1.1

Important variables to consider when designing and reporting on bond strength tests

One of the pervasive themes of the conference was the need for consistent documentation of testing variables utilized to evaluate bond strength. Although the present dental literature is replete with bonding studies, meaningful comparisons and drawing strong conclusions are difficult due to testing and reporting deficiencies in many of the published articles. The future provides significant challenges to establishing standardized testing procedures producing predictive experiments that further our understanding of adhesion in dentistry. Further descriptive detail is required in all aspects of the testing procedure to advance the understanding of dental adhesion beyond the present state.

Table 1 includes a list of variables that should be included when designing a bond strength test and when reporting results of such a test. The authors feel strongly that reviewers and editors of scientific journals also can use these guidelines to help ensure the quality and consistency of studies in this important area. It should be emphasized that the overall goal here is not to establish or dictate one specific test method for bond strength evaluation. The conference speakers pointed out the benefits and drawbacks to the many individual test modalities, making it obvious that no single test can address all of the information being sought. However, irrespective of the test method chosen, the investigator can provide the appropriate details outlined in Table 1 .

2.1

A review of adhesion science

The review of adhesion science served as foundational information for the conference to build upon throughout the three days . The paper presented a brief summary of adhesive science with a focus on relevant aspects for dentistry, providing definitions for adhesion and cohesion, and summarizing the principles for a good interface formation as well as methods to characterize the adhesive joints.

Adhesion, involving dissimilar atoms or molecules sticking together, and cohesion, involving like materials sticking together, can be categorized by the type of physical, chemical, and/or mechanical bonding processes that contribute to the interfacial strength. Physical bonding, even if always present, is very weak. Chemically there are possibilities for covalent, ionic, metallic, and chelation bonding. The most common adhesion mechanism for dental materials remains mechanical interlocking. Requirements for creating good adhesion include a clean surface, the generation of a microscopically rough surface, good wetting of the substratum by the adhesive/cohesive materials, low viscosity adhesives with adequate flow, and resistance to phase separation and adequate adhesive solidification. To characterize the adhesive joints several surface analysis methods are available, including ESCA (electron spectroscopy for chemical analysis), SIMS (secondary ion mass spectroscopy), FTIR (Fourier transform infrared spectroscopy), Raman spectroscopy, SEM/EDS (scanning electron microscopy/energy dispersive X-ray spectroscopy) and AFM (atomic force microscopy). Advantages and disadvantages for all of these systems are described with particular attention to the potential for specimen damage that may affect the interpretation of the interface. SEM/EDS is the major surface analysis technique used in evaluating dental materials with AFM gaining in popularity.

Micro-mechanical bonding, including nano-interdigitation, was presented as the basis for adhesive dentistry. The influence of the permeability coefficient of adhesives monomers and primers into tooth structure and stringing, the phenomenon where strings of adhesive are stretched across the original interface, bridging the gap, and preventing complete crack formation, are considered important parameters to characterize interphase adhesion. Post-test characterization of failures of the adhesive/cohesive joint has been utilized to determine effectiveness of bonding. The use of comprehensive fractographic analysis, with identification of the fracture origin (crack initiation), the direction and pattern of crack propagation, the energetics of the fracture and the phases included along the fracture plane could provide a valuable tool in future dental materials adhesion research.

2.2

Variables related to materials and preparing for bond strength testing

While much effort is focused on direct bonding of materials to the tooth surface, a large part of dentistry involved bonding indirect restorations to prepared tooth surface. In this case, composite cement adheres to both tooth and restoration through bonding resin . To mimic the clinical situation, several preparation procedures have been applied to the restoration side for evaluating adhesion between the bonding composite and restoration in vitro. Surface preparation for bonded materials (metals, ceramics and composites) is performed by three approaches: physical, chemical and combined. The physical approach, etching and sandblasting, involves creating a rough surface in the bonded materials. Deposition of a silica layer on the bonding surface by sandblasting, sol–gel processing or molecular vapor deposition has been reported to be effective when followed by a chemical process, such as silanization. Many of the physical treatments including silica deposition are usually used in combination with the chemical approach. Application of silanes to a silica-deposited surface can result in covalent bonds between the bonding resin and the restorative surface.

