated (Monobond Plus, Ivoclar), and cemented to the preparations at room temperature with resin luting agent (Variolink Veneer Esthetic, Ivoclar) following the manufacturer's recommendations. A constant force was applied with a custom-made device to ensure homogenous cement film thickness. The same operator for all 280 sites repeated this process.

Tensile testing was carried out using a tester (TA. HD Plus Texture Analyser, Stable Micro Systems, Godalming, UK) fitted with a 250 kg load cell. The tooth specimens were secured on the lower grip using a custom acrylic jig. The porcelain specimens were tested to failure at an extension rate of 0.5 mm/s. Force-extension data for each specimen were recorded with plotting software (Exponent software, Stable Micro Systems) and force at failure was also recorded.

Bond strength was described for both surface types, broken down by treatment. Linear mixed models were used to compare bond strength between treatment groups, including a random effect for the teeth to accommodate the repeated measures. For the dentin model only, an interaction between etching and preparation was included initially but only retained if statistically significant, with the main effects model used otherwise. For both dentin and enamel models, the position of the test surface on the tooth was included as a covariate. Standard model diagnostics were performed including examining histograms of conditional residuals and scatter plots of such residuals against fitted values. Log transformations were investigated where there was evidence of positive skew and/or heteroscedasticity and retained where they improved model residuals. Analyses were conducted in Stata 14.1 (StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP, Texas, USA), and a two-sided test P < 0.05 was considered statistically significant. Post hoc tests were only performed when the overall test was statistically significant, and no further adjustments for multiple comparisons were made.

RESULTS

Table 2 shows the relationship between all-etched and nonetched groups for the dentin-bonded groups, while Figure 4 compares the enamel and dentin bonding for the three interventions in the etched groups. In the dentin group, there was no evidence of an etching–preparation interaction (P = 0.804), and hence, the main effects models were investigated. In these, etching increased bond strength by 7.63 MPa (95% confidence interval [CI] 6.02, 9.24; P < 0.001). For preparation, there was strong evidence for a difference between the three groups (P < 0.001), but the only significant post hoc comparison was ultrasonic versus rotary (2.88 MPa higher; 95% CI 1.58, 4.18; P < 0.001) [Figure 4 and Table 2].

Table 2

Summary of results including: mean, standard deviation, confidence interval 95%, and value for all examined groups

	Dentin									Enamel
	No etch			Etch			Overall between treatment effects			Etch			Between treatment effects
	n	Mean	SD	n	Mean	SD	Difference	95% CI	P	n	Mean	SD	Difference	95% CI	P
No intervention	16	11.04	4.52	15	17.73	5.74	0.88	−2.26-4.03	0.582	36	42.48	16.05	−9.88	1.33-18.42	0.023
End cutting burs	31	9.87	4.11	31	17.02	6.55	Ref			38	52.29	17.01	Ref
Ultrasonic	30	12.87	5.58	31	19.68	8.58	2.88	1.58-4.18	<0.001	36	60.50	15.59	7.35	2.66-12.04	0.002
				Difference	95% CI	P
Effect of etching				7.63	6.02-9.24	<0.001

Counts, means, and SDs are for preparation sites and ignore clustering within teeth. Differences, 95% CIs, and P values are from linear mixed models incorporating clustering within teeth and controlling for preparation site position. SDs: Standard deviations, CIs: Confidence intervals

Figure 4

Comparison between enamel and dentin bonding for the three interventions in the etched group (Mpa)

Table 2 shows the relationship between all nonetched groups for the enamel-bonded groups, while Figure 5 compares the dentin bond strength between finishing interventions in both etch and nonetch groups. In the enamel group, there was evidence of overall differences between the three preparations (P < 0.001), in particular that bonding strength was greater following ultrasonic preparation compared to rotary (7.35 MPa; 95% CI 2.66, 12.04; P = 0.002) and that bonding strength following rotary preparation was greater than control (9.88 MPa; 95% confidence level [CL] 1.33, 18.42; P = 0.023), resulting in ultrasonic giving greater bonding strength than control by 17.22 MPa (95% CL 8.62, 25.82; P < 0.001) [Figure 5 and Table 2].

