A STUDY INTO THE LABORATORY TECHNIQUES FOR INTERFACIAL STRENGTH TESTING OF DENTAL

MATERIALS

Cecilia Goracci

UNIVERSITY OF SIENA SCHOOL OF DENTAL MEDICINE PhD PROGRAM:

“DENTAL MATERIALS AND THEIR CLINICAL APPLICATIONS” PhD THESIS OF: Cecilia Goracci TITLE A study into laboratory techniques for interfacial strength testing of dental materials

ACADEMIC YEAR 2003/2004 18 December 2004 Siena, Italy Committee: Promoter Prof. Marco Ferrari Co-Promoter Prof. Franklin R Tay Prof. Piero Balleri Prof. Egidio Bertelli Prof. Carel L Davidson Prof. Michel Goldberg Prof. Manuel Toledano

TITLE A study into laboratory techniques for interfacial strength testing of dental materials CANDIDATE Cecilia Goracci

December 2004

I

CONTENTS Introduction………………………………………………………………………..1

Chapter I Microtensile bond strength to enamel and coronal dentin I.1 Wondering about technique precision and accuracy

I.1.1 Microtensile bond strength tests: SEM evaluation of samples

integrity before testing……………………………………………………………20

I.1.2 Influence of substrate, shape, and thickness on microtensile

specimens’ structural integrity and their measured bond strengths………...38

I.2 Applying the microtensile bond strength test to measure the adhesion on

enamel and crown dentin (stick-forming technique, 1x1mm thick specimens)

I.2.1 Adhesion testing with the microtensile method: effects of dental

substrate and adhesive system on bond strength measurements………….61

I.2.2 Microtensile bond strength to ground enamel and dentin of

simplified adhesives.……………………………………………………………..72

Chapter II Applying the microtensile test to measure bond strength to

radicular dentin II.1 The adhesion between fiber posts and root canal walls: comparison

between microtensile and push-out bond strength measurements…………85

Chapter III Measuring the microtensile bond strength of materials to

non-dental substrates III.1 The adhesion between prefabricated FRC posts and composite resin

cores: microtensile bond strength with and without post silanization……..107

Chapter IV Exploring the application of the push-out test as an

alternative to microtensile. IV.1 Evaluation of the adhesion of fiber posts to intraradicular dentin…….123

IV.2 The contribution of friction to the interfacial strength of endodontic posts

as measured with the thin-slice push-out test………………………………..143

II

Summary, general discussion, conclusions, and future directions….155

Sommario, discussione complessiva, conclusioni e direzioni future.162

Resumé, discussion generale, conclusions et directions futures……170

Resumen, discusion, conclusiones y direcciones futuras……………..178

Acknowledgements…………………………………………………………...189

References……………………………………………………………………...190

Curriculum vitae……………………………………………………………….206

1

INTRODUCTION Adhesive procedures have become routine in the daily practice of dentistry.

The interest in adhesive materials is constantly kept alive by dental

manufacturers, who continuously fight “the battle of the bottles” with the

launch of new products or strategies for bonding.

The need to more thoroughly address clinically relevant aspects of adhesion,

such as bond durability, and the urge to explore still largely unknown fields,

such as bonding to root dentin or to the highly polymerized epoxy resin of

fiber-reinforced composite (FRC) posts, contribute to maintaining a high

level of attention on adhesive materials.

Furthermore, with the introduction of the evidence-based approach, logical

and statistical tools have been provided for a critical appraisal of the in vitro

and in vivo research.

It is known that clinical trials produce the most valuable information.

However, feasibility of in vivo tests is an issue in the dental materials field.

Beside economical and ethical aspects, which are becoming more

demanding, the greatest limitation is represented by the time required to

complete a clinical trial of any scientific value. This is often incompatible with

the frenetic activity and intense productivity of the dental materials market.

Thus, in vitro laboratory tests remain useful in their potential to produce first-

hand information on a newly launched product.

Adhesion can be investigated in vitro through microscopic observations,

microleakage tests, and bond strength measurements. With respect to bond

strength evaluation, the recent years have seen the introduction of a new

method for bond strength testing, the microtensile technique.

Tensile testing at the millimeter and micrometer scale finds several

application in materials science.1-10 In mechanical and electrical engineering

the microtensile technique is applied to assess the tensile properties of weld

joints2 and thin films from metals3,4, polymers5,6, and polycrystalline silicon.7

Polysilicon, in particular, is widely used in the microelectronic industry to

produce tiny sensors and actuators known as microelectromechanical

systems (MEMS).1,7 The microtensile test has also been applied to measure

2

strength and elongation of cotton and wood based cellulose fibers. These

individual cellulose fibers are few millimeters long and tens of microns wide,

and are used as industrial raw materials.8

Additionally, microtensile tests are being performed in orthopedics and

biomedical engineering to measure the stiffness of single trabeculae from

human bone9, as well as to assess the tensile properties of the bone at the

interface with an implant.10

In dentistry the microtensile testing has been extensively applied to measure

the tooth-material interfacial strength, as well as the ultimate tensile strength

of dental materials or substrates. A MEDLINE search with “microtensile” as

the keyword shows that in ten years since the first article on microtensile

appeared in the dental literature (Sano et al., 1994)11, 161 papers have been

published dealing with this technique.

Development of the microtensile technique

Microtensile bond testing was first introduced in order to overcome some of

the limitations of conventional tensile and shear testing, that had been

highlighted by Van Noort and coworkers with finite element analysis.12,13

These authors had pointed out that in conventional tensile testing, when the

load is applied perpendicular to the bonded surface (Fig. 1a), non-uniform

compressive stresses may be introduced along the bonded interfaces if

alignment is not maintained between the adherend and the substrate during

bonding or testing (Fig. 1b).

As far as shear testing is concerned, when the load is applied parallel to the

bonding surface (Fig. 1c), bending moments may develop if the force is

distributed over a relatively extended area of the adherend. The greater the

distance between the point of load application on the adherend and the

substrate, the greater is the bending moment (Fig. 1c).12 If, on the other

hand, the load is applied on a relatively small area at the bonded interface,

there is the risk that the test becomes more of a cleavage rather than a true

shear test (Fig. 1d).

3

Fig. 1 Set-up (a) and force vectors developed in conventional tensile test. (b) Misalignment may lead to non-uniform stress distribution. (c) Shear bond test arrangement. (d) Bending moments and cleavage effect in shear bond strength testing. From Van Noort R et al.13, modified. (a) (b)

(c)

(d)

4

However, the most relevant limitation of conventional tensile and shear tests,

which emerged with the introduction of adhesive systems able to achieve

dentin bond strengths higher than 20 MPa, was the frequent occurrence of

cohesive fracture within the dentin substrate. This is regarded as an

undesirable event as it precludes the measurement of the interfacial bond

strength, and therefore the detection of further improvements in materials

composition and handling. Cohesive failures within the substrate most likely

result from a non-uniform stress distribution.

Microtensile tests, on the other hand, predominantly produce adhesive

failures by testing bonded surface areas of less than 2mm2, where a more

uniform stress distribution is expected to occur.15

Description of the technique, its indications, advantages, and limitations

In a microtensile test, the bonding procedure is performed over the entire

flattened occlusal or buccal surface of the tooth. Then, a composite “crown”

of about 5mm in height is incrementally built over. After complete

polymerization of the resin composite, the tooth is sectioned vertically into a

series of slabs of millimeter thickness by means of a water-cooled diamond

saw.

At this stage, if the non-trimming version of microtensile is followed, each

slab is further sectioned into multiple beam-shaped specimens.

Conversely, if the trimming version is preferred, each slab is trimmed to an

hourglass or dumbbell-shaped profile by means of a water-cooled diamond

bur. The aim is to place the smallest cross-sectional area at the bonded

interface, so as to concentrate the load at this level. Trimming can be done

free-hand, with a bur mounted on a high-speed handpiece, or with the help

of the MicroSpecimen Former developed at the University of Iowa (Fig. 2). In

this device the bur works on the slab like a lathe, thus producing a very

controlled trimming.

A specific indication for the trimming technique is to measure the ultimate

tensile strength of restorative and luting materials or of dental hard tissues.

5

Fig. 2 View of the MicroSpecimen Former device developed at the University of Iowa. From the

MicroSpecimen Former Manual by Armstrong S, Vargas M, Diicks H, Macken D.

As a matter of fact the measurement of tensile properties of mineralized and

demineralized dentin was one of the first uses of the microtensile technique

by Sano et al.16

After the microtensile sticks or hourglasses have been obtained, they are

measured in cross-section, and then loaded in tension until failure by means

of a jig capable of transmitting purely tensile forces, without any torsional

component. Examples of these devices are shown in Fig. 3. Fig. 3 (a) A caliper can be used to apply a purely tensile force to the specimen. (b) The same is accomplished with the microtensile jig designed by Geraldeli. (a) (b)

When testing the bond strength of adhesive materials, following bond failure,

it is advisable to observe the site of failure with an optical or scanning

6

electron microscope, and define the failure mode as either adhesive at the

interface or cohesive within the substrate or the material’s build-up. The

incidence of cohesive failure within the bonding substrate or resin composite

should be minimal in microtensile testing, as adhesive failures can be readily

observed even with bond strengths in excess of 70 MPa.11 The consistent

occurrence of failures at the adhesive interface is desirable, as it allows to

assess the true interfacial strength developed between the bonding material

and the substrate of interest. The ability to more closely reflect the actual

interfacial bond strength is one of the main advantages of microtensile

testing. This property has to do with the more uniform stress distribution

occurring over small-sized specimens, in relation to the presence of less

numerous or less extended defects. Flaws are inevitably present within the

substrates or at the bonded interface, in the form of air bubbles, phase

separations, surface roughnesses or non-uniform film-thicknesses.15 These

defects may represent crack propagation points, from which microfractures

can progress under load to produce the global failure of the specimen. The

lower density of faults is the explanation for the increase in tensile strength

measured for a brittle material with decreasing cross-sectional areas. This

principle was originally formulated by Griffith (1920), who, based on these

observations, promoted the use of high strength glass fibers in materials

science.12,15,7 Later Sano et al. envisaged the possible implications of the

Griffith’s defect theory in testing enamel, dentin, and composite resins, which

are also brittle materials.11

Between Griffith’s intuition and Sano’s application of the principle to dental

substrates and materials testing, Van Noort in 1989 pointed out that the

failure of brittle materials is driven by the density of defects that act as stress

concentration areas. This statement was based on the findings of finite

element analyses.13 Later, a finite element study on bond strength to dentin

conducted by Versluis et al. confirmed Van Noort’s theory of specimen

failure by crack propagation.18

7

The more uniform stress distribution occurring in microtensile specimens has

also a reflection in the lower variance of the data collected with this method,

as compared with conventional testing.14,15

Another immediately evident advantage of microtensile testing is that

multiple specimens can be obtained from a single tooth, thus reducing the

number of teeth necessary to gather a sample of suitable size for statistical

analysis.14 While conventional shear or tensile bond strength tests usually

require sample sizes of 8 to 12 teeth per experimental group, in most

microtensile tests 2 to 4 teeth per group are considered sufficient for the

study’s purpose.19

Microtensile is a versatile technique. Flatness of the substrate is not a

necessary condition for testing, and bond strength can be measured also for

small, irregular surfaces.15 Moreover, local variations in the conditions of

adhesion over a substrate can be assessed. Differences in bond strength

have been detected between occlusal versus middle versus cervical third of

enamel20, between coronal and radicular dentin21, between normal and

adjacent carious dentin22, between occlusal and gingival walls of Class V

cavities23, as well as among the walls of Class II cavities24. It has also been

possible to quantify the effect of enamel prisms and dentin tubules

orientation on bond strength25, as well as the influence of the C-factor and

the layering technique on bond strength to dentin.26

A trade-off for all the mentioned advantages can however be found in the

sensitivity and labor-intensity of the microtensile technique.

A recognized limitation of the method is represented by the test of materials

or regions that produce strengths lower than 5 MPa, as under theses

circumstances many specimens fail prematurely in the cutting, trimming, or

gluing phase.15 Generally, great care must be taken to avoid heat production

during cutting, as well as dehydration of the cut specimens.15

Finite Element Analysis of microtensile specimens

Several studies have been carried out using the Finite Element Analysis

(FEA) to perform a stress analysis in microtensile specimens.

8

The first one, conducted by Phrukkanon et al. in 1998, investigated the effect

of the cylindrical versus rectangular cross-sectional shape and the bonding

surface area on bond strength and stress distribution of specimens trimmed

to an hourglass profile.17 Cylindrical specimens were able to distribute the

stress evenly along the periphery of the bonded surface, whereas in

rectangular specimens stresses concentrated at the corners and the central

areas of the sides. However, there were no statistically significant

differences in bond strength between the two cross-sectional shapes. More

importantly, small surface areas exhibited a more uniform stress distribution

and higher bond strength than larger ones. This finding was related to the

Griffith’s principle regarding the effect of size on the strength of solids.

Adhesive failures at the dentin-resin interface occurred more consistently in

1.1 or 1.5 mm2 specimens.

In a more recent study using FEA, El Zohairy et al.27 questioned that the

higher strength values measured for specimens with smaller cross-sectional

areas are due to the lower occurrence of internal defects and surface flaws.

Based on a finite element model of composite bars of various widths and

thicknesses, the authors concluded that the inverse relationship between

size and strength is mainly due to the attachment of the specimen to the

testing device by the sides. This way of attachment makes the strength

dependent on the thickness of the specimens, whereas holding the

specimen by the top and bottom leads to a more homogeneous stress

distribution. While top and bottom attachment may be feasible for dumbbell-

shaped specimens, it is not easily accomplished with sticks. Therefore, with

beam-shaped specimens the suggestion is to still attach them at their lateral

sides, but to prepare them in the smallest possible thickness, in order to

bring the free surface closer to the path of load application, which

contributes to levelling the stress.

Meira et al.28 pointed out that the mode of specimen fixation has an effect on

the stress concentration in square hourglass microtensile specimens. These

authors emphasized that in scientific papers dealing with the microtensile

method, almost never are parameters such as number of fixation sides,

9

height of the fixed region, and curvature radius of the notch reported. These

geometrical parameters have indeed an influence on the stress

concentration factor kt, which is defined as the ratio between the maximum

and the mean (the nominal) tensile stress. The higher the stress

concentration factor, the higher the probability of specimen failure under

relatively low nominal stresses. Meira et al.28 modelled hourglasses from an

isotropic and homogeneous material with elastic modulus of 9.8 GPa and

Poisson’s ratio of 0.25, supposedly simulating composite resin. In the

hourglasses the neck width and thickness were kept constant to 1mm,

whereas curvature radius of the notch, number of fixation sides, height of the

fixed region varied. Based on the Kt calculated for each geometrical

configuration, the authors suggested fixing the specimen by two sides, rather

than one, as in the former configuration the stress concentration factor is

reduced, and the variations in kt induced by the other geometrical

parameters are limited. It should however be mentioned that fixation by one

side is in actuality never done in microtensile testing.

Additionally, the authors stated that as the stress concentration factor is

increased when the height available for fixation is reduced, the geometrical

configuration of the test may be partly responsible for the lower bond

strength values usually measured in deep dentin. In order to control for this

variation, when the remaining dentin thickness is low, Meira et al. suggested

bonding some composite to deep dentin on the opposite side of the tested

interface so as to increase the specimen height.28

In conclusion, the authors recommended that the geometrical parameters

involved in microtensile testing be as much as possible standardized for the

purpose of reducing variations among studies and for the sake of results

comparability.

Silva et al.§ used the finite element analysis to evaluate whether the

angulation of the adhesive joint has an effect on the stress distribution

§Silva NRFA, Calamia CS, Harsono M, Carvalho RM, Pegoraro LF, Fernandes CAO, Vieira ACB, Thompson VP. Bond angle effects on microtensile bonds: laboratory and FEA comparison. PhD thesis of Dr. Nelson RFA Silva, Bauru School of Dentistry, 2004.

10

represented by the Maximum Principal Stress (MPS). According to the

authors, the study of angled interfaces is relevant as caries-affected dentin

and the dentin of prepared cavities is generally more curved or irregular than

the flat dentin surface usually prepared for microtensile testing. The finding

of the FEA were also validated by the results of microtensile testing of

specimens from human dentin, treated with different dentin bonding agents

and exhibiting differently angled interfaces. FEA and laboratory tests

demonstrated a trend toward decreasing MPS and bond strength values as

the bond angle increased. The authors mentioned two aspects that may

explain this finding. First, in the more angled interfaces the cross-sectional

area is greater and therefore more likely to incorporate flaws. Secondly, the

load applied on an angled joint produces bending moments that reduce the

bond strength values.

Silva et al.§ used FEA also to evaluate the effect of adhesive layer thickness

on MPS in each angled group and reported a tendency for MPS to increase

as the adhesive thickness increases. This finding is explained by the

elasticity of the adhesive layer, in relation with the Poisson’s contraction from

the edges of the adhesive toward the center of the specimen. This

contraction relieves the tensile stress and allows for a greater tolerance of

the load prior to the occurrence of a fracture. The greater the adhesive

thickness, the greater the tensile stress relief.

In the geometrical model proposed in this study constraints were defined

over four surfaces on each component (resin and dentin), so as to closely

simulate the most usual way of gluing the specimen to the loading device.

Additionally, the authors warned against the risk of inaccuracy and

oversimplification lying in FEA analyses, in that dentin was modelled as an

isotropic material. Enamel and dentin are indeed anisotropic substrates,

owing to the presence of prisms and tubules respectively. Given the

complexity and the degree of approximation involved in simulating the

mechanical behaviour of a combination of materials, Silva et al.§

recommended that the results of FEA models should always be validated by

laboratory data.

11

Statistical interpretation of microtensile bond strength data

Some statistical issues concerning the treatment of microtensile bond

strength data still remain unsolved. The most relevant issue is whether or

not premature failures should be included in statistical calculations as “zero

value” bonds.14,24 The question is not idle, as the inclusion or exclusion of

“zero bonds” can remarkably affect the measures of central tendency and

spread of the data set (mean and standard deviation respectively), as shown

by the microtensile bond strength data for a new self-adhesive resin cement

on enamel and dentin under different conditions of substrate hydration (Fig.

4).29

Fig. 4 The histogram represents mean (dark grey) and standard deviation (light grey) values of

the bond strengths calculated excluding or including premature failures for a self-adhesive resin

cement on enamel (E), dry dentin (DD), wet dentin (WD), and dehydrated dentin (DeD). The

inclusion of premature failures lowers the mean and increases the standard deviation. The

change is particularly evident in the “dehydrated dentin” group, where the number of premature

failures was remarkably high. Data from Goracci et al.15

Group Total # of

beams from four teeth

# of intact beams for

testing

Microtensile bond strength (MPa) without considering premature failures

Microtensile bond strength (MPa)

after considering premature failures

Enamel 50 40 12.2 ± 4.1 9.8 ± 6.1 Dry dentin 61 32 14.4 ± 5.2 7.7 ± 8.2

Wet dentin 69 35 15.0 ± 4.4 7.6 ± 8.2 Dehydrated

Dentin 65 2 5.5 ± 1.0 0.2 ± 1.0

0 2 4 6 8

10 12 14 16 18 20

E w/out zeros E w/

zeros DD w/out zeros

DD w/zeros

WDw/outzeros

WD w/zeros

DeDw/outzeros

DeD w/zeros

sdmean

12

The inclusion of premature failures as zero values may also have an effect

on the data distribution, by changing it from a normal, Gaussian-like

distribution to a non-normal one (Fig. 5). The latter limits the use of

parametric statistics for verifying the significance of differences among the

experimental groups26.

Nevertheless, a high frequency of prematurely debonded specimens

logically suggests a greater fragility of the bond. Under these conditions,

basing the calculations solely on the specimens that survived the

preparation may bias the test toward an overestimation of the adhesive

potential.19,24

A similar question may be raised regarding the specimens that fail

cohesively within the substrate or the material’s build-up. Should cohesive

failures be simply disregarded or should they be accounted for as values

greater than the highest measured interfacial bond strength? Fig. 5 Steam-and-leaf plot and normality curve for a set of data from microtensile measurements of the composite-composite bond strength. With the exclusion of premature failures (zeros), the data set follows a normal distribution.

00000000000000000000000000

0.8

1.11122223

1.5577789999

2.01111122233333334444444

2.5556666777777788888889999999

3.0011111222233344444

3.555555666666666677777778888888889999999

4.00000011111111111222222223333334444444444

4.5555666677777777777888888899999

5.000000000012223334

5.555688

6.0001113444

With the intention to rule out the chance of such misinterpretations in the

microtensile method, Reis et al.19 have proposed to calculate for each tested

tooth a bond strength index that considers the relative contribution of all

types of failures, premature, adhesive, cohesive in dentin and in composite

13

resin. The formula of bond strength index assumes the cohesive strength of

dentin and the cohesive strength of resin composite to be the average value

of all the specimens from one tooth that failed in that manner. Premature

failures are instead attributed an arbitrary value, which is approximately half

of the lowest bond strength that could be measured in the study.30

It has also been proposed to replace each premature failure with a value

estimated on the basis of the regression model (linear, cubic etc.) that best

fits the entire data set.24

Attributing premature failures a greater than zero bond strength makes

sense if one considers that it must have taken a certain amount of stress to

produce the failure during preparation.26

In addition, when testing natural teeth, statistics can either be calculated per

tooth, or, regardless of the tooth of origin, by pooling together the specimens

from all the teeth of a same group. By calculating the statistics per group,

rather than per microtensile specimen, the sample size is reduced. As a

consequence the power of the study is lowered, with the end result of

levelling the differences among groups.

In order to account for the tooth-related variability, some statisticians

suggest to treat the bond strength data from specimens of a same tooth as

“repeated measures”. Therefore, the proposed analysis for the significance

of differences among experimental groups is the ANOVA for repeated

measures, with tooth as within subject factor and the variables under study

as between subject factors.31

Some other researchers prefer to add the tooth of origin as a random factor

to the statistical model, in order to correct for the multiple specimens

gathered from one single tooth.32,33

Another approach involves applying the ANOVA test to check for the

existence of significant differences among the teeth within each

experimental group prior to using each microtensile specimen as an

independent unit.30

14

Beside the statistical interpretation of bond strength data, the precision and

accuracy of microtensile measurements are other aspects that deserve

attention.

It is considered precise a measurement “that has nearly the same value

each time it is measured”.34 The accuracy of a measurement, on the other

hand, is defined as “the degree to which it actually represents what it is

intended to represent”.34

Based on these concepts, the microtensile method provides a precise and

accurate measurement of adhesion between two substrates if it is capable of

consistently detecting the actual strength at the bonded interface.

The test may be biased from the assessment of the true interfacial strength

by the presence of faults within the substrates or at the interface, which are

supposed to act as stress raisers.12 The theory founding the microtensile test

claims the density of these intrinsic defects to be relatively low in relation to

the reduced dimension of the specimens.

The first step in this research project was indeed to verify this assertion by

viewing with a scanning electron microscope the surface of beam-shaped

microtensile specimens from human enamel and dentin.

In the following study, different shapes and thicknesses in which microtensile

specimens can be prepared were compared for measured bond strengths

and microscopic aspects, with the aim of identifying the most appropriate

specimen design for microtensile bond testing on enamel and coronal

dentin. Then, the method of specimen preparation which was found to be

most adequate for adhesion testing on coronal hard tissues was applied to

measure the bond strength of simplified adhesives to enamel and dentin.

Additionally, the increasing use of FRC posts adhesively luted inside root

canals for the restoration of endodontically-treated teeth has raised the

interest around bonding to root canal dentin.

Also, the strength of the adhesion between fiber posts and the resin

composites used for luting or core build-ups has drawn the attention of

researchers and manufacturers.

15

A further objective in this project was therefore to evaluate the potentials of

microtensile to assess the adhesion of luted fiber posts, as well as the

strength of the bond between FRC posts and the resin cements or core

materials available on the market.

For measuring the retentive capability of endodontic posts, the push-out test

can also be performed, and the possible contribution of sliding friction to

interfacial strength has to be taken into consideration and possibly

quantified. Two studies were conducted for this purpose.

16

References 1. Sharpe WN. Tensile testing at the micrometer scale: opportunities in

experimental mechanics. Experimental Mechanics 2003; 43: 228-237.

2. Lee J, Ju J, Jang J, Kim W, Kwon D. Weld crack assessments in API

X65 pipeline: failure assessment diagrams with variations in

representative mechanical properties. Materials Science and

Engineering A 2004; 373: 122-130.

3. Augulis L, Tamulevičius S, Augulos R, Bonneville J, Goudeau P,

Templier C. Electronic speckle pattern interferometry for mechanical

testing of thin films. Optics and lasers in engineering 2004; 42: 1-8.