Aging of test pieces, such as dental composites, prior to bonding affects the bond strength. The most common aging is storage in water at a specified temperature. Thermal cycling may represent a more severe aging procedure for specimens than water storage alone. While the effect of thermal cycling is not always obvious, it is suggested to be included in the screening test of new materials and techniques, because some specimens will spontaneously debond under these conditions. To simulate the clinical conditions in the thermal cycling regime, it is important to remember that the temperature change in the mouth is likely to be relatively slow. Mechanical cycling (fatigue) testing can assess the influence of cumulative damage on the bond strength, and cyclic compression is the most common stress in the mouth. In fatigue testing, it is important to describe three parameters, the applied load, the number of cycles and the rate of application of the load. Long-term fatigue tests may better predict clinical performance than monotonic tests.

2.3

Dentin bonding—variables related to the clinical situation and the substrate treatment

Human dentin is a complicated bonding substrate due to its morphological and physical variations . Therefore, laboratory testing of adhesive materials to teeth should attempt to simulate in vivo conditions. Several clinically relevant factors such as dentin wetness, pulpal pressure, remaining dentin thickness, and type of dentin (normal or sclerotic) should be considered when testing adhesive materials in vitro. Differing from the relatively homogenous and highly mineralized enamel, dentin is very heterogeneous, containing less mineral, and more water, and organic matrix, which makes it challenging to achieve durable adhesive bonds. For example, caries-affected dentin contains mineral deposits in the tubules while non-carious cervical lesions may contain hypermineralized dentin and denatured collagen. On the other hand, transparent dentin beneath caries lesions becomes filled with mineral from passive chemical precipitation. Deep dentin contains high tubule density and increased intrinsic wetness, which in turn, negatively affects bond strengths of etch-and-rinse dentin adhesive systems. Simulation of pulpal pressure is also important, since it affects permeability of dentin and consequently resin bond strengths. In order to reduce permeability and its negative effects on dentin bonding, the use of dentin desensitizers such as 3% potassium oxalate or Gluma desensitizer have been suggested post-etching and prior to application of the adhesive.

Hybrid layer degradation and possible means to avoid such degradation have also been extensively studied. Collagen fibrils which are exposed during acid-etching and not enveloped with adhesive resin may be vulnerable to degradation by endogenous matrix metalloproteinases (MMP). However, current studies have reported that the use of chlorhexidine after etching and rinsing, but prior to the application of adhesive resin, reduces the collagenolytic and gelatinolytic activities of the MMPs. This novel approach, although very recent, seems to preserve the integrity of the hybrid layer. In any case, the storage time and conditions of teeth used for in vitro bonding studies is critical as teeth undergo a degenerative process.

2.4

Adhesion to tooth structure: a critical review of “macro” test methods

“Macro” bond strength tests continue to be reported frequently in the literature, primarily due to the ease of specimen preparation associated with these testing methods . Specimen geometry and test mechanics may influence “macro” shear and tensile bond strength. One of the primary difficulties in comparing studies utilizing “macro” tests is the lack of information on specific testing parameters thus making valid comparisons across studies impossible. Specimen design directly impacts test results. There is a trend for increasing bond strength values with smaller bonding areas. The elastic modulus of the resin composite may affect test results, with the use of stiffer composites showing significantly increased bond strength values. Finite element analysis shows that the higher the elastic modulus mismatch between substrates, the higher the stress concentration at the interface, resulting in lower bond strengths. This was found to be more pronounced in shear than tensile loading, and may be dependent upon the adhesive system.

Bond strength values are impacted by variables related to test mechanics including the type of loading, crosshead speed and incidence of cohesive failures. The test assembly configuration and the distance between the bonded interface and point of load application significantly affect load distribution in shear tests. The use of the chisel as a loading device causes the most severe stress concentration. Another variable related to test mechanics is the influence of specimen loading rate as determined by the crosshead speed. Among the studies reviewed, only one reported differences in the bond strength within the crosshead speed range proposed in the ISO/TS 11405. Last, when there is the occurrence of cohesive failures in either substrate, the validity of the reported bond strength is questionable. Classification of failure modes is difficult, especially within the mixed classification. A cohesive failure can be attributed to the mechanics of the test and brittleness of the material, not necessarily the bond strength.