Figure 5

Comparison of the dentin bond strength between finishing interventions in both etch and no etch groups (MPa)

DISCUSSION

Porcelain was bonded to bovine tooth specimens using various preparation treatments and then tensile tested to failure to determine bond strength. A split tooth model was employed to minimize the variability that may be encountered using different teeth and to make more efficient use of resources. Despite randomization of samples to preparation methods, and, for dentin, the use of acid etching, factors such as the age of the animals and their diet may contribute unmodeled variability. While the use of bovine teeth in bonding tests is widely accepted, the data may not be directly comparable to human teeth, as their chemistry and structure are not identical.[1011] The validity of bond strength tests to predict clinical performances of alternative surface preparation techniques or dental adhesives is questionable; recent evidence shows that clinical results can, to a certain extent, be estimated based upon the laboratory results.[12] It is also important to note that the results of this study may not be widely generalized because only a single ultrasonic method and a single rotary method were compared.

The results show that, as hypothesized, ultrasonic instruments provided higher tensile strength values for both enamel and dentin preparations compared to rotary instruments. Acid etching increased the bond strength significantly in the dentin groups, also as hypothesized. This finding around ultrasonic instruments agrees with that of Borges et al., who showed that ultrasonic instruments produce higher bond strengths compared to diamond burs.[13] They reported that dentin surfaces prepared with diamond burs exhibited regions where dentinal tubules were occluded by smear layer and with longer smear plugs when viewed under scanning electron microscope. Conversely, the authors reported that a looser and less compact dentin smear layer was produced when ultrasonic instruments were used. Horne et al. also reported the presence of open dentinal tubules on surfaces finished with the same type of ultrasonic tip.[2] However, the smear plugs were not completely removed. Although ultrasonic treatment does not result in total removal of the smear layer, it does promote general dentinal cleanliness and roughness which facilitates bonding.[14]

Dentin bonding remains unpredictable due to its composition, consisting of a high proportion of moisture, low mineral content, and the presence of smear layer. Dentin contains high water content (21% volume) and <50% volume inorganic material.[15] This makes it mandatory to avoid collapse of the exposed collagen network during dentin bonding, which would inhibit effective infiltration of resin monomers into the dentinal matrix, resulting in lower bond strength. Similarly, smear layer forms tags, which extend into the dentinal tubules, inhibiting effective penetration of resin into the tubules. In contemporary dental practice, smear layer is often either removed or modified to facilitate bonding.

This study also showed that acid etching improved the bond strength of the enamel and dentin. Acid etching exposes the collagenous dentinal network and improves tubules permeability. The effect of etching is more significant than the method of preparation (ultrasonic instruments or diamond burs) in producing higher bond strength, with etching increasing the bond strength by a mean of 7.63 MPa and ultrasonic instruments an additional 2.88 MPa compared to diamond burs.[3] The highest bond strength was achieved through the combination of ultrasonic instruments and etching, followed by rotary burs and acid etching, ultrasonic instruments without acid etching, and finally, rotary burs without acid etching. No interaction between the etching and instrumentation was observed suggesting that the benefits were additive. This highlights that etching can improve bond strength levels, but having a reduced amount of smear layer followed by etching produces the highest bond strength as demonstrated by the ultrasonic plus etching results.

Except periodontology, the uptake of ultrasonics in dentistry has been slow, with ultrasonic scaling reported in 1955 and experimental cavity preparation in 1957. The techniques have required the purchase a benchtop unit, handpieces, and appropriate tips. However, as dentists reequip their surgeries, they may opt to have the ultrasonic machine built in, with its handpiece alongside rotary instruments. Such availability is likely to increase ultrasound use in general practice, in techniques, such as minimally invasive tooth preparation, passive ultrasonic irrigation in endodontics, and post and obstruction removal. The technique's limitation is then the purchase of the specialized tips required. While the increase in bond strength found in this study is small, a further key benefit of ultrasonic preparation methods is their reduced potential to accidentally damage adjacent teeth or restorations.

CONCLUSION

Within the limitations of this in vitro study, several conclusions can be drawn:

Ultrasonic preparation produced higher bond strengths compared to rotary instruments in both etched and nonetched groups

Enamel had greater bonding values than dentin

Etching improved the bond strength across all groups.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.