4. Son D, Kim J, Lim TW, Kwon D. Evaluation of fatigue strength of LIGA

nickel film by microtensile tests. Scripta Materialia 2004; 50: 1265-1269.

5. Sanders JE, Nicholson BS, Mitchell SB, Ledger RE. Polymer microfiber

mechanical properties: a system for assessment and investigation of the

link with fibrous capsule formation. J Biomed Mater Res Part A 2003; 67:

1412-1416.

6. Shi XQ, Wang ZP, Pang HLJ, Zhang XR. Investigation of effect of

temperature and strain rate on mechanical properties of underfill

material by use of microtensile specimens. Polymer Testing 2002; 21:

725-733.

7. Son D, Kim J, Lim TW, Kwon D. Evaluation of fracture properties of

silicon by combining resonance frequency and microtensile methods.

Thin Solid Films 2004, in press.

8. Kompella MK, Lambros J. Micromechanical characterization of cellulose

fibers. Polymer Testing 2002; 21: 523-530.

9. Bini F, Marinozzi A, Marinozzi F, Patanè F. Microtensile measurements

of single trabeculae stiffness in human femur. J Biomech 2002; 35:

1515-1519.

10. Kim DG, Brunski JB, Nicolella DP. Microstructural-level strain-fields

during a microtensile test of bone at a bone implant interface.

17

Proceedings of the American Society of Mechanical engineers (ASME)

2001, Snowbird, Utah.

11. Sano H, Shono T, Sonoda H, Takatsu T, Ciucchi B, Horner JA, Pashley

DH. Relationship between surface area for adhesion and tensile bond

strength – Evaluation of a microtensile bond test. Dent Mater 1994; 10:

236-240.

12. Sudsangiam S, Van Noort R. Do dentin bond strength tests serve a

useful purpose? J Adhes Dent 1999; 1: 57-67.

13. Van Noort R, Noroozi S, Howard IC, Cardew G. A critique of bond

strength measurements. J Dent 1989; 17: 61-67.

14. Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshijama M, Shono

Y, Fernandes CA, Tay F: The Microtensile Bond Strength: A Review. J

Adhes Dent 1999; 1: 299-309.

15. Pashley DH, Sano H, Ciucchi B, Yoshiiama M, Carvalho R. Adhesion

testing of dentin bonding agents: A review. Dent Mater 1995; 11: 117-

125.

16. Sano H, Ciucchi B, Mathhews WG, Pashley DH. Tensile properties of

mineralized and demineralized human and bovine dentin. J Dent Res

1994; 73: 1205-1211.

17. Phrukkanon S, Burrow MF, Tyas MJ. The influence of cross-sectional

shape and surface area on the microtensile bond test. Dent Mater 1998;

14: 212-221.

18. Versluis A, Tantbirojn D, Douglas WH. Why do shear bond tests pull out

dentin? J Dent Res 1997; 76: 1298-1307.

19. Reis A, Loguercio AD, Azevedo CLN, Carvalho RM, Singer JM, Grande

RHM. Moisture spectrum of demineralized dentin for adhesive systems

with different solvent bases. J Adhes Dent 2003; 5: 183-192.

20. Shono Y, Terashita M, Pashley EL, Brewer PD, Pashley DH. Effects of

surface area on resin-enamel tensile bond strength. Dent Mater 1997;

13: 290-296.

18

21. Yoshiyama M, Carvalho RM, Sano H, Horner JA, Brewer PD, Pashley

DH. Regional bond strengths of resin to human root dentin. J Dent 1996;

24: 435-442.

22. Nakayima M, Sano H, Burrow MF, Tagami J, Yoshiyama M, Ebisu T,

Pashley DH. Tensile bond strength and SEM evaluation of caries-

affected dentin using dentin adhesives. J Dent Res 1995; 74: 1679-

1688.

23. Yoshiyama M, Sano H, Ebisu S, Tagami J, Ciucchi B, Carvalho RM,

Johnson MH, Pashley DH. Regional strengths of bonding agents to

cervical sclerotic dentin. J Dent Res 1996; 75: 1404-1413.

24. Bouillaguet S, Ciucchi B, Jacoby T, Wataha JC, Pashley DH. Bonding

characteristics to dentin walls of Class II cavities, in vitro. Dent Mater

2001; 17: 316-321.

25. Carvalho RM, Santiago SL, Fernandes CAO, Suh B, Pashley DH.

Effects of prism orientation on tensile strength of enamel. J Adhes Dent

2000; 2: 251-257.

26. Nikolaenko SA, Lohbauer U, Roggendorf M, Petschelt A, Dasch W,

Frankenberger R. Influence of C-factor and layering technique on

microtensile bond strength to dentin. J Adhes Dent 2004; 20: 579-585.

27. El Zohairy AA, de Gee AJ, de Jager N, van Ruijven LJ, Feilzer AJ. The

influence of specimen attachment and dimension on microtensile

strength. J Dent Res 2004; 83: 420-424.

28. Meira JBC, Ballester RY, Souza RM, Driemeier L. Stress concentration

in microtensile tests using uniform material. J Adhes Dent 2004, in

press.

29. Goracci C, Grandini S, Monticelli F, Tay FR, Ferrari M. Bonding

mechanism of a new self-adhesive resin cement to dental hard tissues. J

Adhes Dent, 2004. In press.

30. Bouillaguet S, Troesch S, Wataha JC, Meyer JM, Pashley DH.

Microtensile bond strength between adhesive cements and root canal

dentin. Dent Mat 2003; 19: 199-205.

19

31. Kurtz JS, Perdigão J, Geraldeli S, Hodges JS, Bowles WR. Bond

strengths of tooth-colored posts. Effect of sealer, dentin adhesive, and

root region. Amer J Dent 2003; 16: 31A-36A.

32. Loguercio AD, Uceda-Gomez N, Oliveira Carrilho MR, Reis A. Influence

of specimen size and regional variation on long-term resin-dentin bond

strength. Dent Mater 2004; in press.

33. De Munck J, Van Meerbeek B, Yoshida Y, Inoue S, Vargas M, Suzuki K,

Lambrechts P, Vanherle G. Four-year water degradation of total-etch

adhesives bonded to dentin. J Dent Res 2003; 82: 136-140.

34. Hulley SB, Cummings SR. Planning the measurements: precision and

accuracy. In Hulley SB, Cummings SR. Designing clinical research.

Baltimore, Williams and Wilkins, 1988.

20

CHAPTER I: MICROTENSILE BOND STRENGTH TO ENAMEL AND

CORONAL DENTIN I.1 WONDERING ABOUT TECHNIQUE PRECISION AND ACCURACY I.1.1 Microtensile bond strength tests: SEM evaluation of samples

integrity before testing. Ferrari M, Goracci C, Sadek FT, Cardoso PEC. The European Journal of

Oral Sciences 2002; 110: 385-391.

Introduction

Shear and tensile bond strength tests have long represented the most

common laboratory trials for evaluating the adhesion of bonding systems to

enamel and dentin. These tests are relatively easy to carry out, are widely

applied in dental research, and they provide the bulk of the currently

published data on bonding systems. However, it has been shown that both

tensile and shear bond strength tests can be greatly affected by the

variability in specimen geometry and experimental loading conditions.1,2 Microtensile tests have been developed3 to overcome some of these

limitations, and are now regarded as the most predictable bond strength

tests than can be performed.4

The microtensile technique offers several advantages over the other

procedures.4-7 One of the main objectives of the method, as it was first

introduced, was to avoid the occurrence of cohesive fractures of dentin on

loading. Cohesive failures in dentin, while the resin-dentin bond remains

intact, have been frequently reported since the introduction of newer

adhesive systems that create bond strengths on dentin of 20 to 25 MPa.4

The occurrence of failure of the substrate itself prevents the measurement of

interfacial bond strength and, hence, the evaluation of improvements in

bonding procedures or formulations. Using the microtensile test results in a

more uniform distribution of loading stresses across a smaller bonded

interface, thereby reducing the frequency of cohesive fractures in dentin, as

compared with conventional bond strength tests.4

21

The structural variability of the substrate in small bonding sites is expected

to be limited, thus allowing for a more accurate analysis of the bonding

mechanism. Indeed, the microtensile technique has found a specific

application in highlighting the differences in bonding characteristics between

small regions of dental tissues,4 such as normal versus adjacent carious

dentin,8 enamel versus dentin, coronal dentin versus root dentin,4,9-11 and

occlusal versus middle versus cervical enamel.12

In addition, since a number of microtensile specimens can be obtained from

a single tooth, collecting suitable numbers of teeth that meet the statistical

criteria becomes easier.4

A trade-off for the simplified sample collection, however, exists in the fact

that microtensile testing is a very “technique-sensitive” procedure. Proper

specimen preparation requires special testing equipment and a skillful

investigator. The method involves cutting a bonded tooth into a number of

slabs, which are then further sectioned into sticks of 0.5-1.5 mm thickness.

Each stick is made up of the two substrates (i.e. resin composite versus

enamel or dentin), which are bonded at the interface to be tested. The stick

can be left in a beam shape in the “non-trimming technique”,5,13-16 or can be

trimmed with burs at the bonding site,5,11,16-17 to create an hourglass profile

that reduces the bonding surface even more, further concentrating the

loading stress.

All the cutting procedures, particularly the bur-trimming of the hourglass

shape, likely transmit vibrations to the specimens. A common occurrence

when preparing microtensile specimens, especially if the bond strengths are

relatively low (5-7 MPa),4 is a premature failure of the specimen that makes

it useless.5,14,16

The number of prematurely failed, discarded specimens in each test is

probably related to the “aggressiveness” of the preparation procedure. In this

regard, the non-trimming technique should prove as the least traumatizing

and most efficient.4,5

SEM observations can be of great help in revealing interfacial voids, gaps or

cracks in microtensile specimens that can be responsible for their premature

22

failure. However, SEM investigations of microtensile specimens in order to

assess the type of failure have only been conducted after tensile testing.18,19

No previous SEM study has evaluated the integrity of microtensile samples

after preparation, but before loading, in order to detect those structural

defects in the substrates or the adhesive interface, that can significantly

affect the bond strength test results.

The purpose of this study was to determine if SEM observations of sticks

prepared from resin-bonded enamel and dentin could identify interfacial

defects prior to microtensile testing.

Material and Methods Twenty-eight sound human molars, recently extracted for periodontal

reasons, were selected for the study. Any residual soft tissue was removed

from the roots with a scaler. The teeth were then rinsed with water, and

stored in a saline solution at 4°C for no longer than 3 months.

On all of the specimens, the roots were cut off at the middle third with a

diamond disc. Each tooth was then randomly assigned to one of two groups:

group A included samples for resin-enamel bond strength testing. On these

specimens, some of the most superficial enamel was cut off from the buccal

or lingual aspect of the tooth with a cooled diamond disk on a Labcut 1010

machine (Extec Corp.,Enfield, CT, USA) (Fig. 1a). Care was taken not to

expose any portion of the underlying dentin, but to create a flat surface of

enamel, which was then polished with wet sand-paper (Fig. 1a). Fig. 1a Enamel specimens preparation involved the removal of a portion of superficial tissue without exposing the underlying dentin.

23

Group B samples were used to evaluate resin-dentin bond strengths. On

these teeth, all of the occlusal enamel and some superficial dentin were

removed with a cooled diamond disk, to obtain a flat surface in mid-coronal

dentin (Fig. 1b). Fig. 1b Tooth prepared for dentin test: the occlusal third was removed with a diamond disc, creating a flat surface.

All polishing of enamel or dentin bonding surfaces was done with wet sand-

papers to create a standard smear layer. Abrasive (SiC) papers of 220, 320,

and 400 grit were used in sequence, each one for 10 seconds, followed by a

final polishing with a 600 grit sand-paper for 60 seconds. Finally, the bonding

surface was rinsed with water, and lightly dried with an air stream.

Within each group, two subgroups were then randomly formed, in which two

different bonding systems were tested. They were a self-etching primer

system (Clearfil SE, Kuraray, Morita, Japan - subgroup 1), and a ‘total-etch

system’ (Excite, Vivadent, Schaan, Liecthenstein - subgroup 2). These

materials were used according to the manufacturer’s instructions (Table 1).

Table 1 Recommended steps for the handling of the tested bonding systems.

Clearfil SE Excite Apply SE Primer and wait 20s Apply H3PO4 for 15 s

Gently air-dry Rinse and dry gently, leaving the surface damp

Apply SE Bond Apply the adhesive for 10s Lightly blow with air Lightly blow with air Light cure for 10s Light cure for 20s

24

After applying the adhesive system, a proprietary composite resin block of

approximately 5 mm x 5 mm x 5 mm was built on the bonding surface,

following the incremental technique. Each layer of composite was

individually cured for 40 seconds, with an Optilux 401 light (Demetron, Kerr

Co, Danbury, CT, USA, 600 mW/cm2 intensity) (Fig. 1c). Fig. 1c Resin build-up over the enamel and the dentin surface.

The bonded specimens were placed in a saline solution at 27° for 24 h.

Using a diamond blade, each bonded tooth was sectioned vertically into a

number of slabs 0.8 mm thick. By rotating the sample 90° and again

sectioning it lengthwise, multiple beam-shaped sticks, each with a cross-

sectional surface area of approximately 0.8x0.8 mm=0.64 mm2, were

obtained. The lower half of the sticks was made up of the dental substrate,

while the upper half was composed of the resin build-up (Fig. 1d). Fig. 1d Cutting of the tooth along the X and Y axis and the resulting stick.

Two to three sticks from each subgroup were randomly selected for

microscopic analysis. A total of 80 sticks, 40 with enamel and 40 with dentin

as bonding substrates, were processed for SEM observations.

25

The preparation involved a gentle surface decalcification of the sticks with

36% phosphoric acid for 10 seconds, and a brief deproteinization of the

surface of the interface between resin and dentin with 2% sodium

hypochlorite solution for 60 seconds. After rinsing with water, the specimens

were dehydrated in ascending acetone concentrations (30, 50, 70, 90 and

100%), and critical-point dried (CPD 030, Balzers, Liecthenstein). Finally,

each stick was mounted on aluminum stubs, sputter-coated with gold by

means of the Edwards Coater S150B device (Edwards Ltd., London, UK),

and observed under a Philips 515 scanning electron microscope. Only two of

the four sides of the sticks could be imaged, as one side was on the stub

and one side was directed away from the scanning beam.

Microphotographs were taken at different standardized magnifications (x120,

x710, x1010). The low magnification provided an overview of each stick (Fig.

2a), whereas the micrographs taken at higher magnifications revealed the

quality of the bonding interface, as well as various structural defects of the

specimens (Fig. 2b). These were classified as microfractures within the

composite resin (CR), within the adhesive resin (AR), the hybrid layer (HL),

dentin (D), or enamel (E). Gaps were sometimes seen running along the

interfaces, such as the interface between composite and adhesive (CR-AR),

between the adhesive and the top of hybrid layer (AR-THL), between the

bottom of the hybrid layer and dentin (BHL-D), or the bottom of the hybrid

layer and enamel (BHL-E) (Table 2). The morphological characteristics of

the hybrid layer and the resin tags penetrating the dentin substrate were also

analyzed (Fig. 2c). Every defect of each sample was counted, evaluating the

location of the defect and including it in the corresponding group.

Results No premature failures of the bonds occurred during the preparation of the

sticks. However, none of the microtensile specimens observed under the

scanning electron microscope appeared free of structural defects. These

faults consisted of microcracks in the dental substrate, either in enamel (Fig.

3), or in dentin at the level of hybrid layer (Fig. 4).

26

Fig. 2 (a) Overview of a stick from the Clearfil SE group under the scanning electron microscope at a low magnification (bar, 0.1mm). (b) Higher magnification of the previous stick to detect the presence of gaps along the adhesive interface or fractures within the dental substrate (bar, 0.1mm). (c) View of the adhesive interface, to reveal the morphological characteristics of the hybrid layer and the resin tags (arrows) penetrating the dental substrate (bar, 10µm).

Fig. 3 Microphotograph of a stick taken to reveal the presence of microcracks in the enamel substrate. An enamel crack running at 90° from the bonded interface is shown (arrows, a). An interfacial gap is also visible (arrows, b) (bar, 10 µm).

a

b

c

a

b

27

Fig. 4 Microphotograph of a dentin stick exhibiting cracks at the top of the hybrid layer (arrows) (D, dentin, HL, hybrid layer, CR, composite resin) (bar 0.1 mm).

Another type of defect that was often seen was the presence of gaps, either

between enamel and resin, or between the hybrid layer and resin; voids

within the resin composite thickness were also sometimes visible (Fig. 5). On

a few dentin specimens, a small portion of residual enamel remained

because of incomplete occlusal preparation (Fig. 6). A very common

observation was the presence of microcracks in the enamel substrate at the

corners of the sticks. Regardless of the material used for bonding, the

finding of microfractures was more frequent in enamel than in dentin

samples. Hybrid layer and resin tag formation was noted in all samples (Figs

2 and 7).

Table 2 summarizes the observed defects and their location for each

subgroup.

HL

D CR

28

Fig. 5 A stick from the Clearfil SE group exhibiting a void (arrow) in the composite layer (CR,

composite resin, D, dentin) (bar, 0.1mm).

Fig. 6 Residual enamel (E and arrows) at the periphery of a stick prepared to test the dentin substrate (D) (CR, composite resin; bar, 0.1 mm).

CR D

CR

D E

29

Fig. 7 Microtensile specimen from the Excite dentin subgroup with a clear adhesive layer (arrows). Integrity was not affected by the preparation procedures (bar 0.1 mm).

Table 2 Frequency and location of structural defects as seen under the SEM

Substrate Adhesive CR AR or HL

D or E CR-AR

AR-THL

BHL-D/E

No linear

Interface

Defect of substrate

Clearfil SE 3 4 4 25 32 1 Dentin Excite 5 5 2 22 34

Clearfil SE 4 14 4 30 20 1 Enamel Excite 3 10 4 15 23 3

CR, defects (voids) within composite resin; AR, defects within adhesive resin; HL, defects within hybrid layer; D, defects within dentin; E, defects within enamel; CR-AR, defects between composite resin and adhesive resin; AR-THL, defects (microcracks) between adhesive resin and top of hybrid layer; BHL-D/E, defects between bottom of hybrid layer and dentin or enamel substrate; no linear interface means that the adhesive interface was not positioned linearly; defects of substrate means that the dental substrate showed defects per se.

Discussion The development of the microtensile method can be regarded as a

significant contribution to the science of adhesion testing.7 One of the limits

of the conventional tensile bond strength technique, revealed through FEA

analysis by Van Noort et al.1 and De Hoffe et al.20, is the highly non-uniform

stress distribution across the bonded surface, on which specimen geometry,

material stiffness, and loading configuration have an effect. In the

30

microtensile technique, the tested surface is so small that the variability

introduced by these factors is greatly reduced. Further, the regional

differences in the structure of dental tissues can be controlled, and the

stress distribution across the bonding surface is thought to be much more

uniform. This allows for a more realistic and reliable appraisal of resin-tooth

bond strengths. In the present study, a scanning electron microscope analysis was

performed in order to detect any structural defects exhibited at resin- enamel

or dentin interfaces in microtensile specimens before loading, as a result of

the bonding or preparation procedures. It was revealed that none of the

specimens was flawless, and that the majority of the flaws were located on

the enamel side or in resin-tooth bonds . In a microtensile bond strength test performed using the same protocol as

the present research (Cardoso et al. 2001, AADR, Chicago), the resin bond

strength measured on enamel were not significantly higher than those

achieved on dentin. As years of research and clinical experience have

clearly demonstrated, bonding to enamel is far more reliable than adhesion

on dentin; it is therefore fair to accept that enamel samples might fail under

relatively lower loading levels, owing to the intrinsic brittleness of this tissue

in the reduced surface areas used in microtensile specimens.21 This calls

into question whether microtensile testing is an appropriate trial for enamel,

which is fragile, anisotropic, and has a water content lower than dentin. The suspicion of an intrinsic weakness of the enamel in microtensile

specimens was confirmed in the SEM analysis of the sticks. As mentioned

earlier, microscopic analysis of the sticks before loading revealed a more

frequent occurrence of microcracks in enamel than in dentin. Microcracks

were also most often located at the periphery of the sticks, suggesting that

these defects were inadvertently introduced by the specimen preparation

procedures, particularly the vibrations of the cutting devices, disks and burs.

Had the enamel cracks developed before resin bonding, they would have

been filled with resin. The fact that the cracks were empty indicates that they

developed after bonding was done.

31

In this research, the “non-trimming” method of specimens preparation was

followed, which was expected to be less traumatic than the methods where

an hourglass profile is created with burs at the bonding interface.5 In the

present study, no premature failure of the resin-enamel or resin-dentin bond

occurred. Nevertheless, all of the bonded interfaces in the sticks exhibited

structural defects at various locations. It might be argued that the use of ascending concentrations of acetone could

have been responsible for the extraction of poorly polymerized materials,

thereby creating voids that did not exist before such treatment. Similar

technique artifacts might have been introduced with the use of ethanol.

However, if these materials were “acetone-extractable”, they were probably

not contributing much to the bond. Conversely, it is also true that some linear

polymers which are soluble in acetone, such as Polymethylmethacrylate and

Poly Hydroxyiethilmethacrylate, can provide good bond strength. In addition,

microscopic images such as the example given in Figs 2 and 7 clearly

demonstrate that when the resin-dentin bond is of good quality, it is able to

withstand also the challenge of exposure to acetone solutions without

developing defects. It should then be pointed out that, in order to be observed under the SEM,

the sticks underwent a vacuum desiccation and that the stress imposed to

the specimens by this procedure may be responsible for some of the

detected defects. If epoxy resin replicas of the specimens had been made

for microscopic evaluation, bubbles and other artifacts might have been

introduced. This technique also tends to lower the resolution of microscopic

images to a certain extent. Detailed, high-resolution imaging up to x5000 can

be performed with the epoxy resin replica technique using a proper

impression material, adequate degassing and a high-quality microscope;

however, handling the size of microtensile samples can be very difficult.

Mannocci et al.,22 in a recent analysis of dentin microtensile specimens

using a confocal microscope at normal atmosphere pressure, frequently

observed fractures or cracks of the dental substrate. In order to limit the

32

occurrence of this phenomenon, the authors suggested preparing

specimens no thinner than 1.5 mm.

The thickness of the specimen seems to be critical in determining its ability

to survive the preparation procedures for microtensile testing. Bouillaguet et

al.5 reported a high incidence of premature failures (26%) during hourglass

trimming of microtensile dentin specimens 0.5 mm thick. They suggested

that, especially when using the trimming technique, the slabs be made

thicker prior to trimming. Phrukkanon et al.23 recommended never reducing

the cross-sectional area at the bonding interface to less than 1.1 mm2, since,

in a pilot study, specimen failures greatly increased below this size. The

authors believed that a 1.5 mm2 cross-sectional surface is the most

appropriate, at least for the trimming technique, as they reported a minimal

percentage of premature failures when handling specimens of this size.23 In

the present study, the specimens were prepared in a beam shape, with a

cross-sectional area of 0.8 mm x 0.8 mm, using the non-trimming technique.

Thus, although none of the specimens failed during preparation, many resin-

enamel beams appear to have developed cracks. As enamel has clearly

proved to be stiffer and consequently more brittle than dentin, it may turn out

that, for proper microtensile bond strength testing, resin-bonded enamel

indeed does require a different specimen size from that of resin-bonded

dentin. However, several limitation of this study must be considered. Cracks

observed between the base of the hybrid layer and the underlying

mineralized dentin might correspond to the ‘submicron hiati’ that can result

from the dehydration artifacts produced during specimen preparation.

Because the specimens beams used in this study were small (0.8 x 0.8 mm),

they easily are susceptible to dehydration. The same may also be true for

the microcracks that were noted along the surface of the hybrid layer.

There is still some controversy as to whether the specimens that fail during

preparation should be included as “zero values” in the computation of the

sample mean bond strength, as proposed by Shono et al.14 Bouillaguet et

al.5, however, preferred to discard the prematurely failed specimens, to avoid

biasing the sample. These authors reported the percentage of specimens in

33

each experimental group that debonded prematurely, and related it to the

mean tensile bond strength measured for that group. Through this analysis it

was revealed that the specimens that failed prematurely most likely had a

bond strength of 13 MPa or lower.5 These specimens had been prepared

with the trimming technique. As the non-trimming method has proved able to

measure bond strengths as low as 5 MPa24, the findings of Bouillaguet’s

study provides some indirect support to the idea that the non-trimming

technique is a less traumatic procedure for microtensile testing.