2.5

Adhesion to tooth structure: a critical review of “micro” bond strength test methods

Bond strength tests are useful for screening new products and studying experimental variables . Currently, numerous tests have been performed in a microshear or microtensile manner, though a consensus standardized approach for the test has not been established. Before bond strength tests can be standardized, it is necessary to understand how measured nominal bond strength is related to stress distribution during testing and ultimately to the clinical performance of the material.

The microshear test generates less stress or damage during the preparation of the specimen prior to testing than the microtensile test; therefore, the microshear test is useful for relatively brittle substrates such as glass ionomers and enamel. Finite element analysis of the microshear mode has found that: tensile stress is responsible for crack initiation, a nonuniform stress distribution is created, and measured nominal bond strength underestimates the true stress. Microtensile bond strength results vary among different laboratories since varying test methods and parameters have been employed. These variables include the type of gripping device (a non-gluing passive device is preferable to minimize unexpected stress concentration), specimen geometry (an hourglass specimen fails at lower stress than stick or dumbbell specimens due to larger stress concentration), specimen preparation (there is a relatively high incidence of pre-testing failures associated with sectioning and trimming) and test speed (1 mm/min crosshead speed was reportedly recommended because of a more uniform stress-time pattern). Until the time that the relationship between bond strength and clinical performance is fully understood, more attainable goals may include: adoption of universally accepted terminology and definitions, standardized reporting of specimen handling and fabrication, inclusion of positive and negative controls during testing, standardized reporting of experimental setup and test mechanics, and full reporting of, or access to a complete data set. Table 1 included in this summary provides guidelines derived from this presentation.

2.6

Review of the fracture toughness approach

The paper presented the theoretical aspects of fracture mechanics and how it can be applied to dental adhesives . Fracture toughness studies performed on dental adhesives were reviewed, focusing on the complexity of conducting fracture mechanics analysis, specimen geometry, load application (i.e. speed) and other testing parameters. Fracture mechanics is based on the fact that fractures are caused by localized stress levels that are high enough to cause crack growth within a material or interface. Therefore, it is useful to compare different materials by their inherent material properties, such as G _c (fracture energy or critical strain energy release rate being defined as the energy required for extending a crack over a unit area) and K _c (critical stress intensity value for crack growth in the material). Parameters, like adhesive thickness and the brittle/plastic character of the adhesive, will however, influence these values. Both of these parameters are often calculated under the assumption that the materials perform truly elastic, but the materials involved in adhesive joints in dentistry are of visco-elastic nature. Consequently, the testing temperature and load rate will affect plastic deformation at the crack tip, which in turn will impact the G _c or K _c values. For this reason, it is recommended to consider plastic–elastic fracture mechanics approaches instead of linear elastic methods.

Performing fracture toughness tests on dental adhesives is complex and errors may occur in the evaluation of fracture toughness values by neglecting the complex stress pattern at the adhesive interface, omitting pertinent parameters like residual stress introduced during polymerization or the generally too small sample size. Furthermore the studies performed by different research groups have used different sample geometries and different loading rates making a direct comparison of the values difficult. Fracture toughness values for adhesives have been shown to vary between 1 and 2 MPa m ^1/2 with a high standard deviations (>25% coefficient of variation). The methods to determine fracture toughness, like the single edge notch tests and the short rod Chevron notch tests on cylindrical, rectangular or prismatic specimens, are further critically compared. Due to the simplicity in the sample preparation in comparison to most other approaches, the Chevron notch test on prismatic specimens may be the most appropriate test. Despite all these limitations, fracture mechanics may be the most proper way for analyzing dental adhesives, and it is most appropriate to report the fracture energy of adhesives rather than fracture toughness values.