For a more thorough and meaningful appraisal of the amount of structural

defects exhibited by each experimental specimen prior to testing, it would be

desirable to develop a method that could non-destructively detect interfacial

defects across the entire bonded surface of the stick. A technique able to

provide good resolution images of the specimens without exposing them to

extreme conditions of pressure, temperature, and humidity, would be ideal.

For that purpose, a field-emission environment microscope (FESEM) might

be ideal.

It would also appear desirable to develop a method for a quantitative

assessment of the structural integrity of resin-bonded interfaces in

microtensile specimens before loading. This method might permit the

recording of all the defects seen through a microscopic section of the whole

surface of each specimen, to finally arrive at a void or defect score, which is

a quantitative indicator of its intrinsic strength.

Going a step further, if statistical analysis revealed the existence of a

correlation between the defect score of the stick before testing and its

measured bond strength, the score could be taken as a predictor of the

specimen’s performance under load. Through this quantitative analysis of

samples’ integrity before testing, the degree of aggressiveness of the

different procedures for preparing microtensile specimens could be better

appreciated.

Furthermore, by selecting from the experimental sample those specimens

which are expected to have about the same intrinsic strength, one would be

more confident that what is being tested is the actual bond strength of the

34

adhesive interface. The proposal of a scientific method for assessing the

structural integrity of microtensile specimens before loading will be the aim

of a future study.

In conclusion, the microtensile technique is a versatile and reliable method

to test the quality of adhesion of dental materials to different substrates.

However, microtensile testing should be regarded as a very “technique-

sensitive” method that should be handled with care. In order to make the test

more accurate, a standard procedure for specimens preparation, which

places the least possible stress on the bonds, should be defined.

Furthermore, a scientific method for assessing the structural integrity of the

sticks before loading should be developed, in order to detect those

specimens which, as a result of an intrinsic weakness, might yield bond

strength values that would bias the outcome of the trial.

35

References

1. Van Noort R, Cardew GE, Howard IC, Noroozi S. The effect of local

interfacial geometry on the measurement of the tensile bond strength to

dentin. J Dent Res 1991; 70: 889-93

2. Van Noort R, Noroozi S, Howard IC, Cardew G. A critique of bond

strength measurements. J Dent 1989; 17: 61-67

3. Sano H, Shono T, Sonoda H, Pashley DH. Relationship between surface

area for adhesion and tensile bond strength-evaluation of a microtensile test.

Dent Mater 1994; 10: 236-240

4. Pashley DH, Ciucchi B, Sano H, Yoshiyama M, Carvalho RM. Adhesion

testing of dentin bonding agents. A review. Dent Mater 1995; 11: 117-125

5. Bouillaguet S, Ciucchi B, Jacoby T, Wataha JC, Pashley D. Bonding

characteristics to dentin walls of Class II cavities, in vitro. Dent Mater 2001;

17: 316-321

6. Inoue S, Vargas MA, Abe Y, Lambrechts P, Vanherle G, Sano H, Van

Meerbeek B. Microtensile bond strength of eleven contemporary adhesives

to dentin. J Adhes Dent 2001; 3: 237-245

7. Schreiner RF, Chappell RP, Glaros AG, Eick JD. Microtensile testing of

dentin adhesives. Dent Mater 1998; 14: 194-201

8. Nakajima M, Sano H, Burrow MF, Tagami J, Yoshiyama M, Ebisu S,

Ciucchi B, Russel CM, Pashley DH. Tensile bond strength and SEM

evaluation of caries-affected dentin using adhesives. J Dent Res 1995; 74:

1679-1688

9. Yoshiyama M, Carvalho RM, Sano H, Horner JA, Brewer PD, Pashley DH.

Regional bond strength of resin to human root dentin. J Dent Res 1996; 24:

435-442

10. Yoshiyama M, Sano H, Ebisu S, Tagami J, Ciucchi B, Carvalho RM,

Johnson MH, Pashley DH. Regional bond strength of bonding agents to

cervical sclerotic root dentin. J Dent Res 1996; 75: 1404-1413

11. Zhang Y, Agee K, Nör J, Sachar B, Russel C, Pashley DH. Effect of acid

etching on the tensile properties of demineralized dentin matrix. Dent Mater

1998; 14: 222-228

36

12. Shono Y, Terashita M, Pashley EL, Brewer PD, Pashley DH. Effects of

surface area on resin-enamel tensile bond strength. Dent Mater 1997; 13:

290-296

13. Pashley DH, Tay FR. Aggressiveness of contemporary self-etching

adhesives. Part II: etching effects on unground enamel. Dent Mater 2001;

17: 430-444

14. Shono Y, Ogawa T, Terashita M, Carvalho RM, Pashley DH. Regional

measurement of resin-dentin bonding as an array. J Dent Res 1999; 78:

669-705

15. Tay F, King NM, Suh BI, Pashley DH. Effect of delayed activation of

light-cured resin composites on bonding of all-in-one adhesives. J Adhes

Dent 2001; 3: 207-225

16. Tay FR, Smales RJ, Ngo H, Wei SHY, Pashley DH. Effect of different

conditioning protocols on adhesion of a GIC to dentin. J Adhes Dent 2001; 3:

153-166

17. Armstrong SR, Keller JC, Boyer DB. The influence of water storage abd

C-factor on the dentin-resin composite microtensile bond strength and

debond pathway utilizing a filled and unfilled adhesive resin. Dent Mater

2001; 17: 268-276

18. Armstrong SR, Boyer DB, Keller JC. Microtensile bond strength testing

and failure analysis of two dentin adhesives. Dent Mater 1998; 14: 44-50

19. Armstrong SR, Keller JC, Boyer DB. Mode of failure in the dentin-

adhesive resin-resin composite bonded joint as determined by strength-

based (TBS) and fracture-based (CNSB) mechanical testing. J Dent 2001;

17: 201-210

20. De Hoff PH, Anusavice KJ, Wang Z. Three-dimensional finite element

analysis of the shear bond test. Dent Mater 1995; 11: 126-131

21. Carvalho RM, Santiago SL, Fernandes CAO, Suh B, Pashley DH.

Effects of prism orientation on tensile strength of enamel. J Adhes Dent

2000; 2: 251-257

37

22. Mannocci F, Scheriff M, Ferrari M, Watson TW. Microtensile bond

strength and confocal microscopy of dental adhesives bonded to root canal

dentin. Am J Dent 2001; 14: 100-104

23. Phrukkanon S, Burrow MF, Tyas MJ. The influence of cross-sectional

shape and surface area on the microtensile bond test. Dent Mater 1998; 14:

212-221

24. Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshiyama M, Shono

Y, Fernandes CA, Tay F . The microtensile bond test : A review. J Adhes

Dent 1999; 1: 299-309

38

I.1.2 Influence of Substrate, Shape, and Thickness on Microtensile

Specimens’ Structural Integrity and Their Measured Bond Strengths Goracci C, Sadek FT, Monticelli F, Cardoso PEC, Ferrari M. Dental Materials

2004; 20: 643-654.

Introduction In recent years an increasing number of researchers in the field of dental

adhesion has turned to the microtensile technique of bond strength testing

as supposedly being able to overcome some of the known limitations of

conventional tensile and shear bond strength tests.1-6 More precisely, in the

microtensile technique a purely tensile load is applied on a very small cross-

section of the bonded interface between the dental substrate and the

adhesive material of interest. Over such a limited surface, stress distribution

is expected to be uniform, thus enabling the test measurements to truly

express the interfacial bond strength between dental tissue and material.1-3,6

This appreciation has often been precluded in conventional testing by the

occurrence of cohesive failures in dentin, when working with adhesives able

to establish on this tissue bond strengths higher than 15-20 MPa.1,6 In

addition, the microtensile method has allowed the mapping of bond strength

in different regions or at different depths of dental tissues, thus also

contributing to the understanding of adhesion mechanisms under clinical

conditions.1,7-10

However, this innovative and versatile method of bond strength testing

requires great care, as the technique is very sensitive and time-consuming,

although multiple specimens can be obtained from one single tooth.1

Furthermore, as with any new measurement method, one should wonder

about its accuracy and precision. In other words, research should verify how

close and how consistently microtensile measurements represent the “true”

interfacial bond strength. This issue has a bearing on the internal and

external validity of microtensile studies. The need to in a sense “put on trial”

the microtensile technique itself appears even more evident when

considering that different procedures for specimens preparation and loading

39

are being followed.1 As a result, specimens of different shapes and cross-

sections are being used for testing the adhesion to substrates, which are

also intrinsically different, such as enamel, coronal and root dentin. In other

words, an effort should be made to standardize the protocols of microtensile

tests. This would eventually take the microtensile technique out of its

“experimental phase”, and would definitely prove the claimed superiority of

the method in comparison with conventional tensile and shear bond strength

tests, whose potential for standardization and reproducibility is known to be

poor.6

The influence of cross-sectional shape and surface area on the microtensile

bond test has been the object of a previous study by Phrukkanon et al.11

These Authors compared specimens that exhibited either a cylindrical or a

rectangular cross-section at the bonded interface. However, all of the

specimens had been prepared following the “trimming” version of the

microtensile technique. No direct comparison has yet been made between

the bond strengths yielded by trimmed and untrimmed microtensile

specimens, although there are in the literature indications that a higher

percentage of premature failures and accordingly lower values of yielded

bond strength are associated with the trimming modality.1,12

Additionally, as regards specimens thickness, suggestions have been given

that for adequate testing the cross-sectional area should not exceed 1.5

mm2 1,13 and not be lower than 0.5 mm2 12, but again these guidelines only

apply to hourglass-shaped specimens. No similar indications have so far

been provided for untrimmed specimens.

As finally regards the substrate, a previous scanning electron microscopy

investigation of unloaded specimens, that in all of them revealed the

presence of structural faults possibly influencing the bond strength yielded

under load, also pointed out that the defects were more common in enamel

than in dentin specimens.14 It was then speculated that the brittleness and

low elasticity of enamel, especially in the reduced thicknesses of

microtensile sticks, may render these specimens intrinsically more prone to

failure. This would also explain the relatively low levels of bond strength

40

recorded on enamel as compared with dentin.15 However, in the

aforementioned study only beam-shaped specimens prepared following the

non-trimming technique were analyzed.

In short, as only scattered and incomplete information is available on the

matter, there is the need to systematically assess whether either the shape

or the thickness into which enamel or dentin specimens are prepared can

have an influence on the bond strength recorded under load, in an attempt to

possibly identify the most adequate specimen design for microtensile testing

of enamel or coronal dentin. With this aim, the null hypothesis that neither

the specimen substrate nor its shape or thickness has a significant influence

on the measured microtensile bond strength was tested.

Materials and Methods Tooth preparation Sixty-four extracted third molars were collected. The teeth, which had to be

free of caries and/or previous restorations, were cleansed from any debris,

and stored in a saline solution at 4°C for no longer than three months. The

teeth were meant to provide differently sized and differently shaped

specimens for microtensile bond strength measurements either on enamel

or on dentin. After cutting off the roots at their middle third with a diamond

disc, each tooth was randomly assigned to one of sixteen groups, which

were thus defined:

Group 1: Hourglass-shaped specimens 0.5 mm x 0.5 mm in cross-section,

from enamel; Group 2: Hourglass-shaped specimens 1 mm x 1 mm in cross-

section, from enamel; Group 3: Hourglass-shaped specimens 1.5 mm x 1.5

mm in cross-section, from enamel; Group 4: Hourglass-shaped specimens 2

mm x 2 mm in cross-section, from enamel; Group 5: Hourglass-shaped

specimens 0.5 mm x 0.5 mm in cross-section, from dentin; Group 6:

Hourglass-shaped specimens 1 mm x 1 mm in cross-section, from dentin;

Group 7: Hourglass-shaped specimens 1.5 mm x 1.5 mm in cross-section,

from dentin; Group 8: Hourglass-shaped specimens 2 mm x 2 mm in cross-

section, from dentin. Group 9: Beam-shaped specimens 0.5 mm wide and

41

0.5 mm thick, from enamel; Group 10: Beam-shaped specimens 1 mm wide

and 1 mm thick, from enamel; Group 11: Beam-shaped specimens 1.5 mm

wide and 1.5 mm thick, from enamel; Group 12: Beam-shaped specimens 2

mm wide and 2 mm thick, from enamel; Group 13: Beam-shaped specimens

0.5 mm wide and 0.5 mm thick, from dentin; Group 14: Beam-shaped

specimens 1 mm wide and 1 mm thick, from dentin; Group 15: Beam-

shaped specimens 1.5 mm wide and 1.5 mm thick, from dentin; Group 16:

Beam-shaped specimens 2 mm wide and 2 mm thick, from dentin. Each

group included four teeth. On the teeth which were used to obtain enamel

specimens (Groups 1-4, 9-12), some of the most superficial enamel was cut

off from the buccal or the lingual aspect of the tooth with a cooled diamond

disc on a Labcut 1010 machine (Extec Corp., Enfield, CT, USA)

(Fig.1a).Thus, a flat surface of enamel was exposed. On the teeth meant to

provide dentin specimens (Groups 9-16), all of the occlusal enamel and

some superficial dentin were removed by cutting with a cooled diamond disc,

so as to create a flat surface in mid-coronal dentin (Fig. 1b). The exposed

dental substrate, be it enamel or dentin, was then polished with wet

carborundum papers, in order to create a standard smear layer. Papers of

220, 320, and 400 grit were used in sequence, each one for 10 seconds,

and a final polishing was done with a 600 grit paper for 60 seconds. After

rinsing and gently drying the exposed substrate, the Clearfil SE Bond Plus

adhesive (Kuraray Co., Japan) was applied, and a composite resin block of

approximately 5mmx5mmx5mm was built up on the bonding substrate, using

the proprietary material (Clearfil AP-X, Kuraray Co., Japan) (Fig. 1c).

After a 24-hour storage into a saline solution at 27°C, each bonded tooth

was sectioned with a diamond blade into a series of slabs of a determined

thickness, which could be 0.5mm, 1mm, 1.5mm, and 2mm depending on the

group. Each slab was made up of the dental substrate for about one half and

of the resin build-up for the remaining portion.

42

Fig. 1 (a-c) (a) Enamel specimens preparation involved the removal of a portion of superficial tissue without exposing the underlying dentin.

(a)

(b) Tooth prepared for test on dentin: the occlusal third was removed with a diamond disc, creating a flat surface.

(b)

(c) Resin build-up over the enamel and the dentin surface.

(c)

43

Fig. 1 (d, e) (d) Cutting of the tooth along the X and Y axis and the resulting sticks.

(d)

(e) Procedure for the preparation of hourglass-shaped sticks: the bonded tooth is sectioned in multiple slabs. On each slab the narrowest cross-section is created at the interface by trimming with a bur.

(e)

Then, the teeth meant to provide beam-shaped specimens for microtensile

bond strength testing according to the “non-trimming technique”, were

rotated 90° on the cutting machine and again sectioned lengthwise. By this

second cut a number of sticks were produced, whose width and thickness

could be 0.5mm, 1mm, 1.5mm, and 2mm depending on the group (Fig. 1d).

On the teeth meant to provide hourglass-shaped specimens, each slab was

trimmed with a 2214 FF (KG Sorensen) bur mounted on a high-speed

handpiece, so as to create at the level of the bonding interface a cross

section of either 0.5mmx0.5mm, or 1mmx1mm, or 1.5mmx1.5mm, or

44

2mmx2mm depending on the group. The handpiece was used free-hand and

under a copious spray of water (Fig. 1e).

Every tooth provided at least eight specimens, leading to a greater than

thirty sample size in each tested group, which was considered enough to

allow for the application of an adequate statistical analysis.

Specimens preparation for SEM analysis

From each group, two to three specimens were selected at random for

microscopic analysis. Each of them underwent a gentle surface

decalcification with a 36% phosphoric acid solution for 10 seconds, and a

brief deproteinization at the interface between resin and dentin with a 2%

sodium hypochlorite solution for 60 seconds. The specimens were then

rinsed with water, dehydrated with ascending acetone concentrations (30,

50, 70, 90 and 100%), and critical point dried (CPD 030; Balzers,

Liechtenstein). Lastly, each specimen was mounted on aluminum stubs,

sputter-coated with gold by means of the Edwards Coater S150B device

(Edwards Ltd., London, UK), and observed using a Philips 515 scanning

electron microscope (Philips Co., Amsterdam, The Netherlands).

Only two of the four sides of each specimen could be imaged, as one side

was on the stub and one side was directed away from the scanning beam.

Images taken at relatively low magnifications (x25-x75) provided an

overview of the exterior structure of each specimen at the level of the

interface, whereas with high magnification views (x600-x3000) the

morphology of the hybrid layer was visualized.

Microtensile bond strength measurements

For the purpose of measuring microtensile bond strength, a cyanoacrylate

material (Zapit, Dental Ventures of America, CA, USA) was used to bond the

ends of each specimen to the two free-sliding parts of a specially designed

holding device. The latter is able to transmit to the specimen purely tensile

forces, without any torquing component, when it is mounted on a universal

loading machine (Kratos Dinamometros, Brazil). The tensile load was

applied at a cross-head speed of 0.5 mm/min, until the fracture of the

45

specimen occurred. At this point the load at failure in Kilograms was

recorded, and the specimen’s fragments cautiously removed from the grips

with a scalpel. The cross-sectional area at the site of fracture was measured

to the nearest 0.01 mm with a digital caliper, in order to calculate the bond

strength at failure in MegaPascals.

Statistical analysis

The differences among the microtensile bond strength values measured in

each group were tested for statistical significance using the Univariate

Analysis of Variance with microtensile bond strength in MegaPascals as the

dependent variable and specimen substrate, shape, and thickness as

factors. The Bonferroni test was applied for multiple comparisons. The

Spearman’s Rank Correlation Test was performed to test the significance of

the correlation between specimens thicknesses and their measured

microtensile bond strengths. For all of the analyses the level of significance

was set at p<0.05.

Results The mean values of microtensile bond strength recorded for each of the

sixteen experimental groups are reported in Graph 1.

Bond strength ranged between the highest value of 63.05 MPa, recorded by

0.5x0.5mm thick dentin sticks, and the lowest value of of 17.17 MPa

measured on 2mmx2mm dentin sticks. The variation associated with these

data was within acceptable limits.

Results of the statistical analysis for microtensile bond strength values

When an appropriate statistical analysis (Univariate ANOVA, with

computation of estimated marginal means) was applied to assess whether

substrate, shape, and thickness of the specimens had an effect on their

recorded microtensile bond strength, it appeared that this effect was actually

significant (p<0.05).

46

In particular, as regards the substrate, dentin specimens (39.04±17.20 MPa)

gave significantly higher values of bond strength than enamel specimens

(31.94±11.51 MPa) (p<0.05, graph 2). Graph 1. Mean and standard deviation of the values of microtensile bond strength in MPa for all the tested combinations of specimens substrate, shape, and thickness. In general, higher bond strengths were obtained by dentin versus enamel specimens, and by sticks versus hourglasses. The only one exception was represented by 2x2mm dentin sticks, which gave the lowest bond strengths of all the groups in the study. The highest degree of variability in bond strength was exhibited by 0.5x0.5 mm dentin sticks.

0.5 mm Group 1 63.05±12.23 MPa 1 mm Group 2 55.97±4.05 MPa

1.5 mm Group 3 40.60±7.06 MPa

Sticks

2 mm Group 4 17.17±6.81 MPa 0.5 mm Group 5 47.15±7.73 MPa 1 mm Group 6 42.05±7.17 MPa

1.5 mm Group 7 26.52±7.13 MPa

Dentin Hourglasses

2 mm Group 8 19.85±6.42 MPa 0.5 mm Group 9 51.12±8.36 MPa 1 mm Group 10 36±2.80 MPa

1.5 mm Group 11 25.75±2.68 MPa

Sticks

2 mm Group 12 18.62±4.50 MPa 0.5 mm Group 13 43.63±3.25 MPa 1 mm Group 14 33.85±3.50 MPa

1.5 mm Group 15 25.06±2.65 MPa

Enamel

Hourglasses

2 mm Group 16 21.50±6.25 MPa

dent

in h

ourg

lass

0.5

dent

in s

tick

1

dent

in h

ourg

lass

1

dent

in s

tick

1.5

dent

in h

ourg

lass

1.5

dent

in s

tick

2

dent

in h

ourg

lass

2

enam

el s

tick

0.5

enam

el h

ourg

lass

0.5

enam

el s

tick

1

enam

el h

ourg

lass

1

enam

el s

tick

1.5

enam

el h

ourg

lass

1.5

enam

el s

tick

2

enam

el h

ourg

lass

2

0

10

20

30

40

50

60

MPa

47

Graph 2. Comparison in microtensile bond strength between all of the specimens from dentin and all of the specimens from enamel. The comparison is based on estimated marginal means.

Dentin Enamel Mean sd Mean sd

39.04 MPa 17.20 MPa 31.94 MPa 11.51 MPa The difference is statistically significant (p<0.05)

3232N =

dental substrate

enameldentin

mic

rote

nsile

bon

d st

reng

th

100

80

60

40

20

0

As far as the specimens shape is concerned, sticks (38.53±17.47 MPa) gave

significantly higher values of bond strength than hourglasses (32.45±11.40

MPa) (p<0.05, graph 3). Graph 3. Comparison in microtensile bond strength between all of the stick specimens and all of the hourglass specimens. The comparison is based on estimated marginal means.

Sticks Hourglasses Mean sd Mean sd

38.53 MPa 17.47 MPa 32.45 MPa 11.40 MPa The difference is statistically significant (p<0.05)

3232N =

specimen shape

stickhourgl

mic

rote

nsile

bon

d st

reng

th

100

80

60

40

20

0

48

As for specimens thickness, when this increased from 0.5 mm to 2 mm, the

recorded microtensile bond strength decreased. The differences among the

four tested thicknesses were all statistically significant (p<0.05, graph 4). Graph 4. Comparison in microtensile bond strength (MPa) among all of the 0.5mm specimens, all of the 1mm specimens, all of the 1.5mm specimens, all of the 2mm specimens. The comparison is based on estimated marginal means.

0.5 mm 1 mm 1.5 mm 2 mm Mean sd Mean sd Mean sd Mean sd 51.23 10.72 41.96 9.84 29.40 8.1 19.28 5.66

All of the differences are statistically significant (p<0.05)

16161616N =

specimen thickness

21.510.5

mic

rote

nsile

bon

d st

reng

th

100

80

60

40

20

0

In addition, the negative correlation between specimen thickness and

recorded microtensile bond strength was statistically significant (ρ=-.83;

p<0.05).

Also, the interaction between the factors substrate and shape, substrate and

thickness, shape and thickness were statistically significant (p<0.05, graphs

5, 6, 7).

Results of SEM observations

The most remarkable finding from SEM analysis was the frequent

observation in hourglass-shaped specimens of lines of fracture in the area of

action of the bur during trimming at the bonding interface. This occurrence

was noticed in the totality of enamel hourglasses, and in the great majority

(70%) of dentin trimmed specimens which had been inspected (Fig. 2 a, b).

In addition, some hourglasses broke during the procedures of specimens

preparation for SEM analysis. These failures, that suggested a relatively

49

higher brittleness of hourglasses as compared with sticks, were again more

frequent on enamel than on dentin specimens (Fig. 3a-d). A good integrity of

the specimen at the bonding interface was seen more often on untrimmed

than on trimmed specimens (Fig. 4a, b). Graph 5. Effect of the interaction between specimens substrate and shape on the measured microtensile bond strength. The table reports the mean values of microtensile bond strength for different combinations of specimens substrate and shape. Dentin sticks gave significantly higher values of bond strength than enamel hourglasses (p<0.05). All of the other differences are not statistically significant (p>0.05). In the graph, the columns representing combinations that yielded similar microtensile bond strength values are underlined by the same segment.

Dentin sticks Dentin hourglasses Enamel sticks Enamel hourglasses 44.2 MPa 33.89 MPa 32.87 MPa 31.01 MPa

dent

in s

tick

dent

in h

ourg

lass

enam

el s

tick

enam

el h

ourg

lass

0

5

10

15

20

25

30

35

40

45

50

specimens subgroups

MPa

50

Graph 6. Effect of the interaction between specimens shape and thickness on the measured microtensile bond strength. Table I reports the mean values of microtensile bond strength for different combinations of specimens shape and thickness. In table II, the star sign indicates the statistically significant differences in microtensile bond strength among all the possible combinations (S= sticks; H= hourglasses; *p<0.05; [-] the difference is negative).

Sticks Hourglasses Table I 0.5 mm

1 mm

1.5 mm

2 mm

0.5 mm

1 mm

1.5 mm

2 Mm

MPa 57.08 45.98 33.17 20.67 45.39 37.95 25.80 17.90

stic

k 0.

5

stic

k 1

hour

glas

s 0.

5

hour

glas

s 1

stic

k 1.

5

hour

glas

s 1.

5

stic

k 2

hour

glas

s 2

0

10

20

30

40

50

60

MPa

Table II S 0.5mm S: 1mm 1.5mm* 2mm* H: 0.5mm* 1mm* 1.5mm* 2mm* S 1 mm S: 0.5mm 1.5mm* 2mm* H: 0.5mm 1mm 1.5mm* 2mm*

S 1.5mm S: 0.5mm*[-] 1mm*[-] 2 mm*

H: 0.5mm*[-] 1mm 1.5mm 2mm*

S 2mm S: 0.5mm*[-] 1mm*[-] 1.5mm*[-]

H: 0.5mm*[-] 1mm*[-] 1.5mm 2mm

H 0.5mm H: 1mm* 1.5 mm* 2 mm* S: 0.5mm*[-] 1mm 1.5mm* 2mm* H 1mm H: 0.5mm*[-] 1.5mm*

2mm* S: 0.5mm*[-] 1mm*[-] 1.5mm 2mm*

H 1.5mm H: 0.5mm*[-] 1mm*[-] 2mm*

S: 0.5mm*[-] 1mm*[-] 1.5mm 2 mm

H 2mm H: 0.5mm*[-], 1 mm*[-], 1.5 mm

H: 0.5 mm*[-], 1mm*[-], 1.5mm*[-], 2 mm

51

Graph 7. Effect of the interaction between specimens substrate and thickness on the measured microtensile bond strength. Table III reports the mean values of microtensile bond strength for different combinations of specimens substrate and thickness. In table IV, the star sign indicates the statistically significant differences in microtensile bond strength among all the possible combinations. (E= enamel; D= dentin; *p<0.05; (-) the difference is negative).

Table III Enamel Dentin mm 0.5 1 1.5 2 0.5 1 1.5 2 MPa 47.37 34.92 25.4 20.06 55.10 49.01 33.56 18.51

dent

in 0

.5

dent

in 1

enam

el 0

.5

enam

el 1

dent

in 1

.5

enam

el 1

.5

enam

el 2

dent

in 2

0

10

20

30

40

50

60

Table IV E 0.5mm E: 1mm* 1.5mm* 2 mm* D: 0.5mm 1mm 1.5 mm* 2 mm* E 1mm E: 0.5mm*[-] 1.5mm*

2mm* D: 0.5mm*[-] 1mm*[-] 1.5mm* 2mm*

E 1.5mm E: 0.5mm*[-] 1mm*[-] 2mm*

D: 0.5mm* 1mm* 1.5mm* 2mm*

E 2mm E: 0.5mm*[-] 1mm*[-] 1.5mm

D: 0.5 mm* 1mm* 1.5mm* 2mm

D 0.5mm D: 1mm 1.5mm* 2mm* E: 0.5mm 1mm* 1.5mm* 2mm* D 1mm D: 0.5mm* 1.5mm* 2 mm* E: 0.5mm 1mm* 1.5mm* 2 mm*

D 1.5mm D: 0.5mm*[-] 1mm*[-] 2mm*

E: 0.5mm*[-] 1mm*[-] 1.5mm*[-] 2mm*[-]

D 2mm D: 0.5mm*[-] 1mm*[-] 1.5mm*[-]

E: 0.5mm*[-] 1mm*[-] 1.5mm*[-] 2mm

52

Fig. 2 (a) Microphotograph of an enamel hourglass-shaped specimen 2mmx2mm in thickness. At the bonding interface, in the area of action of the bur, a line of fracture is visible (x25, bar 1 mm, C=composite, E=enamel). (b) The discontinuity is more evident at a higher magnification (x1010, bar 0.1 mm, C=composite, E=enamel).

a

b

C E

C E

53

Fig. 3 (a) Enamel hourglass-shaped specimen 0.5x0.5 mm in thickness that broke as a result of the preparation procedures for SEM observations. The failure occurred through the hybrid layer (x40, bar 1mm, hl=hybrid layer). The area of the interface in square brackets is shown at a higher magnification in Fig. 3b (x1550, bar 10µm, HL=hybrid layer, E=enamel).

a

b

HL

E

hl

54

Fig. 4 (a) Microphotograph of a D stick 0.5x0.5mm in thickness (x75, bar 1mm, C=composite, D=D). (b) At a higher magnification (x600, bar 0.1 mm) a good integrity of the specimen at the bonding interface can be noticed (C=composite, D=D).

a

b

C

D

C

D

55

Discussion The study’s findings reveal that specimen substrate, shape, and thickness all

have a significant influence on the recorded microtensile bond strength.

As regards in particular the effect of the substrate, regardless of specimen

shape and thickness, dentin samples consistently yielded higher values than

enamel samples.

This outcome is in disagreement with the bulk of research data on dental

adhesion provided by conventional tensile and shear bond strength tests,

and in conflict with years of clinical experience.

The relatively lower levels of adhesion achieved in enamel specimens could

be related to the use of a self-etching system as a bonding, as these

simplified adhesives have shown to provide more satisfactory conditions of

bonding on dentin than on enamel.16-18 However, with the microtensile

method, the recording of bond strength values on enamel as high as, if not

lower than on dentin seems to be recurrent. This same result was indeed

already reported in some previous trials, regardless of whether a self-etching

primer or a total-etch adhesive was used as a bonding system15, (Cardoso et

al.: Personal Communication 2001, AADR, Chicago). In the already

mentioned microscopic investigation, where microtensile specimens were

observed using the SEM before being loaded,14 structural defects possibly

undermining their integrity were seen to more often occur in enamel than in

dentin samples. It was then hypothesized that the inherent fragility of enamel

in the small cross-sections of microtensile specimens was responsible for

their failure under relatively low loading levels, as compared with dentin. The

issue was then raised whether for accurate microtensile testing of interfacial

bond strength of enamel, a particular shape or cross-section of the specimen

could be more appropriate.

This question has been addressed in the present study, which showed that,

as for the thickness, despite the fragility of the enamel tissue, the highest

levels of bond strength were associated with the smallest cross sections

(0.5x0.5mm and 1x1mm).

This was in fact true not only for enamel, but also for dentin specimens.

56

Another factor that might have had an effect on microtensile bond strength

measurements on enamel is the orientation of the enamel prisms. As

recently pointed out by Carvalho et al.19, although enamel is considered to

be a very strong substrate, its tensile strength is greatly dependent on the

orientation of the prisms, results being significantly lower when the stress is

directed perpendicular to the long axis of the prisms. For research purposes,

it would be ideal to obtain enamel specimens with their prisms perfectly

oriented either parallel or perpendicular to the applied load. However “this is

practically impossible, due to the random and tortuous disposition of rods

along the enamel structure”.19

The inverse relationship between a specimen’s cross section and measured

bond strength demonstrated in the present study had actually already been

noted in microtensile testing by Sano et al.3, Phrukkanon et al.11, and has

been explained in the light of the Griffith’s Law, dealing with stress

distribution over solids (1920).1-3,11 According to this principle, since internal

defects of specimens are seen as potential stress raisers, smaller specimens

yield higher bond strengths as they contain a lower number of internal

defects, thus allowing for a more homogeneous stress distribution.

As regards the influence of specimen shape on the measured microtensile

bond strength, this investigation revealed that regardless of the dental

substrate and of the specimen cross-section, sticks tended to give higher

values of bond strength than hourglasses. This finding confirms the

suspicion already raised in previous microtensile studies that the trimming

method, by placing an extra-stress at the interface, may in fact weaken the

adhesive bond. Pashley et al.1 as well as Bouillaguet et al.12, after noticing a

higher incidence of premature failures of microtensile specimens with the

trimming technique, prefer the non-trimming method, in particular for

materials or specimens exhibiting relatively low bond strengths. The action

of the bur can be particularly aggressive for the bonding interface when the

trimming is done free-hand, as uneven cutting forces are applied.1,2 In

addition, as the present investigation has pointed out, the trimming

technique can be especially traumatizing for enamel specimens, again in

57

relation to the intrinsic fragility of this tissue. As a matter of fact, when

considering the effect of the interaction between specimen substrate and

shape (Graph 4), enamel hourglasses yielded the lowest values of

microtensile bond strength. Also Carvalho et al.19, in their recent microtensile

study on enamel, mention that enamel microcracks are likely to be created

during specimen preparation due to the action of burs and the generated

heat. Although with fine diamond burs the development of microcracks can

be limited, however it can not be avoided, and, even more importantly, it is

difficult to prevent any transmission of heat to the specimens during

preparation. According to the Authors, these surface microcracks can

seriously weaken a brittle material, such as enamel.19 Also the microscopic

analysis carried out on untested specimens in the present study detected

more numerous signs of compromised structural integrity in hourglasses

than in sticks (Fig. 2, 3, 4). Although it is known that the desiccation

procedure during specimen preparation for SEM analysis may be

responsible for the introduction of cracks, however it is a fact that defects

were consistently seen in a higher number and with greater dimensions in

trimmed specimens as compared with the untrimmed ones. It is therefore

realistic to assume that desiccation may actually have just exaggerated

flaws which were present in the first place.

Another issue that should not be overlooked when dealing with hourglass

specimens regards the geometry of the interface. Especially if the trimming

is done free-hand, it is quite unlikely that the operator will be able to

consistenly produce a perfectly squared or rectangular profile in cross-

section at the bonded interface. More realistically, a trapezius or even a less

defined shape may come out of the trimming. In order to be accurate with

bond strength measurements, it would then become necessary on each

tested specimen to calculate the bonded surface area, by applying the

geometrical formula most appropriate for the cross-sectional shape created

at the trimmed portion.

The present research was prompted by the intention to get an insight into

the different modalities of the microtensile technique, and by the wish to

58

collect some information useful to standardize the procedures. A move in

this direction is needed, as an increasing number of researchers are

resorting to the microtensile technique for dental adhesion testing. Protocols

should be better defined, to make sure that the method is reliable not only

for in-house assessments, but also for comparison of results among different

research laboratories. This dependability has been hoped for from

conventional tensile and shear bond strength tests, but proved quite

disappointing in terms of consistency and potential for reproducibility.6

In an attempt to identify the most appropriate specimen design for

microtensile testing in enamel and coronal dentin, the present study has

come to the following conclusions:

- The null hypothesis that neither the specimen substrate nor its shape or

thickness has a significant influence on the measured microtensile bond

strength has to be rejected;

- Sticks on dentin are the microtensile specimens that yield the highest

levels of bond strength;

- It appears advisable to avoid the trimming action, especially on enamel

specimens;

- If the hourglass shape is preferred, the cross-section area should not

exceed 1mmx1mm, and the trimming should not be done free-hand, in order

to standardize the geometry of the interface.

59

References 1. Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshiyama M, Shono Y,

Fernandes CA and Tay F . The microtensile bond test : A review. J Adhes

Dent 1999; 1: 299-309.

2. Pashley DH, Ciucchi B, Sano H, Yoshiyama M and Carvalho RM.

Adhesion testing of dentin bonding agents. A review. Dent Mater 1995; 11:

117-125.

3. Sano H, Shono T, Sonoda H and Pashley DH. Relationship between

surface area for adhesion and tensile bond strength-evaluation of a

microtensile test. Dent Mater 1994; 10: 236-240.

4. Van Noort R, Cardew GE, Howard IC and Noroozi S. The effect of local

interfacial geometry on the measurement of the tensile bond strength to

dentin. J Dent Res 1991; 70: 889-93.

5. Van Noort R, Noroozi S, Howard IC and Cardew G. A critique of bond

strength measurements. J Dent 1989; 17: 61-67.

6. Sudsangiam S and Van Noort R. Do dentin bond strength tests serve a

useful purpose? J Adhes Dent 1999; 1: 57-67.

7. Yoshiyama M, Carvalho RM, Sano H, Horner JA, Brewer PD and Pashley

DH. Regional bond strength of resin to human root dentin. J Dent Res 1996;

24: 435-442.

8. Yoshiyama M, Sano H, Ebisu S, Tagami J, Ciucchi B, Carvalho RM,

Johnson MH and Pashley DH. Regional bond strength of bonding agents to

cervical sclerotic root dentin. J Dent Res 1996; 75: 1404-1413.

9. Shono Y, Terashita M, Pashley EL, Brewer PD and Pashley DH. Effects

of surface area on resin-enamel tensile bond strength. Dent Mater 1997; 13:

290-296.

10. Cardoso PEC, Braga RR and Carrilho MRO. Evaluation of micro-tensile,

shear and tensile tests determining the bond strength of three adhesive

systems. Dent Mater 1998; 394-398.

11. Phrukkanon S, Burrow MF and Tyas MJ. The influence of cross-sectional

shape and surface area on the microtensile bond test. Dent Mater 1998; 14:

212-221.

60

12. Bouillaguet S, Ciucchi B, Jacoby T, Wataha JC and Pashley D. Bonding

characteristics to dentin walls of Class II cavities, in vitro. Dent Mater 2001;

17: 316-321.

13. Mannocci F, Sheriff M, Ferrari M and Watson TF. Microtensile bond

strength and confocal microscopy of dental adhesives bonded to root canal

dentin. Am J Dent 2001; 14: 100-104.

14. Ferrari M, Goracci C, Sadek F and Cardoso PEC. Microtensile bond

strength tests: scanning electron microscopy evaluation of sample integrity

before testing. Eur J Oral Sci 2002; 110: 385-391.

15. Ferrari M, Goracci C, Monticelli F, Sadek FT and Cardoso PEC.

Adhesion testing with the microtensile method: effects of dental substrate

and adhesive system on bond strength measurements. J Adhes Dent 2002;

4: 291-297.

16. Tay FR, Pashley DH. Aggressiveness of contemporary self-etching

systems. I: depth of penetration beyond dentin smear layers. Dent Mat 2001;

17: 296-308.

17. Pashley DH, Tay FR. Aggressiveness of contemporary self-etching

adhesives Part II: etching effects on unground enamel. Dent Mat 2001; 17:

430-444.

18. Ibarra G, Vargas MA, Armstrong SR, Cobb DS. Microtensile bond

strength of self-etching adhesives to ground and unground enamel. J Adhes

Dent 2002; 4 115-124

19. Carvalho RM, Santiago SL, Fernandes CAO, Suh B and Pashley DH.

Effects of prism orientation on tensile strength of enamel. J Adhes Dent

2000; 2: 251-257.

61

I.2 APPLYING THE MICROTENSILE BOND STRENGTH TEST TO MEASURE THE ADHESION ON ENAMEL AND CROWN DENTIN (stick-forming technique, 1x1mm thick specimens) I.2.1 Adhesion testing with the microtensile method: effects of dental

substrate and adhesive system on bond strength measurements. Cardoso PEC, Sadek FT, Goracci C, Ferrari M. The Journal of Adhesive

Dentistry 2002; 4: 291-297.

Introduction For many years, research in the field of dental materials has employed

conventional shear and tensile bond strength tests to collect and compare

data on the quality of the adhesion effected by bonding systems on dentin

and enamel. However, some in-depth investigations about the physics of

these tests have revealed that their results can be greatly influenced by the

specimen’s geometry and the experimental conditions.25,26

In the attempt to control this variability, the microtensile technique was

developed.17 Successfully applied to several substrates and materials, the

microtensile has proved to be a very versatile and reliable technique.13

The innovative aspect of the microtensile method is to be seen in the small

dimensions of the specimens. These are obtained from the experimental

tooth through a series of cross- and longitudinal sections, and consist of

either hour-glass or beam-shaped sticks, one half of which consists of the

dental tissue, and the other of the restorative material. The two substrates

are held together by an adhesive system at an interface, which has a very

small cross-sectional area, from 0.5 to 1.5 mm2 depending on the technique.

This reduced bonding surface, supposedly with a more uniform structure,

allows for an even distribution of the stress, and thus for a more realistic

appraisal of the bonding mechanism.7,16,17,19,20 As a result of more uniform

stress distribution, the occurrence of cohesive fractures within dentin - that

would invalidate the test by preventing the correct assessment of the

interfacial bond strength - is less frequent. Such events have become

62

unacceptably frequent since the introduction of adhesive systems able to

achieve levels of bond strength to dentin as high as 20-25 MPa.

Another advancement in adhesion testing made possible by the microtensile

method is the ability to better investigate the differences in the bonding

conditions offered by small, contiguous portions of dental tissues, such as

enamel vs dentin, coronal dentin vs root dentin,13,27,28 superficial vs deep

dentin,3,27,28 normal vs adjacent carious, and dry vs wet dentin.11

The ability to obtain multiple specimens from each experimental tooth is

another advantage of the method, as it speeds up and simplifies the sample

production step of the study.9,18

On the other hand, the technique is more demanding when it comes to

specimens preparation and loading. In particular, the cutting and/or trimming

of the microtensile specimens is critical, as it may lead to premature failures

of the bond, most likely due to the transmitted vibrations.3,9,27,28 Obviously,

premature failure makes the specimen useless. Creating the hourglass

shape at the bonding interface with burs is more prone to causing premature

failures. Alternatively, the nontrimming technique, that leaves the

microtensile specimens with a beam shape, appears to be less traumatic,

and has been able to measure bond strengths as low as 5 MPa.3,6,13

Since its introduction, the microtensile method has been variously applied for

the evaluation of several bonding materials, such as traditional three-step

systems, one-bottle adhesives, and self-etching primers.

In the present study, the nontrimming variety of the microtensile technique

was utilized to evaluate the quality of the adhesion created by a self-etching

primer, as compared to a one-bottle adhesive. Specifically, the study aimed

to compare the microtensile bond strength values of Clearfil SE and Excite

on both enamel and dentin. The null hypotheses tested were: a) there is no

difference between the two adhesive materials, and b) sample preparation

can not affect the bond strength results.

63

Materials and Method

Twenty-eight sound human molars which had been extracted for orthodontic

reasons were collected and kept in a saline solution at 37°C until use, but in

any case no longer than three months. Two groups of teeth were randomly

formed. Group A included the samples meant for the testing of the enamel

substrate. On the buccal or lingual aspects of these teeth, a portion of

enamel was removed, in order to create a flat surface in this tissue, without

exposing the underlying dentin (Fig 1a). Fig 1a Tooth prepared for enamel test.

In order to test adhesion to dentin, the entire thickness of enamel layer of

teeth in Group B was cut away with a cooled diamond disk, thus producing

an even surface in middle dentin (Fig 1b). Fig 1b Tooth prepared for the dentin test. One-third of the crown removed using a diamond

disk, obtaining a flat surface.

With the purpose of forming a standard smear layer on the bonding

surfaces, both enamel and dentin were polished following a specific

procedure. This involved grinding the substrate with progressively finer

sandpapers, of 220-, 330-, and 400-grit, for about 10 s with each paper.

Final polishing of the surface was done with a 600-grit sandpaper for 60 s.

The bonding surface was finally washed with water and gently air dried.

64

Each group was then randomly divided into two subgroups on the basis of

the bonding system to be applied. A self-etching primer system (Clearfil SE,

Kuraray, Osaka, Japan), subgroup 1, and a one-bottle adhesive (Excite,

Vivadent, Schaan, Liechtenstein) subgroup 2, were chosen for the trial, and

used according to the manufacturers’ instructions (Table 1). Table 1 Manufacturer-recommended steps for handling the tested

bonding systems Clearfil SE Excite

Apply SE Primer and wait 20s Gently air dry Apply SE Bond Lightly blow with air Light cure for 10s

Apply H3PO4 for 15s Rinse and dry gently Apply the adhesive for 10s Lightly blow with air Light cure for 20s

Following the application of the adhesive system, a 5x5x5 mm composite

resin block (Tetric Ceram, Vivadent) was built on the bonding surface

through the sequential application of 1- to 2-mm-thick layers of material,

each one cured for 40 s with an Optilux 401 light (Demetron, Orange, CA,

USA; 600 mW/cm2 intensity) (Fig 2). Fig 2 Resin block built up over the dentin and enamel surfaces.

After 24-h storage in a saline solution at 37°C, each tooth was multiply

cross- and longitudinally sectioned with a diamond blade, so as to obtain a

variable number of beam-shaped sticks, according to the nontrimming

technique.19 Each stick had a cross-sectional area of about 0.8 mm2 (Fig 3). Fig 3 Sectioning the tooth along the x and y axis to obtain the stick-shaped specimens.

65

Before being tested for bond strength, each stick was carefully checked

under a stereomicroscope at 25x (Bausch&Lomb, Rochester, NY, USA), in

order to verify that the adhesive interface was perpendicular to the long axis

of the stick. This condition is necessary for loading to apply pure tensile

forces, without any undesired torque component. All specimens in which the

adhesive interface was slanted relative to the long axis of the stick were

discarded.

A total of 221 specimens, 124 in dentin and 97 in enamel, were judged

suitable for microtensile testing. This was done with a universal loading

machine (Kratos Dinamometros, Embu, SP, Brazil), and required that the

ends of each stick be carefully glued with Super Bonder gel (Henkel Loctite,

Itapevi, SP, Brazil) to specially designed grips on the machine. The tensile

load was applied at a cross-head speed of 0.5 mm/min, until the stick

fractured. At this point, the load at failure in Kgf was recorded. Each stick

was observed under a stereomicroscope at 25X (Bausch & Lomb) in order to

verify the failure mode, and the stick’s fragments cautiously removed from

the grips with a scalpel. The cross-sectional area at the site of fracture was

measured to the nearest 0.01 mm with digital calipers in order to calculate

the bond strength at failure in MPa. Only the specimens with adhesive failure

were used to calculate the bond strength (Table 2). Table 2 Failure mode of tested specimens

Substrate Adhesive % of cohesive failures

Dentin

Enamel

Clearfil SE Bond Excite

Clearfil SE Bond

Excite

3.1 4.2

5.2 7.0

After checking for normal distribution, the differences in bond strength values

among the four subgroups of specimens were tested for statistical

significance with a two-way ANOVA and Tukey test. The effects of dental

material, substrate, and the interactions were assessed. Significance was

set in advance at p=0.05.

66

Results

In terms of bond strength, Excite obtained mean values for enamel of

45.8±4.7 MPa and 42.9±7.1 for dentin, and Clearfil SE Bond obtained mean

values of 44.5±7.7 for dentin and 38.9±4.8 for enamel (Table 3). Table 3 Bond strength results (MPa)

Substrate Adhesive Mean (SD) Dentin

Enamel

Clearfil SE Bond Excite

Clearfil SE Bond

Excite

44.5 (7.7) 42.9 (7.1)

38.9 (4.8) 45.8 (4.7)

Even though these results are lower than the ones obtained in the other

groups, there were no statistically significant differences among the four

groups analyzed (p>0.05). The standard deviations were low, indicative of

the low variability inherent to the microtensile test.

Discussion

As already mentioned, microtensile testing is a very accurate, but also

technique-sensitive method. A positive finding of the present investigation

was that none of the specimens failed prematurely during preparation. This

is most likely the result of both choosing the allegedly less aggressive

procedure for specimen preparation, the nontrimming technique, and of

carefully applying it. In the original technique, the additional stress placed on

the bond by bur-trimming an hour-glass profile at the interface seems to

increase the likelihood of a premature failure of the bond, as compared with

the nontrimming technique.3,6,13 Additionally, in order to apply the purest

possible tensile load, a point was made of testing only those specimens that,

under microscopic observation, exhibited an adhesive interface

perpendicular to the long axis of the stick. Furthermore, particular care was

taken to fix the microtensile samples in the machine with glue able to

develop adhesive forces on both the specimen and the grip higher than the

loading forces applied.

The cross-head speed of the loading machine was set at 0.5 mm/min.

Different levels of cross-head speed have been used in previous

67

microtensile studies, ranging from 0.2 to 1 mm/min. In a recent investigation

by Inoue et al,9 microtensile bond strength values measured at a 10 µm/min,

20 µm/min, and 50 µm/min cross-head speed were compared; the speed of

20 µm/min was deemed preferable, as related to the best stress distribution.

At the cross-head speed of 50 µm/min the highest values of bond strength

were recorded.

The bond strengths measured in this study were also quite high, ranging

from 38.9 to 45.8 MPa. However, high bond strength values are a

characteristic result of the microtensile method,4,5,12 as compared with those

measured in conventional testing. Furthermore, the observation of relatively

low values of standard deviation is fairly common when dealing with

microtensile, presumably as a result of the accuracy of the method, that

tends to leave little room for variability. In the present study too, a limited

dispersion of the data around the means was noticed (Table 3).