2.7

Direct comparison of the bond strength results of the different test methods: a critical literature review

The literature review was based on a search of all dentin bond strength studies reported for 6 specific adhesives with four tests (shear, tensile, microshear, microtensile) . The results from 147 references were utilized from a PubMed search for years 1998 through 2009. Specific selection criteria were established prior to the search. Results were analyzed for variability in bond strength between tests for the same adhesive, coefficient of variation of bond strengths within tests for a same adhesive, mode of failure, and ranking of products by tests. Bond strength values differed significantly for the same adhesive measured in different laboratories with the same test as well as when using different tests. The overall results comparing macro testing with micro testing generally showed that the smaller the bonding surface, the higher the bond strength values. Caution is stressed in making any comparisons based upon results due to the variations in specimen geometries, loading configurations and different modulus of elasticity of the restorative resins. The rankings for the individual products changes depending on which test is used. The coefficient of variation of bond strengths, including those from the microtensile bond test, is high. Almost every testing variable has an influence on the stress state and thus the bond strength values measured. The classification of the mode of failure for bond strength tests remains problematic as there is no uniformly accepted classification system.

Specific recommendations are put forth for consideration. First, all broken specimens that show cohesive failures in dentin or resin composite should be discarded. Only adhesive failures or mixed failures with small areas (less than 10%) should be considered. The classification of the failure requires a microscopic evaluation. Second, the use of Weibull statistics should be systematically applied. Studies should utilize a minimum of 30 non-cohesive failed specimens. Finally, the move to a fracture mechanics approach looking at fracture toughness or the strain energy release rate is encouraged.

2.8

Relationship between bond strength tests and other in vitro phenomena

Bonding performance of adhesive resin materials has been evaluated by macro, micro tensile, and shear bond tests to enamel and dentin . Not only the bond strength tests but microleakage tests, nanoleakage tests, observation of marginal gap and morphological analyses of the bonding interface have been applied to the evaluation of the adhesives. The bond strength is reportedly not correlated with the results of microleakage tests or gap formation tests since the procedures for fabrication of the test specimens are different. Many factors, including the configuration of the cavity and the filling, the mechanical properties of the materials evaluated, the bonding procedures, and the polymerization shrinkage of the filling materials, are considered to significantly affect the results, as previously stated.

Dentin hybridization with adhesives is very important for bonding. Poor impregnation of adhesives into the demineralized dentin allows for space for nanoleakage within the hybrid layer. Water penetrates into the space through nanoleakage channels and induces hydrolytic degradation of the layer. The degradation affects the durability of dentin bonding. The author states that through evaluation of the extent of nanoleakage channels in the hybrid layer using silver penetration one might be able to predict bond durability. The formation of an acid–base resistant zone in dentin recently has been reported. This zone is located adjacent to the hybrid layer in self-etch adhesive systems and may also influence the bond durability as it is more chemically and mechanically stable than normal dentin. Mechanical properties of all of the components at the bonding interface also are considered to be significant factors contributing to the bond strength. Included among these properties is the water absorption and solubility of the adhesives.

2.9

Relationship between bond strength tests and clinical outcomes

Extensive laboratory research is published yearly on bonding performance of adhesive materials, however, the clinical relevance of this data is still questionable . This paper reported on a literature review of laboratory bond strength test methods and resulting data. The second component was a literature review of clinical effectiveness of different adhesives by their retention rates in adhesive class V restorations. The two literature reviews were compared, looking for a potential correlation between laboratory bond strength data and clinical outcomes.

Current adhesive approaches can be divided into an etch-and-rinse, a self-etch, or a self-adhesive approach. Etch-and-rinse adhesive systems are preferred for enamel bonding due to the micro-mechanical retention post-etching and bonding. However, when bonding to dentin, a mild self-etch approach is superior, as it also involves ionic bonding with residual hydroxyapatite, which contributes to bond durability. When bonding to both enamel and dentin, selective etching of enamel followed by the application of the 2-step self-etch adhesive to both enamel and dentin currently appears to provide the most effective and durable bond to tooth tissue. Different laboratory macro and micro test methodologies that measure bond strength and sealing ability of adhesive systems have been reviewed and related to retention rates of adhesive Class-V restorations. There appears to be a correlation between laboratory bond strength and clinical retention rates of Class-V restorations. This is particularly true with ‘aged’ bond-strength data with medium-term retention rates. Therefore, in order to predict clinical effectiveness, laboratory testing of ‘aged’ bond strength should be encouraged in addition to the usual immediate test results.