The statistical analysis did not reveal any significant difference in the

adhesion achieved either by the two types of bonding system or on the two

dental substrates. The materials on trial were chosen to represent two

different categories of products currently available on the market for

adhesion to dental tissues. Excite is a one-bottle adhesive, to be used in

combination with phosphoric acid. This is a strong acid that effectively

etches the enamel, whereas on dentin substrate it removes the smear layer

and demineralizes the subsurface dentin, opening the tubules. On the other

hand, Clearfil SE belongs to the class of self-etching primers, less

aggressive acidic solutions which effect a mild demineralization of the dental

tissues, and do not solubilize the smear layer.14 The similarity between the

conditions of adhesion created by Clearfil SE and Excite on dentin is in

agreement with the results of a previous investigation comparing the same

two materials using microtensile testing.24 In contrast, Pashley et al14

reported significantly lower microtensile bond strengths for samples treated

with Clearfil SE as a conditioner (11.6 MPa), in comparison with those in

which the substrates had also been etched with phosphoric acid (27 MPa).

68

A surprising result of the present and a similar previous study5 was the

comparable levels of bond strength achieved by the two adhesives on

enamel and on dentin. This is contrary to the bulk of research data1,4,9 and

years of clinical experience which confirm that enamel offers bonding

conditions more favorable and consistently reliable than dentin. This raises

the question of whether microtensile may is suitable for enamel, which is a

very brittle and anisotropic tissue. In other words, it is plausible that

microtensile enamel specimens fail under relatively low loading forces, due

to the intrinsic fragility of the tissue in thin sections. A parallel scanning

electron microscope investigation of the structural integrity of microtensile

specimens before loading has revealed the presence of a higher number of

microfractures in enamel, as compared with dentin samples.8 This

observation supports the suspicion that enamel may be intrinsically too weak

in the reduced thicknesses of microtensile sticks.

As to the pattern of failure, the majority of the loaded samples exhibited an

adhesive fracture (93% to 96.9%), which is the type most commonly seen

when loading microtensile specimens.16 On the other hand, if the cross

section of samples exceeds 2 mm2, the frequency of cohesive failures

increases.

Conclusions

Within the limits of this study, the null hypotheses are accepted that 1) there

is no difference between the self-etching primer and the one-bottle adhesive

in terms of bond strength, and 2) the bonding conditions provided by either

bonding material on enamel were not significantly better than on dentin. The

majority of the specimens failed adhesively under load.

When carefully handled, the microtensile test provides an accurate and

dependable method for evaluating dental adhesion. The question of whether

enamel’s fragility in thin sections may be a limiting factor for reliable testing

of this substrate needs to be addressed further. Acknowledgments This study was partially supported by NAPEM (Núcleo de Apoio à Pesquisa

em Materiais Dentários). The authors are also grateful to Paulo Santos for the drawings.

69

References 1. Armstrong SR, Boyer DB, Keller JC. Microtensile bond strength testing

and failure analysis of two dentin adhesives. Dent Mater 1998; 14: 44-50.

2. Armstrong SR, Keller JC, Boyer DB. Mode of failure in the dentin-

adhesive resin-resin composite bonded joint as determined by strength-

based (TBS) and fracture-based (CNSB) mechanical testing. J Dent 2001;

17: 201-210.

3. Bouillaguet S, Ciucchi B, Jacoby T, Wataha JC, Pashley D. Bonding

characteristics to dentin walls of Class II cavities, in vitro. Dent Mater 2001;

17: 316-321.

4. Burrow MF, Tagami J, Negishi T, Kikaido T, Hosoda H. Early tensile

bond strengths of several enamel and dentin bonding systems. J Dent Res

1994; 73: 522-528.

5. Cardoso PEC, Mallmann A, Burmann PA. Micro-tensile of self-etching

primer adhesive systems in enamel and dentin [abstract 12]. J Dent Res

2001; 80: 61.

6. Della Bona A, Anusavice KJ, Shen C. Microtensile strength of composite

bonded to hot-pressed ceramics. J Adhes Dent 2000; 2: 305-313.

7. Erickson RL, Glasspoole, Retief DH. Influence of test parameters on

dentin bonding strength measurements [abstract 1543]. J Dent Res 1989;

68: 374.

8. Ferrari M, Goracci C, Sadek F, Cardoso PEC. Microtensile bond

strength tests: SEM evaluation of samples integrity before testing. Eur J Oral

Sci 2002; 110: 385-391.

9. Inoue S, Vargas MA, Abe Y, Lambrechts P, Vanherle G, Sano H, Van

Meerbeek B. Microtensile bond strength of eleven contemporary adhesives

to dentin. J Adhes Dent 2001; 3: 237-245.

10. Mannocci F, Scheriff M, Ferrari M, Watson TW. Microtensile bond

strength and confocal microscopy of dental adhesives bonded to root canal

dentin. Am J Dent 2001; 14: 200-204.

11. Nakajima M, Sano H, Burrow MF, Tagami J, Yoshiyama M, Ebisu S,

Ciucchi B, Russel CM, Pashley DH. Tensile bond strength and SEM

70

evaluation of caries-affected dentin using adhesives. J Dent Res 1995; 74:

1679-1688.

12. Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshiyama M, Shono

Y, Fernandes CA, Tay F. The microtensile bond test : A review. J Adhes

Dent 1999; 1: 299-309.

13. Pashley DH, Ciucchi B, Sano H, Yoshiyama M, Carvalho RM. Adhesion

testing of dentin bonding agents. A review. Dent Mater 1995; 11: 117-125.

14. Pashley DH, Tay FR. Aggressiveness of contemporary self-etching

adhesives. Part II: etching effects on unground enamel. Dent Mater 2001;

17: 430-444.

15. Pereira PNR, Okuda M, Nakajima M, Sano H, Tagami J, Pashley DH.

Relationship between bond strength and nanoleakage: Evaluation of a new

assessment method. Am J Dent 2001; 14: 100-104.

16. Phrukkanon S, Burrow MF, Tyas MJ. The influence of cross-sectional

shape and surface area on the microtensile bond test. Dent Mater 1998; 14:

212-221.

17. Sano H, Shono T, Sonoda H, Pashley DH. Relationship between surface

area for adhesion and tensile bond strength-evaluation of a microtensile test.

Dent Mater 1994; 10: 236-240.

18. Schreiner RF, Chappell RP, Glaros AG, Eick JD. Microtensile testing of

dentin adhesives. Dent Mater 1998; 14: 194-201.

19. Shono Y, Ogawa T, Terashita M, Carvalho RM, Pashley DH. Regional

measurement of resin-dentin bonding as an array. J Dent Res 1999; 78:

669-705.

20. Shono Y, Terashita M, Pashley EL; Brewer PD, Pashley DH. Effects of

surface area on resin-enamel tensile bond strength. Dent Mater 1997; 13:

290-296.

21. Tay F, King NM, Suh BI, Pashley DH. Effect of delayed activation of

light-cured resin composites on bonding of all-in-one adhesives. J Adhes

Dent 2001; 3: 207-225.

71

22. Tay F, Pashley DH. Aggressiveness of contemporary self-etching

systems. I: Depth of penetration beyond dentin smear layers. Dent Mater

2001; 17: 296-308.

23. Tay FR, Smales RJ, Ngo H, Wei SHY, Pashley DH. Effect of different

conditioning protocols on adhesion of a GIC to dentin. J Adhes Dent 2001; 3:

153-166.

24. Van Meerbeek B, Vargas M, Inoue S, Yoshida Y, Peumans M,

Lambrechts P, Vanherle G. Adhesives and cements to preserve

preservation dentistry. Oper Dent 2001; (suppl 6): 119-144.

25. Van Noort R, Cardew GE, Howard IC, Noroozi S. The effect of local

interfacial geometry on the measurement of the tensile bond strength to

dentin. J Dent Res 1991; 70: 889-893.

26. Van Noort R, Noroozi S, Howard IC, Cardew G. A critique of bond

strength measurements. J Dent 1989; 17: 61-67.

27. Yoshiyama M, Carvalho RM, Sano H, Horner JA, Brewer PD, Pashley

DH. Regional bond strength of resin to human root dentin. J Dent Res 1996;

24: 435-442.

28. Yoshiyama M, Sano H, Ebisu S, Tagami J, Ciucchi B, Carvalho RM,

Johnson MH, Pashley DH. Regional bond strength of bonding agents to

cervical sclerotic root dentin. J Dent Res 1996; 75: 1404-1413.

29. Zhang Y, Agee K, Nör J, Sachar B, Russel C, Pashley DH. Effect of acid

etching on the tensile properties of demineralized dentin matrix. Dent Mater

1998; 14: 222-228.

72

I.2.2 Microtensile bond strength to ground enamel and dentin of

simplified adhesives. Goracci C, Sadek FT, Monticelli F, Cardoso PEC, Ferrari M. Journal of

Adhesive Dentistry 2004; 6.

Introduction Since their relatively recent introduction8,9,19, self-etching adhesives have

gained a position of respect on the bonding-system market, especially

attracting the attention of those clinicians who appreciate a simplification of

chairside procedures.

The “user friendliness” of self-etching primers actually goes beyond mere

chairside time savings; it also entails a greater tolerance for the different

moisture conditions of the substrate compared with total-etch adhesives,

which in this regard are more sensitive materials.16,23,24

In addition, since the resin monomers of self-etching primers are supposedly

able to diffuse through the smear layer and the underlying dentin as deeply

as the front of demineralization, the occurrence of post-operative tooth

sensitivity is not expected with these materials.22

Although convincing data have been collected on some of these materials’

bond strength to dentin4,23, the clinical reliability of the adhesion achieved on

enamel still remains questionable.10,11,16,19 It has also been pointed out that

self-etching primers can act as a semi-permeable membrane,21,22 allowing

the water to diffuse from the bonded hydrated dentin to the area between the

adhesive and the uncured composite, thus possibly undermining the long-

term survival of bonded restorations.20-22

As a result of manufacturers becoming aware of these possible limitations,1

some developments have been made in the technology of self-etching

primers. The results of this latest evolution are self-etch adhesives such as

Adper Prompt-L-Pop, Xeno CF II, and AdheSE, which have recently been

launched.

Using the microtensile bond test, the purpose of this trial was to assess the

adhesion achieved on dentin and ground enamel with the self-etching

73

materials mentioned above, in comparison with a conventional total-etch

two-step system, which was tested as a control.

The null hypothesis was that the bond strengths measured for Adper

Prompt-L-Pop, Xeno CFII, and AdheSE were similar to that of the total-etch

system Excite.

Materials and Methods Forty extracted, caries-free human molars lacking previous restorations were

collected and kept in 37°C saline solution (0.9% sodium chloride in water) no

longer than one month before being used in the experiment. The teeth were

randomly divided into the two equally sized samples E and D, for enamel

and dentin bond strength testing, respectively. From each of these samples,

four subgroups (n=5) were randomly formed, and each of them was at

randomly assigned to one of the tested adhesives. The experimental

subgroups were then:

E(1): Adper Prompt-L-Pop (3M ESPE, St Paul, MN, USA, Batch

No.20031104) on enamel

E(2): Xeno CF II (Sankin Kogyo, Tokyo, Japan, Batch No.0302000262)

E(3): AdheSE (Ivoclar-Vivadent, Schaan, Liechtenstein, Batch No.E35883)

on enamel

E(4): Excite (Ivoclar-Vivadent, Schaan, Liechtenstein, Batch No.E37320) on

enamel

D(1): Adper Prompt-L-Pop on dentin

D(2): Xeno CF II on dentin

D(3): AdheSE on dentin

D(4): Excite on dentin

Table 1 reports the chemical composition of the adhesives tested, as well as

the pH values of the self-etching materials.

74

Table 1 Chemical composition of the bonding systems tested.

Adper Prompt-L-Pop

Xeno CF II AdheSE Excite

Universal Catalyst Primer Bond Methacrylated phosphoric acid ester

Water Photoinitiator

(BAPO) Stabiliser Fluoride

complex with zinc

Parabenes

Water Ethanol HEMA

Stabilisers

ME-Pyrophosphate

UDMA Fluoride- releasing

phosphazene monomer Microfiller

Photoinitiators

Phosphonic Acid Ether Acrylates

Bisacrylamide Water

Initiator and stabiliser

Dimethacrylates HEMA Silica

Initiator and stabiliser

HEMA TEGDMA

Phosphoric acid

acrylate Silicon dioxide

Initiators Stabilizers

Alcohol

For bond strength testing on ground enamel, the most superficial portion of

enamel, on the buccal aspect of the tooth was removed by means of an

abrasive paper, in order to create a flat enamel surface (Fig. 1a). Fig. 1a Enamel specimens preparation involved the removal of a portion of superficial tissue without exposing the underlying dentin.

On the teeth meant for dentin bond strength testing, a mid-dentin substrate

was exposed by cutting off the overlying tooth substance with a water-cooled

diamond blade mounted on an Isomet saw (Isomet, Buehler, Lake Bluff, IL,

USA) (Fig. 1b). A clinically relevant smear layer was created both on enamel

and dentin by wet grinding the substrate with 180-grit sandpaper.

75

Fig. 1b Tooth prepared for dentin test: the occlusal third was removed with a diamond disk, creating a flat surface in middle dentin.

At this point, the adhesive systems were applied to the substrate of interest

as recommended by the manufacturers. Following the bonding procedure, a

5 x 5mm composite block was built on the substrate (Fig. 1c), utilizing Tetric

Ceram (Ivoclar-Vivadent, Schaan, Liechtenstein), which were individually

cured for 40s with an halogen light (Optilux 401, Kerr/Demetron, Danbury,

CT, USA; intensity 750 mW/cm2). In order to ensure the proper

polymerization of each added layer of composite, the light tip was positioned

as close as possible to the composite surface. Fig. 1c Composite resin build-up over the enamel and the dentin surface.

Once the composite block had been built up, each tooth was again placed in

a 37° saline solution, where it was kept for 24 h. Then the tooth was secured

with sticky wax on an acrylic resin cylinder, which was mounted on an

Isomet cutting machine (Buehler). By means of a water-cooled diamond

blade, the tooth was sectioned into a series of 0.9-mm-thick slabs.

Subsequently, the tooth was rotated 90 degress and again sectioned

lengthwise, yielding 15 to 20 sticks per tooth with a cross-sectional area of

about 0.9mm x 0.9mm (Fig. 1d).

76

Fig. 1d Cutting of the tooth along the X and Y axis and the resulting stick-shaped specimens.

Each stick was made up for about a half of its length of the substrate and for

the rest by the composite build-up, and the two halves were joined together

at their interface by the bonding material of interest.

After having precisely measured the width and thickness of each stick with a

digital caliper, the specimen was glued with cyanoacrylate (Zapit, DVA,

Corona, CA, USA) to a Geraldeli’s device.18 This jig is made of two parts that

are connected by posthole joints. The Geraldeli’s device was placed in a

Bencor unit, which was mounted on an Instron machine (Instron model

5565, Canton, MA, USA). When the loading machine was activated in

tension, the two rods of the Bencor device moved away from each other,

and so did the two parts of the Geraldeli’s device, following the guidance of

the posthole joints, in such a way that purely tensile forces were applied to

the microtensile stick. The test was run a at a crosshead speed of 0.5

mm/min until specimen failure occurred. The load at failure was recorded in

Newtons and bond strength was calculated in MPa.

Statistical Analysis

The distribution of bond strength data was first checked for normality with

the Kolmogorov-Smirnov test. In order to take into account the tooth-related

variance, bond strength data were considered per tooth, and for statistical

significance calculations the two-way ANOVA for repeated measures was

applied, with the variable “tooth” as a within-subject factor and the variables

“substrate” and “adhesive system” as between-subject factors. Tukey’s test

was then used for multiple comparisons. All analyses were processed using

SPSS 11.0 software (SPSS, Chicago, IL, USA), with significance set at the

95% probability level.

77

Results

The means and standard deviations of the microtensile bond strength values

measured for all of the tested subgroups are reported in Table 2.

Table 2 Mean and standard deviation (sd) of the values of microtensile bond strength in MPa. Dentin Enamel

Bonding system Mean sd Signif. Mean sd Signif.

Adper Prompt-L-Pop 20.16 2.07 c 23.90 4.13 b

Xeno CF II 27.22 2.74 b 27.86 3.28 b

AdheSE 28.48 4.71 b 22.74 4.03 b

Excite 45.80 5.79 a 42.92 4.80 a

Subgroups that are statistically similar (p>0.05) are indicated by the same alphabetical letter

The statistical analysis revealed that neither the dental substrate nor the

interaction of this with the bonding system had a significant effect on bond

strength (p>0.05), whereas the bonding system significantly affected bond

strength (p<0.05).

Both on enamel (42.92±4.80 MPa) and on dentin (45.80±5.79 MPa) Excite

achieved the highest bond strengths of any of the groups, and the difference

was statistically significant (p<0.05). Among self-etching systems on dentin

AdheSE (28.48±4.71 MPa) and Xeno CF II (27.22±2.74 MPa) produced

significantly stronger adhesion than Adper Prompt-L-Pop (20.16±2.07 MPa)

(p<0.05). On enamel, all the self-etching adhesives performed similarly

(p>0.05).

Specimen failures during sample preparation (when sectioning by Isomet)

occurred with Adper Prompt-L-Pop (8 of 76 specimens in dentin, 12 of 78

specimens in enamel); Xeno II (6 of 77 specimens in dentin, 9 of 83

specimens in enamel) and AdheSE (10 of 96 specimens in dentin, 13 of 88

specimens in enamel). No premature failures were recorded for Excite.

Discussion In the present study, newly introduced self-etching adhesives were

evaluated for their bonding ability to enamel and dentin. The microtensile

78

method was chosen for it is considered by the majority of researchers in the

field as the most reliable technique for assessing the “true” interfacial bond

strength between an adhesive material and the substrate of interest. As a

matter of fact, over the small cross section of microtensile specimens, a

uniform stress distribution is expected to occur.4,6,7,14,20

Another advantage of the microtensile method is that multiple specimens

can be obtained from one tooth, thus reducing the number of teeth

necessary to obtain a data set of sufficient statistical power.14 In this trial, for

example, the five teeth per group provideda total of 75 to 100 microtensile

sticks, which can be considered a large enough sample for proper statistical

interpretation.

Among the different varieties of the microtensile technique, the nontrimming

modality producing 0.9mmx0.9mm sticks was preferred, as the outcome of a

preliminary research suggested that this thickness and avoidance of

trimming result in specimens which are less prone to premature failure.7 The

present trial confirmed this, in that the premature failures were within

acceptable limits of frequency for all the tested groups (ranging between

7.8% to 14.8%).

The total-etch one-bottle adhesive was chosen as a control material, since

data from previous microscopic investigations and bond strength tests

indicate the bonding ability of this adhesive to be satisfactory on both

enamel and dentin.1-4,12,13 Excite relies on 37% phosphoric acid for enamel

and dentin conditioning. Microscopic images of the bonding interfaces

developed with Excite have shown that the smear layer is effectively

removed, thus allowing the resin to penetrate inside the tubules and to

infiltrate the underlying demineralized dentin.4

In contrast, several investigations of self-etching adhesives have pointed out

that these materials are able to only modify or partially remove the smear

layer, and the hybrid layer formed is relatively thin compared to that

produced by total-etch adhesives.16,23 These microscopic aspects of a less

solid bond are reflected by the lower values of bond strength measured with

the microtensile method.16,23

79

Although significantly inferior to that of total-etch systems, the bond strength

of the self-etching adhesives tested in this trial exceeded 20 MPa, which is

considered sufficient to resist contraction stresses and to attain gap-free

margins in resin composite restorations.5,15

Among the tested self-etching materials, AdheSE performed acceptably on

both substrates. The presence of silica particles as filler, thought to add to

the intrinsic strength of the adhesive, may have contributed to this positive

result, if one accepts Pashley’s theory that a correlation exists between the

strength of the adhesive and that of the adhesive-hard tissue bond.23

The same can be said of Xeno CF II, which, according to the manufacturer,

also contains filler particles of nanometric dimension.

In terms of Adper Prompt-L-Pop, one of the improvements that the

manufacturer claims to have made over the original version of the material

regards viscosity, which has been increased in order to make the fluid

thicker and firmer on application. One of the limitations of Prompt-L-Pop is

its low viscosity. The consistency of the adhesive is such that, in order for it

to achieve an adequate strength and avoid being dislodged upon placement

of the composite resin, it has been advised to apply the material in more

than one layer.17 Despite these improvements, the performance of Adper

Prompt-L-Pop in terms of bond strength was not impressive especially on

the dentinal substrate, where the material exhibited significantly lower bond

strength values than the other self-etching adhesives tested.

The susceptibility of polymerized resin matrices to hydrolytic degradation

has become a concern in particular for the self-etching materials containing

high concentrations of acidic resin monomers that draw water,21,24 and

especially since it has been proven that single-step adhesives behave as

semi-permeable membranes.22 As a result, they allow for a movement of

water between the interface and the underlying dentin that may accelerate

the process of resin leaching, thus undermining the long-term durability of

the bond.24

In this regard it is worth mentioning that the system AdheSE is provided with

new resin monomers, the phosphonic acid ether acrylates, which are

80

claimed to have a high hydrolytic stability, thus facilitating a long-lasting

high-strength bond. Unfortunately, the present study could only quantify the

early bond strength of the adhesives, whereas no inference could be made

from the collected data about the durability of the bond.

Self-etch adhesives rely on acidic resin monomers for substrate

conditioning. According to the classification of Tay and Pashley23 based on

acidity, all the tested self-etching systems could be categorized as

“aggressive”, low-pH materials (Table 1). Since the etching potential of the

tested materials as determined by pH is comparable, it can be hypothesized

that it was indeed the intrinsic strength of the adhesive that made the

difference in bond strength between AdheSE and Xeno on the one hand,

and adheSE and Adper Prompt-L-Pop on the other. Tay & Pashley23 came

to similar conclusions for Prompt-L-Pop, after having measured relatively low

values of microtensile bond strength, in spite of the presence of acceptably

thick hybrid layers.

With the materials tested in this trial, adhesion does not seem to be

significantly affected by the substrate. However, it should be mentioned that

Xeno CF II and Adper Prompt-L-Pop achieved a higher level of bond

strength on enamel than on dentin. This may be an indication that the newer

self-etching systems have been improved in their ability to bond to enamel,

as compared with the firstly produced materials which have shown

limitations in that application.10,11,16,19

Conclusion The results of the test lead to a rejection of the null hypothesis: the bond

strengths achieved on enamel and dentin by recently developed self-etching

adhesives, though similar and compatible with clinical use, are significantly

lower than those of a total-etch one-bottle adhesive tested as control. Self-

etching systems, though simplifying the handling procedure as compared

with the total-etch systems, do not equal this material in terms of bond

strength. The results of this microtensile study should ideally be

81

complemented with microscopic observations and validated by the findings

of an in vivo trial.

Clinical relevance Notwithstanding the simplified handling, the tested self-etching adhesives

established a significantly weaker bond to enamel and dentin compared with

a total-etch one-bottle adhesive.

82

References 1. Armstrong S, Vargas MA, Fang Q, Laffoon JE. Microtensile bond

strength of a total-etch 3-step, total-etch 2-step, self-etch 2-step, and a self-

ecth 1-step dentin bonding system through 15-month water storage. J Adhes

Dent 2003;5:47-56.

2. Bouillaguet S, Gysi P, Wataha JC, Ciucchi B, Cattani M, Godin C, et al.

Bond strength of composite to dentin using conventional, one-step, and self-

etching adhesive systems. J Dent 2001; 29:55-61.

3. Brackett WW, Covey DA, St Germain HA, Jr. One-year clinical

performance of a self-etching adhesive in class V resin composites cured by

two methods. Oper Dent 2002;27:218-222.

4. Cardoso PE, Sadek FT, Goracci C, Ferrari M. Adhesion testing with the

microtensile method: effects of dental substrate and adhesive system on

bond strength measurements. J Adhes Dent 2002;4:291-297.

5. Davidson CL, de Gee AJ, Feilzer A. The competition between the

composite-dentin bond strength and the polymerization contraction stress. J

Dent Res 1984;63:1396-1399.

6. Ferrari M, Goracci C, Sadek F, Cardoso PEC. Microtensile bond

strength tests: scanning electron microscopy evaluation of sample integrity

before testing. Eur J Oral Sci 2002;110:385-391.

7. Goracci C, Sadek FT, Monticelli F, Cardoso PEC, Ferrari M. Infulence of

substrate, shape, and thickness on microtensile specimen’s integrity and

their measured bond strength. Dent Mater 2004;20:643-654.

8. Gordan VV, Vargas MA, Cobb DS, Denehy GE. Evaluation of adhesive

systems using acidic primers. Am J Dent 1997;10:219-223.

9. Hannig M, Reinhardt KJ, Bott B. Self-etching primer vs phosphoric acid:

an alternative concept for composite-to-enamel bonding. Oper Dent

1999;24:172-80.

10. Hara AT, Amaral CM, Pimenta LA, Sinhoreti MA. Shear bond strength of

hydrophilic adhesive systems to enamel. Am J Dent 1999;12:181-184.

83

11. Ibarra G, Vargas MA, Armstrong SR, Cobbb DS. Microtensile bond

strength of self-etching adhesives to ground and unground enamel. J Adhes

Dent 2002;4:115-124.

12. Inoue S, Vargas MA, Abe Y, Yoshida Y, Lambrechts P, Vanherle G, et

al. Microtensile bond strength of eleven contemporary adhesives to dentin. J

Adhes Dent 2001 ;3 :237-245.

13. Manhart J, Chen HY, Mehl A, Weber K. Hickel R. Marginal quality and

microleakage of adhesive Class V restorations. J Dent 2001;29:123-130.

14. Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshiyama M, Shono

Y, et al. The microtensile bond test: a review. J Adhes Dent 1999;1:299-309.

15. Pashley DH, Sano H, Ciucchi B, Yoshiyama M, Carvalho R. Adhesion

testing of dentin bonding agents: A review. Dent Mater 1995;11:117-125.

16. Pashley DH, Tay FR. Aggressiveness of contemporary self-etching

adhesives. Part II: etching effects on unground enamel. Dent Mater

2001;17:430-444.

17. Pashley EL, Agee KA, Pashley DH, Tay FR. Effects of one versus two

applications of an unfilled, all-in-one adhesive on dentine bonding. J Dent

2002;30:83-90.

18. Perdigão J, Geraldeli S, Carmo ARP, Dutra HR. In vivo influence of

residual moisture on microtensile bond strengths of one-bottle adhesives. J

Esthet Rest Dent 2002;14:31-38.

19. Perdigão J, Lopes L, Lambrechts P, Leitao J, Van Meerbeek B,

Vanherle G. Effects of a self-etching primer on enamel shear bond strengths

and SEM morphology. Am J Dent 1997;10:141-6.

20. Sano H, Shono T, Sonoda H, Takatsu T, Ciucchi B, Carvalho R, et al.

Relationship between surface area for adhesion and tensile bond strength-

evaluation of a micro-tensile bond test. Dent Mater 1994;10:236-240.

21. Tay FR, King NM, Chan KM, Pashley DH. How can nanoleakage occur

in self-etching adhesive systems that demineralize and infiltrate

simultaneously? J Adhes Dent 2002;4:255-269.

22. Tay FR, Pashley DH, Suh BI, Carvalho RM, Itthagarun A. Single-step

adhesives are permeable membranes. J Dent 2002;30:371-382.

84

23. Tay FR, Pashley DH. Aggressiveness of contemporary self-etching

systems. I: Depth of penetration beyond dentin smear layers. Dent Mater

2001;17:296-308.

24. Tay FR, Pashley DH. Dental adhesives of the future. J Adhes Dent

2002;4:91-103.

25. Tay FR, Pashley DH, Suh BI, Carvalho RM, Itthagarun A. Single-step

adhesives are permeable membranes. J Dent 2002;30:371-82.

85

CHAPTER II APPLYING THE MICROTENSILE TEST TO MEASURE BOND STRENGTH TO RADICULAR DENTIN II.1 The adhesion between fiber posts and root canal walls: comparison

between microtensile and push-out bond strength measurements. Goracci C, Tavares AU, Fabianelli A, Monticelli F, Raffaelli O, Cardoso PEC,

Tay F, Ferrari M. European Journal of Oral Sciences 2004; 112: 353-361.

Introduction A new interest has been directed toward bonding to root canal dentin since

the introduction of endodontic fiber posts, which are retained by resin

cements used in combination with adhesive systems.

The histologic characteristics of dentin on endodontically-treated canal walls,

as well as the properties of the different materials available for bonding

make the cementation of fiber posts a unique adhesive procedure. The

action of sodium hypochlorite, hydrogen peroxide, EDTA, or other irrigants

on dentin collagen,1 the peculiar conditions of hydration in root canal dentin

as a result of pulp removal, the type of agent used for substrate conditioning,

the polymerization stress of resin cement in root canals with unfavorable

cavity configuration factors, and the chemical and physical properties of the

posts are all variables that can possibly influence the quality of adhesion at

the post-cement-adhesive-dentin interfaces.2

Moreover, the quality of adhesion to root dentin is affected by the density

and orientation of dentin tubules at different levels of the root canal walls,3

and the accessibility of the coronal, middle, and apical third of the root during

handling of the materials.4

Bond strength has been measured through the conventional tensile test on

external root dentin,5 or on the endodontic surface with the pull-out6-11 and

the push-out methods.12-14 The latter has the benefit of more closely

simulating the clinical condition.15 However, it was suggested that a highly

non-uniform stress may be developed at the adhesive interface when the

86

push-out test is performed on the whole post,14 or on thick root

sections.12,15,16 This may explain the relatively low levels of bond strength

that have been reported when applying this method of adhesion testing.16

Using small-sized specimens, the microtensile method of adhesion testing

permits a more uniform stress distribution along the bonded interface.17 In

addition, this technique enables the measurement of bond strength to very

small areas such as the inside of a root canal. It also allows the assessment

of regional differences in adhesion at the three levels of the root canal.2

The microtensile method has already been applied to evaluate bond strength

to root canals that were treated with different irrigants and bonding systems.

However, only the bond strength of adhesive cements to root dentin has so

far been assessed.1,2,16,18,19 On the other hand, the microtensile method has

not yet been applied to quantify the adhesion achieved at the post-cement-

adhesive-dentin interfaces, when a prefabricated post is luted into a

prepared root canal simulating the clinical condition.

Both the “trimming” and “non-trimming” modalities of the microtensile

technique have been utilized to obtain microtensile specimens from root

dentin.1,2,16,18,19 When looking at some of the collected data, however, a

rather frequent occurrence of premature failures of the specimens, together

with a fairly high variability of the values can be observed.2,18

In addition, the findings of different studies regarding the regional differences

in bond strength among the three levels of root canals are not

consistent.1,2,16,18-20 In one study, the most satisfactory conditions of

adhesion were reported to be present in the apical and coronal third of the

root canals.18 These observations are in contrast with the results of other

microscopic studies, showing that the most reliable bond was usually

created in the coronal third, due to the easier access available in this portion

of the root canals.3,4

Based on these considerations, the first objective of the present study was to

compare the push-out method and both the “trimming” and “non-trimming”

variants of the microtensile technique in their ability to accurately measure

the bond strength of fiber posts luted into prepared root canals. The

87

collected bond strength data were analyzed with specific emphasis on the

variability of the data distribution, which was taken not only as indicators of

the consistency and reproducibility of bonds to root dentin, but also as a

proof of the reliability of each experimental method, in measuring the

adhesion to this peculiar substrate.

The second objective of this study was to investigate whether each testing

method revealed the existence of significant differences in the conditions of

adhesion created at different root levels by two resin cements.

Materials and Methods Thirty upper anterior teeth, that had been extracted for periodontal reasons,

were endodontically treated and restored with a cylindrical 1.6 mm diameter

glass fiber post (Ghimas White posts, Ghimas, Casalecchio di Reno,

Bologna, Italy), and a composite core (Table I). Table I

Fiber posts Bonding system and resin cement Ghimas White

Excite DSC, VariolinkII RelyX Unicem

ExciteDSC Powder Liquid HEMA, TEGDMA Phosphoric acid acrylate Silicon dioxide Initiators Stabilizers Alcohol

Glass fillers Silica Calcium hydroxide Self-cure initiators Pigments

Methacrylated phosphoric esters Dimethacrylates Acetate Stabilizers Self-cure initiators Light-cure initiators

VariolinkII

Longitudinal glass fibers of 12 µm in diameter embedded in epoxy resin. -Fiber density: 30/mm². -Elastic modulus: 20 GPa

Base Dimethacrylates Silicon dioxide Self-cure initiators Light-cure initiators

Catalyst Dimethacrylates Silicon dioxide Self-cure initiators Light-cure initiators

The crown portion of each tooth was removed by cutting with a diamond

blade under copious water cooling below the cementum-enamel junction,

and perpendicularly to the long axis of the tooth. The roots were

endodontically instrumented at a working length of 1 mm from the apex with

a #35 master apical file. A step-back technique was used with stainless-steel

K-files, Gates-Glidden drills #2 to #4 (Union Broach, New York, NY, USA),

and a 2.5% sodium hypochlorite irrigation. The roots were obturated with

88

thermoplasticized injectable gutta-percha (Obtura, Texceed Corp., Costa

Mesa, CA, USA), and a resin sealer (AH-26, DeTrey, Zurich, Switzerland).

Then, part of this filling material was removed with burs, and the canal wall

of each specimen was enlarged with low-speed post drills provided by the

manufacturer, in order to create a 9 mm deep post space, as measured from

the cementum-enamel junction on the buccal aspect of the tooth.

At this point, the specimens were randomly divided into two groups of 15

teeth each, depending on the materials used to lute the post.

In Group A, the Excite DSC bonding system was used in combination with

the Variolink II dual-cure resin cement (Ivoclar-Vivadent, Schaan,

Liechtenstein, Table I). The root canal walls were etched with 37%

phosphoric acid gel (Total-Etch, Ivoclar-Vivadent, Schaan, Liechtenstein).

The gel was introduced in the canal through a needle, and after 15 seconds

it was washed with an endodontic syringe. Excess water was removed from

the post space with a gentle blow of air and with paper points, leaving the

dentin slightly moist. Then, by means of the Excite DSC Endo microbrush

two coats of the dual-cure one-bottle adhesive were applied into the root

canal. The priming-adhesive solution in excess was absorbed by a paper

point. The base and catalyst components of Variolink II were then mixed

according to the manufacturer’s instructions (Table I). The resin cement was

brought into the root canal space with a lentulo drill, and the post was

seated. The excess resin was removed and light-curing was performed

through the post for 20 seconds with a halogen light (Optilux 501

Kerr/Demetron, Orange, CA, USA, 750 mW/cm2).

In Group B the RelyX Unicem (3M-ESPE, St. Paul, MN, USA) self-adhesive

resin cement was used to lute the posts (Table I) (21). In handling this new

material, the manufacturer’s instructions were followed rigorously. As RelyX

Unicem does not require any pretreatment of tooth structures, the root canal

was gently dried with an air syringe and paper points, always avoiding

overdrying. The mixed cement was applied only onto the post surface, as

recommended by the manufacturer. Then, the post was directly inserted into

89

the root canal and left undisturbed for the time required by the cement to

auto-cure.

All the post-cemented roots were placed in water at room temperature.

Within one week, specimens for microtensile and push-out bond strength

testing were prepared from each bonded assembly.

Preparation of hourglass-shaped specimens for microtensile bond strength

testing

For this test, twelve roots were randomly selected, with six roots derived

from Group A and the other six from Group B.

Each post-cemented root was sectioned perpendicular to its long axis into a

series of 1mm thick slices with a water-cooled diamond blade on an Isomet

machine (Buehler, Lake Bluff, IL, USA) (Fig. 1a). Fig. 1. Procedure for the preparation of trimmed microtensile specimens. (a) The post cemented root is tranversely sectioned into a series of 1 mm-thick slices. (b) By means of a water-cooled diamond bur, each slice is trimmed to an hourglass profile. As a result of the trimming the tensile load is concentrated at the post-cement-dentin interfaces.

a

b Serial sections were made up to the gutta-percha of the apical filling. A total

of 59 microtensile slabs were obtained from the Group A roots, and 51

specimens were obtained from the Group B roots (Table II).

90

The thickness of each slab was precisely measured with a digital caliper and

was always within the 0.9-1.1 mm range.

Specimen preparation was completed by trimming each slab to an hourglass

profile by means of a flame-shaped diamond bur (2214FF, KG, Sorensen),

mounted on a high speed handpiece and used under continuous water spray

(Fig. 1b). The bur progressed all the way through the dentin, the adhesive,

and the cement thickness up to reach the post surface. The trimming action

of the bur was checked under a stereomicroscope (Bausch&Lomb,

Rochester, NY, USA), to make sure that the bur tip would only lightly touch

the post surface at one point, without disrupting its structure.

As a result of the trimming, when the two opposite sides of the hourglass are

placed under tension (Fig. 2), the tensile stress is concentrated at the post-

cement interface, over a bonding area which can be calculated by

measuring the post diameter and the thickness of the slice with a formula

used previously by Bouillaguet et al.2 (Fig. 2). Fig. 2. Mathematical formula used to measure the area of the post surface at the post-root interface of microtensile trimmed specimens.

Each trimmed slice was glued with cyanoacrylate (Zapit, Dental Ventures of

America, CA, USA) to the two free sliding components of a holding device,

which was mounted on a universal testing machine (Kratos Dinamometros,

arc θ

arc=r x 2sin θ - 1 x (c/2r) arc

r c θ

arc=r x 2sin θ - 1 x (c/2r) bonded area = arc x slice thickness

91

Brazil). This set-up was conceived to apply purely tensile forces to the two

opposite halves of the hourglass-shaped specimen, transmitting the stress to

the post-cement-adhesive-dentin interfaces. Each specimen was loaded at a

cross-head speed of 0.5 mm/min until failure occurred. Bond strength was

expressed in megaPascals (MPa), by dividing the load at failure (Newtons)

with the bonding surface area (mm2).

Preparation of beam-shaped specimens for microtensile bond strength

testing

To prepare microtensile specimens with the beam version of the technique,

roots with the lowest possible degree of curvature were selected, so that the

post can be inserted with its long axis very parallel to the long axis of the

root. A total of six roots with this characteristic were chosen, three from

Group A and three from Group B.

In order to allow for adequate specimen gripping on the loading machine, a

composite build-up was made on the external root surface (Fig. 3a).

The outer surface of each root was cleaned from any remnant of the

periodontal ligament, pumiced, and etched with 37% phosphoric acid for 15

seconds. Then, the one-bottle adhesive Excite DSC (Ivoclar-Vivadent,

Schaan, Liechtenstein) was applied, air-thinned, and light-cured for 5

seconds. For building up the bulk of composite, Tetric Ceram (Ivoclar-

Vivadent, Schaan, Liechtenstein) was used following the incremental

technique.

A first cut was initially made along the long axis of the root at the outermost

periphery of the post with the diamond saw under copious water cooling, so

as to expose it throughout its length (Fig 3b). A similar second cut was done

to expose the post surface on the opposite side (Fig. 3b), so as to create a

slab of uniform thickness with the post in the center of the slab, and the root

dentin portion overlaid by the composite build-up on each side (Fig. 3c).

The precision of the two longitudinal cuts was checked with magnifying

glasses. From the slab, beams about 1 mm by 1mm in thickness were then

serially sectioned (Fig. 3d; Fig. 4a). The width and depth of each stick were

precisely measured with a digital caliper.

92

Fig. 3. Procedure for the preparation of beam-shaped microtensile specimens. (a) A composite build-up is created on the external root surface to allow for adequate specimen gripping on the loading machine. (b) As a result of two longitudinal cuts running at the post periphery throughout its length, a root slab is created exhibiting the luted post in the center, overlaid on each side by the root dentin portion and the composite build-up (c). (d) From the slab 1-mm thick beams are serially sectioned.

Two sticks from the Variolink II group and two sticks from the RelyX Unicem

group were processed for scanning electron microscope (Philips 515, Philips

Co., Eindhoven, The Netherlands).

The preparation procedure for SEM observations involved gentle

decalcification with 32% phosphoric acid for 30 seconds, and

deproteinization by immersion in a 2% sodium hypochlorite solution for 2

P

c c

d d

cb

cb

a

b

c d

93

minutes. Then, after abundant rinsing with water and careful drying, the

sticks were sputter-coated with gold (Edwards Coater, Ltd, London, UK), and

observed at different magnifications (Fig. 4a). Fig. 4. SEM images of beam-shaped microtensile specimens from the Excite-Variolink II group (P=post, C=resin cement, D=dentin). (a) Intact unloaded specimen (bar 1mm ) (b) Prematurely failed specimen (bar 1mm).

a

b

PC

D

94

The rest of the prepared sticks were meant for microtensile bond strength

testing with the loading apparatus previously described.

As the bonded interface was curved, its area had to be calculated using the

same mathematical formula used for hourglass-shaped specimens.2

Preparation of specimens for the push-out bond strength test

Six Group A and six Group B roots were tested with this protocol. The

portion of each root into which the post extended was horizontally sectioned

into five or six 1 mm-thick serial slices with a water-cooled diamond blade

(Labcut 1010 machine Extec Corp., Enfield, CT, USA) (Fig. 5a). A total of

34, 16 coronal and 18 middle-apical sections, were obtained from Group A.

Group B roots provided 33 slices, 15 from the coronal and 18 from the

middle-apical level (Table II).

After measuring the thickness of each slice with a digital caliper, the post

was loaded with a 1.5 mm-diameter cylindrical plunger. The plunger tip was

sized and positioned to touch only the post, without stressing the

surrounding root canal walls. The load was applied on the apical aspect of

the root slice and in an apical-coronal direction, so as to push the post

toward the larger part of the root slice, thus avoiding any limitation to the

post movement possibly due to the root canal taper. Care was also taken to

ensure that the contact between punch tip and post section occurred over

the most extended possible area, in order to avoid any notching effect of the

punch into the post surface, which would have interfered with the

assessment of bond strength at the post-root interface (Fig. 5b).

Loading was performed on a testing machine (Controls S.p.A., Milano. Italy)

at a speed of 0.5 mm/min until bond failure occurred. Bond failure was

manifested by the extrusion of the post from the root section.

In order to express the bond strength in MPa, the load at failure recorded in

Newtons was divided by the area of the bonded interface, which was

calculated through the following formula:

A= 2πr x h

where π is the 3.14 constant, r is the post radius, and h is the thickness of

the slice in mm.

95

Fig. 5. Procedure for the preparation of “micro-push-out” specimens. (a) The post cemented root is sectioned into 1 mm-thick slices. (b) On each slice the post is loaded until bond failure occurs and the post fragment is extruded from the root slice.

Table II. Number of prepared specimens and number of specimens providing useful bond strength measurements with the testing methods under investigation. Microtensile trimming Microtensile non-trimming Push-out

Group A

59 3 34 Number of prepared

specimens Group B

51 2 33

Group A

49 0 34 Number of tested specimens

Group B

37 0 33

Statistical Analysis

In each set of data the coefficient of variation, i.e. the ratio between standard

deviation and mean, was calculated as a parameter of the consistency or

reproducibility of the bonds. In addition, the Kolmogorov-Smirnov test was

used for checking for the normality of data distribution.

Then, in order to assess the significance of the differences in microtensile

bond strengths recorded by the two luting systems on trial with each different

testing modality, the t-test or its non-parametric equivalent, the Mann- Whitney U test; was applied, with the probability level defined at 95%

(p<0.05).

Assessing the regional variability of bond strength to root dentin

A further objective of the statistical analysis was to test the hypothesis that

the coronal, middle, and apical portions of the root canals provide different

a b

96

conditions of adhesion. For this purpose, the microtensile bond strength data

from the same root level of each group were pooled together, regardless of

the root of origin, and the levels of adhesion achieved in the three distinct

portions of the roots were compared for each material.

Results Microtensile bond strength test, trimming technique

The first striking finding of this part of the study was the frequent occurrence

of premature failures. In Group A, 10 out of 59 specimens (16.9%) failed

prematurely during the cutting or gluing phases. In Group B the number of

premature failures was even higher, with 14 out of 51 (27.5%) specimens

debonded during the cutting or gluing phases (Table II).

When all the null bond strength values were included in the statistical

computations, the microtensile bond strength for Group A was 12.3±11.1

MPa, with a coefficient of variation of 0.90. The bond strength in Group B

was 9.1±10.3 MPa, with a coefficient of variation of 1.13.

In addition, the Kolmogorov-Smirnov test revealed that the recorded data

were not normally distributed (Fig. 6).

When null bond strength values were excluded from the data, the standard

deviations of both groups remained quite high. The coefficients of variation

were 0.71 and 0.80 for Group A and Group B respectively, and the data

were still not normally distributed.

The statistical results did not alter significantly when a cubic model of

regression was employed, with each zero bond value replaced by a figure

estimated on the basis of the values trend expressed by the root of origin.2

The microtensile bond strength data remained quite spread out (coefficient

of variation: 0.72 for Group A, and 0.84 for Group B), and their distribution

was still not normal.

When comparing the microtensile bond strength of the two experimental

groups with the Mann-Whitney U test, which is the appropriate test for this

purpose when dealing with non-parametric data, no significant difference

was found between the two materials tested in any of the evaluated

97

statistical settings, i.e. including, excluding or replacing null bond strengths

(p>0.05). Fig. 6. Measures of central tendency and spread of the microtensile bond strength values in

MPa measured with the trimming variant of the microtensile technique. Mean, standard

deviation, and coefficient of variation are calculated taking into account the zero bonds. In the

chart, the length of each box represents the interquartile range of microtensile bond strength

data in MPa for the materials on trial. Also, the size of each sample is reported on the x-axis. Groups Mean and Standard

Deviation Coefficient of variation

1. Excite DSC+Variolink II 12.34±11.13 0.90 2. RelyX Unicem 9.12±10.32 1.13

5951N =

ADHCEM

Excite+Variolink+UniUnicem+FiltekFlow

MIC

RO

TEN

50

40

30

20

10

0

-10

When the data from the same level of the root in each material were pooled

together, posts that were luted with RelyX Unicem (Group B) achieved the

highest bond strength at the coronal level (11.2±9.9 MPa), whereas the

Excite-Variolink II combination (Group A) exhibited progressively weaker

adhesion from the apical to the middle, and to the coronal third of the root

(15.7±11.5 MPa, 11.6±12.0 MPa, and 9.5±9.2 MPa respectively). However,

none of these differences was statistically significant (Kruskal-Wallis Non-

Parametric Analysis of Variance, p>0.05). Inconsistent bond strengths were

observed from specimens that were derived from the same root level. The

Kolmogorov-Smirnov test revealed a non normal distribution of the data from

all the subsgroups, except those of RelyX Unicem at the coronal level and of

98

Excite-Variolink II at the apical level. The findings did not change much after

excluding or replacing the zero bond strength values.

Microtensile bond strength test, non-trimming technique

Almost all of the specimens prepared through this method failed prematurely

during the cutting phase (Fig. 4b). Only five beams, three from Group A and

two from Group B, survived the preparation procedure (Table II). The

number of obtained specimens was so low that the application of any

statistical parameter or test to the collected bond strength data would be

meaningless. Therefore, the specimens were not loaded, but used only for

microscopic observations (Table II, Fig. 4).

Push-out technique

None of the prepared specimens failed prematurely (Table II). The

measured bond strength values were normally distributed according to the

Kolmogorov-Smirnov test, and their variability was within acceptable limits

(Fig. 7). The bond strength achieved by Excite-VariolinkII (6.89±3.77 MPa)

was significantly higher than that of RelyX Unicem (5.01±2.63 MPa, p<0.05,

Fig. 7). Fig. 7. The table reports mean, standard deviation, and coefficient of variation of the bond strengths measured in MPa with the push-out technique. In the chart, the length of each box represents the interquartile range of bond strength data for the tested materials.

Groups Mean and standard deviation Coefficient of variation

Excite DSC+VariolinkII 6.89±3.77 0.54 RelyX Unicem 5.01±2.63 0.52

MATERIAL

UnicemVariolink

BON

DST

RE

20

10

0

-10

99

When the data were pooled together for root level (Fig. 8), it appeared that

with Excite-VariolinkII the bond strength was higher at the coronal

(7.76±4.17 MPa) than at the middle-apical third (5.65±2.76 MPa). This

difference was at the threshold of statistical significance (t-test), being the p-

value equal to 0.06. For RelyX Unicem the bond strength was statistically

similar at the coronal and the middle third of the root (5.18±2.94 MPa and

4.78±2.26 MPa respectively, t-test, p>0.05). Fig. 8. The table shows mean and standard deviation of the push-out bond strength values in MPa, pooled together for root level. In the chart, the length of each box represents the interquartile range of the bond strengths measured at two different root levels for each of the materials on trial. The differences in bond strength between coronal and middle-apical levels were not statistically significant for RelyX Unicem (p>0.05), and at the threshold of statistical significance for Excite-Variolink II (p=0.06).

Mean and standard deviation Groups Coronal Middle-apical

Excite DSC+VariolinkII 5.82±3.25 5.55±2.66 RelyX Unicem 5.37±2.95 4.60±2.26

When comparing the levels of adhesion achieved at the coronal root level by

the two systems on trial, it appear that the bonding ability of Excite DSC-

Variolink II was significantly greater than that of RelyX Unicem (t-test,

p<0.05). In the middle-apical section of the roots, on ther other hand, the two

materials performed similarly (p>0.05).

Variolink

ROOTLEV

middle-apicalcoronal

BON

DST

RE

20

10

0

Unicem

ROOTLEV

middle-apicalcoronal

BON

DST

RE

16

14

12

10

8

6

4

2

0

-2

100

Discussion The most remarkable findings regarding the data collected with the trimming

variant of the microtensile technique were the extremely high number of

premature failures and the large spread of the data distribution. The high

variability affected the data set even after eliminating or replacing the zero

bonds.

During bur-trimming of the hourglass shaped specimens vibrations are

transmitted to the interfaces, which are stressed in an uncontrolled way.22

Also, it is practically very difficult to limit the contact of bur tip on the post

surface to one point, so as to just divide the bonded interface into two

halves, without exerting any disrupting action on the cement-post-dentin

interface.

To the uncontrolled stress applied by the bur, particularly if trimming is done

free-hand, may reasonably be credited the variability seen in bond strength

data collected with the trimming technique. As the standard deviations were

almost as high (Group A) or even higher (Group B) than the mean values,

one should question the reliability or reproducibility of such a bond testing

method.

Previous studies have shown that the “non-trimming” version of the

microtensile technique may be less traumatizing to the bonding interfaces,

suggesting that the “non-trimming” technique may be more useful in

evaluating interfaces with low bond strengths.2,17,22 In this study, we had

adopted this concept to measure the bond strength of luted fiber posts.

Contrary to the expectations, the non-trimming technique was unsuccessful

in providing intact specimens for bond testing. Most of the beams failed

while being cut, suggesting that the root-post bond strength was too low to

resist the stresses transmitted to the interfaces, particularly when sectioning

at the post periphery.

This procedure for obtaining microtensile beams from posted roots was

therefore extremely critical from a technical standpoint, and produced an

unacceptably high number of useless specimens. In addition, the

101

applicability of the technique solely to large and straight roots inevitably

limits its use only to upper centrals and canines.

The “non-trimming” version of microtensile bond test has previously been

applied to measure the bond strength of adhesive materials to root

dentin.1,6,18,19 However, in these studies, no post was inserted inside the root,

and the bond between resin cement and dentin was assessed just on one

side of the root, as the other one was ground in order to gain unrestricted

access to the canal walls during bonding procedures. Although this setting is

ideal from an experimental point of view, it is not truly representative of the

clinical situation.

It can therefore realistically be concluded that, as a result of the intervention

of factors such as difficult handling of the material, heterogeneity of the

substrate, adverse cavity configuration, the levels of bond strength

achievable in the clinical setting when a fiber post is adhesively luted to root

canal walls are in fact very low, as already shown in previous studies.2,12,20

Using the trimming version of the microtensile bond test, Bouillaguet et al.2

examined the bond strength of posts that were custom-made for each root

from Z100 composite resin (3M ESPE), and luted with different resin

cements. In their investigation, when the posts were cemented into intact

roots as done clinically, the levels of bond strength achieved were low for all

of the tested adhesive cements. In addition, substantial premature failures

were reported, and the data variability was large.

We were faced with the same limitations with the data collected from

hourglass-shaped specimens in the present study. Thus we opined that

even the trimming version of the microtensile bond test does not provide a

reliable comparison between the adhesive properties of the Excite DSC-

Variolink II combination versus the self-adhesive resin cement RelyX

Unicem.

On the contrary, the results of this study showed that each prepared

specimen in the push-out test provided a useful measurement, and the data

variability was limited. This test also appeared to be able to realistically

record low levels of bond strength for both of the materials used to lute fiber

102

posts. Taking into account the relative weakness of the post-root bond, the

push-out test seems to be the most accurate and reliable technique to

measure the bonds of fiber posts to root dentin.

Based on the experience with the microtensile method,2 the push-out

technique employed in this study was designed as a sort of “micro-push-out”

test, reducing the specimens size for the benefit of a more uniform stress

distribution. The objection of a highly non-uniform stress distribution raised

by some authors against the push-out method15 is indeed justified

considering the way the test has so far been carried out, loading thick

specimens12,13 or the whole post.14 This limitation of the original push-out

technique was overcome in the present investigation by slicing the posted

root into 1mm-thick specimens. This modified technique also enabled us to

test for regional differences in bond strength inside the root canals.

Further light on the micro-push-out test could be shed by a microscopic

evaluation of the pattern of bond failure in loaded specimens, as well as by

the assessment of stress distribution during testing through the method of

finite element analysis. A study is currently being performed with the aim of

providing also this information on the test.

As for the comparison between the tested materials, the bond strength

measurements from both push-out and microtensile trimmed specimens

gave the same indication that the adhesion achieved after phosphoric acid

etching and the application of Excite DSC and Variolink II is stronger than

that established by the self-adhesive resin cement RelyX Unicem.

The difference was not statistically significant according to the data collected

through the microtensile. On the other hand, the difference in the push-out

bond strength between Excite-Variolink II and RelyX Unicem was statistically

significant.

A possible explanation for the finding of a weaker adhesion with RelyX

Unicem could be that the methacrylated phosphoric esters, responsible for

substrate conditioning in this adhesive, are not as effective as phosphoric

acid at dissolving the thick smear layer created on canalar walls during the

post space preparation. Microscopic observations of the adhesive cement-

103

root dentin interfaces developed with Variolink II and RelyX Unicem are

being performed in order to verify this supposition.

Factors possibly interfering with the development of high bond strengths to

root dentin are the non-uniform adaptation of the bonding material or its

incomplete polymerization, both related to the difficult access of root canal

walls during handling. These factors may account for the lower bond

strengths achieved by the adhesive cements in the middle-apical root

sections (Fig. 8). In particular, as regards RelyX Unicem, the material might

have failed to uniformly reach the deepest portion of the root as a result of

being applied only onto the post surface.

In addition, the C-factor of a root canal is highly unfavourable and

contributes to maximizing the polymerization stress of resin based materials

along the root canal walls. Morris et al.1 have figured that C-factors in root

canals can range from 20 to 100, depending on the diameter and length of

the canal. According to Bouillaguet et al.2, the C-factor when a 150µm-thick

layer of resin cement is used to lute an endodontic post may even exceed

200. Especially with light-curing materials, the curing stress generated in the

adverse geometrical configuration of the root canal may be so intense that

the resin composites detach from the dentin walls, thus creating interfacial

gaps.

In a recent microtensile test, Ari et al.19 report that Variolink II in combination

with Clearfil Liner Bond 2V measured significantly lower bond strength to

root canal dentin than the self-cure slow-setting C&B Metabond. There is

therefore the indication that the adhesion to canal walls would benefit from a

better control of the resin cement polymerization rate, such as to allow most

of the curing stress to be absorbed by flow.2,19

The mentioned study by Ari et al.19 is the only one providing data on the

bond strength to radicular dentin of Variolink II. No similar research data are

yet available for RelyX Unicem, more recently introduced on the market.

Within its limitations, the present study intended to add a piece of

information to the still largely unexplored field of bonding to root canals. The

104

focus was in particular on assessing feasibility and reliability of the different

adhesion testing techniques that can be applied to radicular dentin.

In the light of this study’s findings it can be concluded that, when measuring

the bond strength of fiber posts adhesively luted to root canal dentin, the

push-out test appears to be more efficient and dependable than both the

trimming and non-trimming versions of the microtensile technique. In the

present investigation, the push-out technique was effectively applied to

compare the strength of the adhesion of fiber posts luted with a new self-

adhesive resin cement and a dual-cure cement that was used in conjunction

with phosphoric acid-etching and a dentin adhesive. AcknowledgementsThe authors are thankful to Mr. Paulo Santos and NAPEM for their help

with this project.

105

References 1. Morris MD, Kwang-Won Lee, Agee KA, Bouillaguet S, Pashley DH.

Effects of sodium hypochlorite and RC-prep on bond strengths of resin

cement to endodontic surfaces. J Endod 2001; 27: 753-757.

2. Bouillaguet S, Troesch S, Wataha JC, Krejci I, Meyer JM, Pashley DH.

Microtensile bond strength between adhesive cements and root canal

dentin. Dent Mat 2003; 19: 199-205.

3. Ferrari M, Mannocci F, Vichi A, Cagidiaco MC, Mjör IA. Bonding to root

canal: structural characteristics of the substrate. Am J Dent 2000; 13: 255-

260.

4. Ferrari M, Vichi A, Grandini S. Efficacy of different adhesive techniques

on bonding to root canal walls: an SEM investigation. Dent Mater 2001; 17:

422-429.

5. Nikaido T, Takano Y, Sasafuchi Y, Burrow MF, Tagami J. Bond

strengths to endodontically treated teeth. Am J Dent 1999; 12: 177-180.

6. Drummond JL. In vitro evaluation of endodontic posts. Am J Dent 2000;

13: 5B-8B.

7. Mitchell CA, Orr JF, Connor KN, Magill JPG, Maguire GR. Comparative

study of four glass ionomer luting cements during post pull-out tests. Dent

Mater 1994; 10: 88-89.

8. Qualthrough AJ, Chandler NP, Purton DG. A comparison of the retention

of tooth-colored posts. Quint Int 2003; 34: 199-201.

9. Garcia Varela S, Bravos Rabade L, Rivas Lombardero P, Linares Sixto

J, Gonzalez Bahillo J, Ahn Park S. In vitro study of endodontic post

cementation protocols that use resin cements. J Prosthet Dent 2003; 89:

146-153.

10. Purton DG, Love RM. Rigidity and retention of carbon fibre versus

stainless steel root canal post. Int Endod J 1996; 29: 262-265.

11. Prisco D, De Santis R, Mollica F, Ambrosio L, Rengo S, Nicolais L. Fiber

post adhesion to resin luting cements in the restoration of endodontically-

treated teeth. Oper Dent 2003; 28: 515-521.

106

12. Patierno JM, Rueggeberg FA, Anderson RW, Weller RN, Pashley DH.

Push-out and SEM evaluation of resin composite bonded to internal cervical

dentin. Endod Dent Traumatol 1996; 12: 227-236.

13. Boschian Pest L, Cavalli G, Bertani P, Gagliani M. Adhesive post-

endodontic restoration with fiber posts: push-out tests and SEM

observations. Dent Mater 2002; 18: 596-602.

14. Gallo JR, Miller T, Xu X, Burgess JO. In vitro evaluation of the retention

of composite fiber and stainless steel posts. J Prosthodont 2002; 11: 25-29.

15. Sudsangiam S, van Noort R. Do dentin bond strength tests serve a

useful purpose? J Adhes Dent 1999; 1: 57-67.

16. Ngoh EC, Pashley DH, Loushine RJ, Weller N, Kimbrough F. Effects of

eugenol on resin bond strengths to root canal dentin. J Endod 2001; 27:

411-414.

17. Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshiyama M, Shono

Y, Fernandes CA, Tay F. The microtensile bond test: A review. J Adhes Dent

1999; 1: 299-309.

18. Gaston BA, West LA, Liewehr FR, Fernandes C, Pashley DH.

Evaluation of regional bond strength of resin cement to endodontic surfaces.

J Endod 2001; 27: 321-324.

19. Ari H, Yasar E, Belli S. Effects of NaOCl on bond strength of resin

cement to root canal dentin. J Endod 2003; 29: 248-251.

20. Mannocci F, Sherriff M, Ferrari M, Watson TF. Microtensile bond

strength and confocal microscopy of dental adhesives bonded to root canal

dentin. Am J Dent 2001; 14: 200-204.

21. Goracci C, Ferrari M, Grandini S, Monticelli F, Tay FR. Bonding of a self-

adhesive resin cement to dental hard tissues. J Adhes Dent (in press).

22. Goracci C, Sadek FT, Monticelli F, Cardoso PEC, Ferrari M. Influence of

substrate, shape, and thickness on microtensile specimens’ structural

integrity and their measured bond strength. Dent Mater 2004; 20: 643-654.

107

CHAPTER III: MEASURING THE MICROTENSILE BOND STRENGTH OF

MATERIALS TO NON-DENTAL SUBSTRATES

III.1 The adhesion between prefabricated FRC posts and composite resin cores: microtensile bond strength with and without post

silanization. Goracci C, Raffaelli O, Monticelli F, Balleri B, Bertelli E, Ferrari M. Dental

Materials 2004, in press.

Introduction Prefabricated FRC posts are today diffusely accepted as a viable alternative

to cast posts for the restoration of endodontically treated teeth.1

Prefabricated FRC posts are adhesively luted inside root canals and, in case

the loss of coronal structure is substantial, they provide retention to a core

portion2, which is directly built up onto the post with a composite resin.

For the core build-up procedure a large variety of composite resin materials

are available to the clinician, from packable to microhybrid to flowable

composites.3,4

Some laboratory and clinical evidence has recently been collected that

supports the use of flowable materials for building up the core portion in

prefabricated FRC post restorations.3 When the structural integrity of the

core material and its adaptation onto the post surface were evaluated

through SEM observations of post-and-core units prepared in vitro, flowable

composites exhibited better results than hybrid composites and composites

marketed as core materials.3 The use of flowables as core materials has

also been validated by the findings of an in vivo trial, where post-and-core

restorations thus performed gave proof of a satisfactory clinical service over

a two-year follow-up period.1

In addition to the microscopic appearance of the prefabricated FRC post-

composite core interface, also the strength of the adhesion between the two

materials should be assessed. In particular, it should be verified whether the

108

bond between prefabricated FRC posts and composite resin cores can be

significantly enhanced by silanization of the post surface.

Silane coupling agents have been diffusely utilized in dentistry since the

advent of glass-reinforced and resin-based materials.

In FRC post technology glass or quartz fibers are coated with a silane in

order to improve the adhesion at the fiber-resin matrix interface, protect

fibers from damage during handling, modify the catalytic and wettability

properties of fiber surfaces5, and increase the chemical resistance of the

fiber-matrix interface especially to water.6

Fiber pull-out tests have been performed to measure the interfacial bond

strength at the fiber-matrix interface.7 On the other hand, limited information

is available as regards the potential effect of silane treatment on the

adhesion between the fiber post as a whole and the resin-based material

with which the post is to interface, i.e. luting agents and core materials. Only

one study has lately been published on the effect of different post surface

treatments on the bond strength to resin cements.8 Conversely, data are still

missing regarding the influence of post silanization on the bond strength to

core materials.

This study was therefore conducted with the aim of measuring, through the

microtensile non-trimming technique, the bond strength between two types

of translucent prefabricated FRC posts and two types of flowable composites

used as core materials, with and without silanating the post surface prior to

building up the core.

The formulated null hypothesis was that the bond strengths achieved at the

post-core interface with the various combinations of post material, core

material, and surface treatment tested were not significantly different.

Materials and Methods Twenty-eight FRC Postec size 3 with a maximum diameter of 2 mm (Ivoclar-

Vivadent, Schaan, Liechtenstein; Group 1), and twelve Light-Post size 2 with

a 1.8 mm diameter (RTD, St. Egève, France; Group 2) were used for testing.

109

FRC Postec posts feature unidirectional glass fibers (61.5% weight),

embedded in a polymer matrix of triethylene-glycol-dimethacrylates

(TEGDMA) and urethane-dimethacrylates (UDMA) monomers, in

combination with highly dispersed silicon dioxide.

DT Light-Posts are made of unidirectional pre-tensed quartz fibers (60%

volume), bound in an epoxy resin matrix.

On half of the posts from each group the surface was treated with a silane

coupling agent (Monobond-S, Ivoclar-Vivadent, Schaan, Liechtenstein).

Monobond-S contains 3-methacryl-oxypropyltrimethoxysilane (MPS) as the

effective silane (1% in weight), in a solution of ethanol (52% in weight) and

distilled water (47% in weight), and has a ph of 4.9 Following manufacturers

instructions, Monobond-S was applied on the post surface with a brush.

After having allowed for a 60 second contact at room temperature, the post

surface was dried with air, and the core portion was built up onto the post

with a flowable composite resin. UnifilFlow (GC, Tokyo, Japan) and Tetric

Flow (Ivoclar-Vivadent, Schaan, Liechtenstein) were used for core build-ups.

The following experimental groups were thus formed:

Group 1.1: FRC Postec posts and UnifilFlow (n=7);

Group 1.2: Silanated FRC Postec posts and UnifilFlow (n=7);

Group 1.3: FRC Postec posts and Tetric Flow (n=7);

Group 1.4: Silanated FRC Postec posts and Tetric Flow (n=7);

Group 2.1: Light-Post posts and UnifilFlow (n=3);

Group 2.2: Silanated Light-Post posts and UnifilFlow (n=3);

Group 2.3: Light-Post posts and Tetric Flow (n=3);

Group 2.4: Silanated Light-Post posts and Tetric Flow (n=3).

For the core build-up procedure, each post was positioned upright on a glass

slab, and secured with a drop of sticky wax. Then, a cylindrical plastic matrix

was placed around the post and adjusted so that the post would be exactly

in the middle. The matrix was 10 mm in diameter. In height, the matrix was

extended only to the cylindrical portion of the post (about 10 mm in Light-

Post and 5 mm in FRC Postec posts), since, for an appropriate cutting of the

microtensile specimens, it is desirable that the post diameter be constant

110

throughout the post length. The flowable composite was applied onto the

post directly from the syringe in 1-2 mm thick increments, which were

carefully adapted on the post surface and singularly cured for 20 seconds

with a halogen curing light (Optilux 401, Kerr/Demetron, intensity 750

mW/cm2). The material was always irradiated directly from the open upper

side of the matrix and through the post. Never was irradiation done through

the plastic matrix.

When the matrix was completely filled, the cylinder was taken off the glass

slab, and a further 20-second irradiation was done on the side of the cylinder

that had faced the glass slab, in order to ensure complete polymerization of

the core material. At this point the plastic matrix was cut off, and the cylinder

of composite resin built up around the post was separated (Fig. 1a).

The procedure of specimens cutting and loading was started immediately, as

the intention was to quantify the bond strength reached by the materials

around the time when clinically the procedures of core preparation,

impression, and temporary crown adaptation and cementation are

performed. It is in fact during these phases that the bond between post and

core material is first stressed by vibrations, tensile and shear forces.

For cutting, each cylinder of material was secured on the holding device of

an Isomet machine (Buehler, Lake Bluff, IL, USA; Fig. 1a). Then, by means

of a water-cooled diamond blade, two longitudinal cuts were performed on

two opposite sides of the post at its outermost periphery (Fig. 1b), so as to

expose the post surface throughout its length. As a result, a slab of uniform

thickness was created, that presented with the post in the center and the

core build-up on each side (Fig. 1b). From the slab, 1-mm thick beams were

then serially sectioned (Fig. 1c). Thirty to thirty-five sticks were obtained in

each group. Every stick was glued with cyanoacrylate (Zapit, Dental

Ventures of America, CA, USA) to the two free sliding components of a jig,

which was mounted on a universal testing machine (Controls, Milano, Italy).

This set-up was conceived to apply purely tensile forces to the two opposite

post-core interfaces.

111

Fig. 1 Procedure for the preparation of microtensile specimens. (a) A cylinder of core material was built up onto the glass fiber post by progressively adding small increments of composite resin. With two longitudinal cuts running at the periphery of the post, the post surface was exposed throughout its length. A slab of with the post in the center and the core material on the sides was thus created (b). From the slab, 1-mm thick beams were serially sectioned (c; s=stick).

a

b

c

Each specimen was loaded at a cross-head speed of 0.5 mm/min until

failure occurred at either one of the two stressed interfaces. Bond strength

Core material

Core material Post

S S

112

was expressed in MegaPascals (MPa), by dividing the load at failure

(Newtons) with the bonding surface area (mm2). As the bonded interface

was curved, its area was calculated using a mathematical formula previously

applied by Bouillaguet et al.10 for similar purposes.

Each failed specimen was observed with an optical microscope at 20

magnifications (Bausch&Lomb, Rochester, NY, USA), in order to classify the

type of failure as adhesive between post and core, cohesive within post, or

cohesive within core.

Statistical Analysis

The Two-Way Analysis of Variance was applied with microtensile bond

strength in MPa as dependent variable and type of post, core material, and

silanization procedure as factors. The level of significance was set at p=0.05.

The statistical analysis was processed by the SPSS 11.0 software (SPSS

Inc., Chicago, IL, USA). Results

The mean and standard deviation of the post-core microtensile bond

strength values for the eight experimental Groups are reported in Table I.

The statistical analysis revealed that neither the type of post, nor the core

material, or the interactions among the factors had a significant influence on

bond strength at the post-core interface (p>0.05). Only the post silanization

procedure had a significant effect (p<0.05). In other words, regardless of the

type of post or core material used, the adhesion at the interface was

significantly enhanced by post surface treatment with a silane coupling agent

(Table I, Graph 1).

As regards the type of failure, in none of the loaded specimens fractures

developed within the post or the core portion. Failures always occurred

adhesively at the post-core interface.

113

Table I. Mean and standard deviation values of the microtensile bond strengths measured in all the experimental groups. The graph shows that with any combination of post and core material, post silanization increased the interfacial bond strength (FRC = FRC Postec Posts; UF = UnifilFlow; S = silane; TF = Tetric Flow; LP = Light-Post).

Type of post Core material FRC Postec Light-Post

No Silane Group 1.1 Group 2.1 Mean (sd) 9.05 (5.69) 8.36 (5.46)

Silane Group 1.2 Group 2.2 UniFil Flow

Mean (sd) 11.11 (2.49) 12.22 (2.57) No Silane Group 1.3 Group 2.3 Mean (sd) 10.74 (5.65) 7.87 (2.65)

Silane Group 1.4 Group 2.4 Tetric Flow

Mean (sd) 12.88 (3.16) 13.43 (3.05)

GROUPS

LP+S+TF

LP+TF

LP+S+UF

LP+UF

FRC+S+TF

FRC+TF

FRC+S+UF

FRC+UF

30

20

10

0

-10

no silane

silane

114

Graph 1. The graph reports the mean and standard deviation (in parentheses) values calculated when the bond strengths of all the silanated and non-silanated posts were pooled together, regardless of the post and the core material used. Significantly higher bond strengths were measured after silanization (p<0.05), and the variability of the data was lower than that observed with non-silanated posts.

SILANIZA

silaneno silane

MTB

S

30

20

10

0

-10

Discussion Adhesive posts restorations rely for their retention on the strength of the

bonds established at different interfaces. Among them, the interface

between root dentin and resin cement has been the object of several

studies, involving both bond strength tests,10-15 and microscopic

investigations.16-20

Since the introduction of prefabricated FRC posts, a continuous effort has

been made to improve the bonding potential of current adhesive systems

inside root canals,11,12,14,18-20 as radicular dentin has been shown to offer far

less favourable conditions for adhesion than coronal dentin and enamel.10

As a matter of fact, the findings of clinical trials indicate that in case of

debonding of prefabricated FRC post restorations, an adhesive failure at the

cement-dentin interface is most ot the times involved.1

Although it is then the root-cement bond to represent the weakest link,

however, also the post-cement and post-core interfaces deserve attention.

8.92 (4.79) MPa 12.52 (2.87) MPa

115

In particular, it is from the strength of the chemical and micromechanical

interaction between a fiber-reinforced material and a composite resin that

depends the retention of the core portion onto the post. This bond has to

rapidly reach levels of strength sufficiently high to resist the stress

transmitted during core trimming and adaptation of the provisional crown.

In this investigation the adhesion of two flowable composites to two types of

prefabricated FRC posts were assessed with the microtensile technique at

completion of the core build-up procedure. The most relevant finding of the

study was that with any tested combination of post and core material, the

interfacial bond strength was significantly enhanced if the post surface had

been preliminarily coated with a silane coupling agent. The increase in bond

strength as a consequence of post silanization was more remarkable with

Light-Post than with FRC Postec posts (Table I).

On the other hand, since the statistical analysis revealed the post and core

materials interaction to be non significant, it can be inferred that using a

prefabricated FRC post and a composite resin from the same manufacturer

does not result in a significantly enhanced interfacial bond strength. As a

matter of fact, the adhesion achieved by Tetric Flow on FRC Postec posts

was not significantly stronger than that established by UnifilFlow on the

same type of posts (Table I). According to recently published data [8],

treating the post surface with a silane coupling agent is advisable also to

enhance the adhesion of the resin cement used for luting. Beside

silanization, also post sandblasting and the combination of this with silane

coating were found to significantly increase the bond strength of resin

cements to glass fiber posts.8 It may then be of interest to verify whether and

to what extent also the adhesive potential of the composite core can be

improved with these procedures of post surface treatment.

In this study the silane agent was used at room temperature. However, it

may be worth testing whether heat treatment of the silanized post further

adds to the FRC post-core bond strength, similarly to what has been shown

for the porcelain-composite bond.21

116

Different theories have been proposed in order to explain the bonding

mechanism through silane coupling agents. According to the oldest one, the

chemical bonding theory, the coupling action of the silane involves the

formation of covalent bonds from the reaction of the organo-functional group

(R) and the hydrolysed alkoxy groups (R’O)3 respectively with the resin

matrix and the mineral substrate (glass or silica) of the composite material.22

Today the reversible hydrolytic bond mechanism theory is more widely

accepted as, though keeping some concepts of the chemical bond theory, it

provides a better explanation for the hydrolytic stability of bonding through

silanes. The theory states that the bonds between silane and mineral

substrate are reversibly broken and remade in the presence of water,

allowing for stress relaxation without loss of adhesion.22

As the silane agent is only able to chemically bridge resins and OH-covered

inorganic substrates, at the fiber post-composite core interface the chemical

bond is possible only between the resin of the core material and the exposed

fibers of the post. On the other hand, the highly cross-linked polymers of the

matrix in FRC posts do not have any functional group available for reaction.

In FRC materials such as everStick (StickTech, Turku, Finland), an attempt

to solve the problem of adhering to highly cross-linked polymers has been

made by using semi-Interpenetrating Polymer Network structures.23 In this

technology, the fibers are preimpregnated with a polymethylmetahcrylate

(PMMA), that can be partially dissolved with the application of a light-curing

resin for five minutes. As a result of the partial dissolution at the surface of

the fiber frame, grooves and undercuts are created where micromechanical

bonding can be established in addition to the chemical adhesion. According

to the manufacturers the post surface is thereby “reactivated” to offer

considerably more favourable conditions for adhesion to the core or the

luting material.24

The prefabricated FRC posts tested in this study do not contain semi-IPN

structures. Since the contribution of the chemical bond in coupling post and

core materials through silanes can be expected to be low, the mechanism

most likely involved the enhancement of the post-core bond seen in the

117

present study can be identified in the improvement in post surface wettability

following silane coating.

The surface wetting theory recognizes a key role to the wetting capacity of

the silane for improved adhesion. According to this theory, the silane, thanks

to its low viscosity, would assist substrate wetting, and once an intimate

contact between the interfacing materials is established, also van der Waals’

forces would become effective, providing a physical adhesion, which adds

up to the chemical reactions.6

Although the results of this study provide clear evidence that silane coating

of the post surface increases the post-core bond strength, some uncertainty

remains around the mechanism actually responsible for the enhancing

effect. To a similar conclusion regarding in general the use of silane in

dentistry have come Matinlinna et al. in their recent extensive literature

review on dental silanes: although the majority of clinical results indicate a

significant role of silanes in the adhesion process, however the silane

reaction mechanisms still remain not fully understood.9

To testify the action of the silane as a post-core adhesion promoter in this

microtensile study was not only the recording of higher values of bond

strength recorded after silanization, but also the finding of no premature

specimen failures with silanated posts. On the other hand, few specimens

from untreated posts failed during cutting or gluing, and were taken into

account in the statistical calculations as zero values. The inclusion of “zero

bonds” may have contributed to the greater spread of the bond strengths of

non-silanated posts. Also the finding of more limited standard deviations in

Groups 1.2, 1.4, 2.2, 2.4 may be interpreted as the indication that with post

silanization a more uniform bond with the core material is developed. Silanes

coupling agents are claimed to exert other favourable actions in interfacial

adhesion. They are believed to increase the resistance of the bonds to

chemical dissolution, particularly from water.6 In addition, due to their elastic

properties, silanes would be able to absorb the stress that may develop at

the interface as a result of differences in thermal expansion coefficients of

the interfacing materials.22 Unfortunately, these supposed effects of

118

silanization could not be assessed in this investigation. As the microtensile

bond strength tests were performed right after specimen preparation, no

inference can be made from this data regarding the durability of the bonds.

In order to investigate this aspect, a fatigue test, possibly involving

thermocycling, could be performed on post and core units prepared in vitro,

and a further validation of the results could be provided by a longitudinal

clinical trial of adhesive post and core restorations performed with and

without post silanization.

As regards the materials on trial, although the manufacturers of FRC Postec

and Light-Post posts recommend to use respectively a microhybrid

composite, Tetric Ceram, and a core material, Lumiglass, to build up the

abutment portion, however in this study it was decided to test two flowable

composites as core materials. This choice was suggested by the findings of

a previous microscopic investigation, where the abutments prepared with

Tetric Flow and UnifilFlow exhibited the highest structural homogeneity and

the best adaptation on the post.3

It should be pointed out that on carbon FRC posts silanization would not be

so effective at enhancing the post-core bond as it is on glass FRC posts.

Since carbon fibers do not have a significant number of hydroxyl groups on

the surface, their reaction with silanes is improbable.6

The microtensile technique was adopted in this trial as it is currently

regarded as the most reliable method for bond strength testing.25 The small

size of the specimens is condition for a more uniform distribution of the

stress on loading, which limits the chance of cohesive failures, thus allowing

for an accurate assessment of the interfacial bond strength.25,26

In particular, the non-trimming variant of the technique was chosen as there

are indications in the literature that it is less aggressive than the variant

which involves trimming the specimen to an hourglass shape at the bonded

interface.25,27 As a matter of fact in this trial very few specimens broke

prematurely, whereas in a previous attempt to apply the trimming method to

measure the bond strength between post and core materials, a great

119

number of premature failures were recorded, that considerably elevated the

standard deviation of the data set.28

Alternatively, the push-out technique could be applied to measure the post-

core bond strength, provided that, as done in microtensile, the specimen

size is kept small for improved stress distribution,27 thus performing what

could be called a “micro-push-out test”. Practically this would involve cutting

the cylinder of post and core build-up into 1 mm thick slices, from which the

post portion is then extruded by means of an appropriately sized loading

punch.

Conclusion The results of this microtensile study support the use of a silane agent as an

adhesion promoter at the interface between translucent FRC posts and

composite resin cores. The exact mechanism by which this enhancing effect

takes place remains not fully understood.

120

References 1. Monticelli F, Grandini S, Goracci C and Ferrari M. Clinical behavior of

translucent fiber posts: A 2-year prospective study. Int J Prosthod, 2003; 16:

593-596.

2. Sorensen JA and Engelman MJ. Ferrule design and fracture resistance

of endodontically treated teeth. J Prosthet Dent, 1990; 63: 529-536.

3. Monticelli F, Goracci C, Grandini S, García-Godoy F and Ferrari M.

Scanning electron microscopic evaluation of fiber post-resin core units built

up with different resin composite materials. Am J Dent (in press).

4. Monticelli F, Goracci C and Ferrari M. Micromorphology of the fiber post-

resin core unit: a scanning electron microscopy evaluation. Dent Mater,

2004; 20: 176-183.

5. Ishida H. Structural gradient in the silane coupling agent layers and its

influence on the mechanical and physical properties of composites. In:

Ishida H, Kumar G, editors. Molecular characterization of composite

interfaces. New York: Plenum Press; 1985: 25-50.

6. Plueddemann EP. Silane coupling agents. New York: Plenum Press;

1991.

7. Debnath S, Wunder SL, MCCool JI and Baran GR. Silane treatment

effects on glass/interfacial shear strengths. Dent Mater, 2003; 19: 441-448.

8. Sahafi A, Peutzfeldt A, Asmussen E and Gotfredsen K. Bond strength of

resin cement to surface-treated posts of titanium alloy, glass fiber, and

zirconia, and to dentin. J Adhes Dent, 2003; 5: 153-162.

9. Matinlinna JP, Lassila LVJ, Özcan M, Yli-Urpo A, Vallittu P. An

introduction to silanes and their clinical applications in dentistry. Int J

Prosthod 2004; 17: 155-164.

10. Bouillaguet S, Troesch S, Wataha JC, Krejci I, Meyer JM and Pashley

DH. Microtensile bond strength between adhesive cements and root canal

dentin. Dent Mater, 2003; 19: 199-205.

11. Morris MD, Kwang-Won Lee, Agee KA, Bouillaguet S and Pashley DH.

Effects of sodium hypochlorite and RC-prep on bond strengths of resin

cement to endodontic surfaces. J Endod, 2001; 27: 753-757.

121

12. Ngoh EC, Pashley DH, Loushine RJ, Weller N and Kimbrough F. Effects

of eugenol on resin bond strengths to root canal dentin. J Endod, 2001; 27:

411-414.

13. Gaston BA, West LA, Liewehr FR, Fernandes C and Pashley DH.

Evaluation of regional bond strength of resin cement to endodontic surfaces.

J Endod, 2001; 27: 321-324.

14. Ari H, Yasar E and Belli S. Effects of NaOCl on bond strength of resin

cement to root canal dentin. J Endod, 2003; 29: 248-251.

15. Prisco D, De Santis R, Mollica F, Ambrosio L, Rengo S and Nicolais L.

Fiber post adhesion to resin luting cements in the restoration of

endodontically treated teeth. Oper Dent, 2003; 28: 515-521.

16. Ferrari M, Mannocci F, Vichi A, Cagidiaco MC and Mjör IA. Bonding to

root canal: structural characteristics of the substrate. Am J Dent, 2000; 13:

255-260.

17. Mannocci F, Sherriff M, Ferrari M and Watson TF. Microtensile bond

strength and confocal microscopy of dental adhesives bonded to root canal

dentin. Am J Dent, 2001; 14: 200-204.

18. Ferrari M, Vichi A and Grandini S. Efficacy of different adhesive

techniques on bonding to root canal walls: an SEM investigation. Dent

Mater, 2001; 17: 422-429.

19. Ferrari M, Grandini S, Simonetti M, Monticelli F and Goracci C. Influence

of a microbrush on bonding fiber posts into root canals under clinical

conditions. Oral Surg, Oral Med, Oral Path, Oral Rad, and Endod, 2002; 94:

627-631.

20. Ferrari M, Vichi A, Grandini S and Goracci C. Efficacy of a self-curing

adhesive-resin cement system on luting glass-fiber posts into root canals: an

SEM investigation. Int J Prosthod, 2001; 14: 543-549.

21. Barghi N, Berry T, Chung K. Effects of timing and heat treatment of

silanated porcelain on the bond strength. J Oral Rehab 2000; 27: 407-412.

22. Pape PG and Plueddemann EP. Methods for improving the performance

of silane coupling agents. J Adhesion Sci Technol, 1991; 5: 831-842.

122

23. Kallio TT, Lastumäki TM, Vallittu PK. Bonding of a restorative and

veneering composite resin to some polymeric composites. Dent Mater, 2001;

17: 80-86.

24. Väkiparta TM, Yli-Urpo A, Vallittu PK. Flexural properties of glass fiber

reinforced composite with multiphase biopolymer matrix. J Mat Sci Mater

Med 2004, 2004; 15: 7-11.

25. Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshiyama M, Shono

Y, Fernandes CA and Tay F. The microtensile bond test: A review. J Adhes

Dent, 1999; 1: 299-309.

26. Sudsangiam S and van Noort R. Do dentin bond strength tests serve a

useful purpose? J Adhes Dent, 1999; 1: 57-67.

27. Goracci C, Sadek FT, Monticelli F, Cardoso PEC and Ferrari M.

Influence of substrate, shape, and thickness on microtensile specimens’

structural integrity and their measured bond strength. Dent Mater (in press).

28. Goracci C, Monticelli F, Tavares AU, Sadek F, Cardoso PEC and Ferrari

M. The adhesion between fiber posts and composite resin cores:

microtensile bond strength of different combinations of materials. J Dent

Res, 2003; 82: B170, Abstract No. 1268.

123

CHAPTER IV: EXPLORING THE APPLICATION OF THE PUSH-OUT TEST AS AN ALTERNATIVE TO MICROTENSILE

IV.1 Evaluation of the adhesion of fiber posts to intraradicular dentin. Goracci C, Sadek FT, Fabianelli A, Tay FR, Ferrari M. Operative Dentistry

2004, in press.

Introduction Fiber posts that are bonded to root canal dentin via resin cements are now

routinely employed for the restoration of endodontically treated teeth. The

similarity in elastic moduli of the fiber post, resin cement, core material, and

dentin1 was perceived to be advantageous in improving the performance of

these restorations, as compared with cast metal post and core

restorations.2,3 Although the occurrence of root fractures, the most frequent

cause of failure with metallic posts, is rare with the use of fiber posts,4 recent

clinical trials indicated that fiber post restorations may fail via the dislodging

of the bonded posts.5-8 Previous bond strength and morphologic studies

have shown that bonding to root canals may be influenced by endodontic

procedures prior to post cementation,9-12 the variability of intraradicular

dentin,13-15 the compatibility of resin cements with dentin adhesives,16-18 and

the cement film thickness.19-21

Coupling of resin-based cements traditionally requires the adjunctive use of

dentin adhesives that are either total-etch or self-etch in nature.22 Total-etch

resin cements utilize phosphoric acid-etching that completely dissolves the

smear layer and creates a zone of partially demineralized dentin. On rinsing

of the acid conditioners, adhesive primers and resins are applied to the

demineralized dentin to achieve micromechanical bonding. Conversely, self-

etch resin cements utilize adhesives containing increased concentrations of

acidic resin monomers to simultaneously demineralize and infiltrate the

smear layer-covered dentin. A further reduction in working steps has been

accomplished with the recent introduction of a self-adhesive resin cement

(RelyX Unicem, 3M ESPE, St. Paul, MN, USA) that does not require any

124

pre-treatment of tooth substrates.23 Apart from the marketing data supplied

by the manufacturer, little information is available on the efficacy of the new

self-adhesive resin cement for luting of fiber posts to intraradicular dentin.

Bonding to intraradicular dentin presents challenges to clinicians due to the

radically different bonding conditions that are present.13 For example, the

highly unfavorable cavity configuration factors within the dowel spaces

warrant the use of self-cured, slow-setting resin cements for more effective

dissipation of polymerization shrinkage stresses.24 Although microtensile

bond strength studies are available on the use of resin cements in root

canals, the dowel spaces in these studies were filled completely with resin

cements alone,11,25 ground flat in order to gain unrestricted access to canal

walls during the bonding procedure,26,27 or bonded with custom-made posts

that were fabricated out of pre-polymerized hybrid resin composites.24 While

these studies provide significant contributions to the adhesive mechanisms

of different types of resin cements to root dentin, they are more akin to

simplified modeling approaches and do not truly reflect the dislocation

resistance of bonded posts in intact dowel spaces that involve multiple

interfaces. The reliability of the microtensile test in assessing the bonding of

fiber posts to intact dowel spaces was challenged in a recent study,28 with

the frequent observation of premature bond failures when specimens were

prepared for bond testing using either the trimming or non-trimming

technique. The propensity of these premature bond failures was thought to

be caused by the presence of pre-existing interfacial gaps and/or the

superimposition of in-service stresses that were generated during specimen

preparation for microtensile bond testing, upon the potentially destructive

macroscopic (Type I) residual stress fields29 that were present in the bonded

canals.

Similar to the retention of implant attachments in bone,30 the resistance to

dislocation of fiber posts bonded to intact root canals with resin- or glass-

ionomer-based cements may be considered as a net sum of

micromechanical interlocking, chemical bonding and sliding friction.31 For

this reason, pull-out and push-out test results have been more successfully

125

employed as indicators of the interfacial strengths of bonded fiber posts in

root canals.1,32,33 Similar to the microtensile bond test, an additional

advantage with the use of the “thin slice” push-out test34,35 is that multiple

specimens may be retrieved from a bonded root canal.

This study utilized a “thin-slice” push-out test and transmission electron

microscopy (TEM) to examine the interfacial strength and ultrastructure of a

total-etch, a self-etch, and a self-adhesive resin cement that were employed

for bonding of glass fiber posts to intact root canals. The null hypothesis

tested was that there is no difference in the efficacy of the three resin

cements for bonding of glass fiber posts to intraradicular dentin.

Materials and Methods Twenty-seven single-rooted teeth that had been extracted for periodontal

reasons were used in the study. The crown portion of each tooth was

removed with a water-cooled diamond saw (Isomet, Buehler, Lake Bluff, IL,

USA), and the root was endodontically instrumented at a working length of 1

mm from the apex to a #35 master apical file. A step-back preparation

technique was followed, using stainless-steel K-files (Union Broach, New

York, NY, USA), Gates-Glidden drills #2 to #4 (Union Broach), and 2.5%

sodium hypochlorite irrigation. For canal obturation, thermoplasticized,

injectable gutta-percha (Obtura, Texceed Corp., Costa Mesa, CA, USA) and

a resin sealer (AH-26, DeTrey, Zurich, Switzerland) were employed. In each

root-treated tooth, a 9-mm deep dowel space was prepared with low-speed

drills provided by the post manufacturer, and a 1.3 mm diameter translucent

glass fiber post (FRC Postec, Ivoclar-Vivadent, Schaan, Liechtenstein) was

tried-in and cut to the adequate length. After sectioning to the appropriate

length, the glass fiber post was cleaned in ethanol, silanized with Monobond-

S (Ivoclar-Vivadent), and left to air-dry before coating with the adhesive or

resin cement.

The teeth were randomly divided into three groups of nine specimens each.

For each group, a different resin-based luting agent was utilized for fiber

post cementation. The tested materials were Excite DSC/Variolink II (Ivoclar-

126

Vivadent, Schaan, Liechtestein, Group I), ED Primer/Panavia 21 (Kuraray

Medical Inc. Japan, Group II), and RelyX Unicem (Group III). Variolink II is a

dual-cured cement that requires phosphoric acid for substrate conditioning

and the application of the self-activated Excite DSC as a coupling dentin

adhesive. Panavia 21 is an auto-cured resin cement that is used in

combination with the proprietary one-step self-etching primer (ED primer).

RelyX Unicem is a dual-cured cement that is claimed by the manufacturer to

be self-adhesive in nature and does not require pre-treatment of the tooth

substrates. The components of the three resin cements are shown in Table

I. Table I. Chemical composition of the resin cements tested in the study

Group I Group II Group III

RelyX Unicem Excite DSC HEMA, TEGDMA, phosphoric acid acrylate, silicon

dioxide, initiators, stabilizers, alcohol

ED Primer HEMA, MDP, 5-NMSA,

sodium benzene sulfinate N, N-diethanol

p-toluidine, water

VariolinkII Dimethacrylates,

silicon dioxide, self-cure initiators, light-

cure initiators

Panavia 21 Glass and silica powder,

sodium fluoride bis-phenol A polyethoxy

dimethacylate, 10 MDP, hydrophilic and

hydrophobic dimethacylates, self-cure

initiators

Powder Glass fillers,

silica, calcium hydroxide, self-cure initiators,

pigments

Liquid Methacrylated

phosphoric esters, dimethacrylates,

acetate, stabilizers, self-cure initiators, light-cure initiators

The materials were handled according to the manufacturers instructions

(Table II).

After storage in water for 24 h at room temperature, seven roots from each

group were randomly selected for the evaluation of push-out strength, and

the remaining two roots that were left behind were used for TEM

examination.

127

Table II. Manufacturers’ instructions for the handling of the resin cements

Excite DSC/Variolink II -Etch with 37% phosphoric acid gel for 15 seconds, rinse with water from an endodontic syringe, remove excess water with paper points. -Apply Excite DSC in four to five layers by means of the self-activating microbrush. remove the excess adhesive with paper points. -Mix base and catalyst of Variolink II, carry the cement into the root canal with a lentulo drill. -Insert the post and light-cure the cement for 20 seconds through the post (Optilux 501 Kerr/Demetron, Orange, CA, USA, 750 mW/cm2).

ED Primer/Panavia 21 -Mix equal amounts of ED Primer liquids A and B. Apply the mix inside the canal with a brush and leave it undisturbed for 60 seconds. Remove excess adhesive with paper points, dry with a gentle air flow. -Mix equal amounts of base and catalyst for 20 seconds, apply the cement onto the post with a brush. -Insert the post and let the cement cure without any interference. RelyX Unicem -Dry the canal with paper points and a gentle blow of air. -Mix powder and liquid by triturating the activated capsule. -Apply the cement onto the post surface. -Insert the post and let the cement cure initially without any interference, followed by light-curing for 20 s through the post.

Push-out Strength Evaluation

The portion of each root that contained the bonded fiber post was sectioned

into five to six 1 mm-thick serial slices with the Isomet saw under water

cooling (Figure 1). Seven bonded roots were used for each group, resulting

in 36-37 slices per group for push-out strength evaluation.

The thickness of each slice was individually measured by means of a digital

caliper, and then firmly fixed with cyanoacrylate glue to a loading fixture. A

compressive load was applied on the apical aspect of the slice via a

universal testing machine (Controls S.P.A., Milano. Italy) that was equipped

with a 1mm-diameter cylindrical plunger (Figure 1).

The plunger was positioned so that it only contacted the bonded post on

loading, introducing shear stresses along the bonded interfaces. The loading

force was exerted in an apical-coronal direction, so as to move the post

toward the larger part of the root slice. Loading was performed at a speed of

0.5 mm/min until failure, as manifested by the extrusion of the post segment

128

from the root slice. This was further confirmed by the appearance of a sharp

drop along the load/time curve recorded by the testing machine.

The interfacial strength (MPa) was computed by dividing the load at

debonding by the area (A) of the bonded interface. The latter was calculated

through the formula A=2πrh, where r represents the post radius, and h the

thickness of the slice in mm. Fig. 1 A schematic representation of the preparation of thin root slices containing the bonded fiber post, and the set up for the push-out test.

Statistical Analysis

The interfacial strength data were first verified by the Kolmogorov-Smirnof

test for their normal distribution, and then using regression analysis to

ensure that the root of origin was not a significant factor for differences in the

strength measurements. A one-way analysis of variance was subsequently

performed, to assess the significance of the differences in interfacial strength

among three resin cements, followed by the Tukey test for post-hoc

comparisons. In all the analyses the level of significance was set at the 95%

probability level.

TEM Examination

A root slice was retrieved from the center portion of the bonded fiber post of

each of the two remaining teeth in each experimental group. The root slices

were fixed initially in Karnovsky’s fixative, post-fixed with osmium tetroxide,

dehydrated in an ascending ethanol series, and processed for epoxy resin

embedding according to the protocol reported by Tay, Moulding & Pashley36

129

for undemineralized TEM specimen preparation. After complete

polymerization of the laboratory epoxy resin (TAAB 812, TAAB Laboratories,

Aldermaston, UK) at 60°C for 48 h, the bulk of the fiber post was carefully

removed from the resin blocks with a 30-flute tungsten carbide dental bur

under an endodontic microscope (OPMI pico, Carl Zeiss, Oberkochen,

Germany). 2X2 mm blocks containing the resin cement-root dentin

interfaces were sectioned from the root slice and further embedded in epoxy

resin for ultramicrotomy. 90-120 nm thick, undemineralized sections were

prepared and collected on carbon- and formvar-coated single slot copper

grids (Electron Microscopy Sciences, Fort Washington, PA, USA) and

examined without further staining, using a TEM (Philips EM208S,

Eindhoven, The Netherlands) operated at 80 kV.

Results The results of push-out strength measurements are represented in Table III. Table III. “Thin-slice” push-out test results

Resin cement systems

Number of specimens tested

Push-out strength (MPa) *

Excite DSC/ Variolink II 37 10.18 ± 2.89 A

ED Primer/ Panavia 21 36 5.04 ± 2.81 B

RelyX Unicem 37 5.01 ± 2.63 B

Interfacial strengths to root dentin achieved by Excite DSC/Variolink II

(Group I) was significantly higher (P<0.05) than those obtained for ED

Primer/Panavia 21 (Group II) and RelyX Unicem (Group III). There was no

difference between the interfacial strengths in Groups II and III (P>0.05).

The resin-dentin interface in Excite DSC/Variolink II specimens revealed

complete dissolution of the smear layer and the formation of a 8-10 µm thick

hybrid layer in which the collagen matrix was completely demineralized by

the phosphoric acid etchant (Figure 2). There was no separation of the

hybridized dentin from the resin cement.

130

When ED Primer/Panavia 21 was used for fiber post luting, the root dentin

smear layer was almost completely dissolved, and smear plugs were

retained within the tubular orifices. A 1-1.5 µm thick hybridized complex

could be seen, consisting of the some smear layer remnants and an

underlying partially demineralized dentin matrix (Figure 3). Gaps were

present between the unfilled primer layer and the surface of the hybridized

dentin.

Conversely, RelyX Unicem was unable to dissolve or etch through the 3-4

µm thick smear layer created in the intraradicular dentin (Figure 4). The

smear layer remained heavily mineralized and no hybrid layer was seen in

the intact root dentin. Although the self-adhesive cement appeared to adhere

to the smear layer, separation occurred between the smear layer and the

underlying root dentin. Fig. 2 Representative TEM micrograph of the interface between the root dentin and Excite DSC/Variolink II. A completely demineralized, 6-8 µm thick hybrid layer (H) was created after the root dentin (RD) was etched with phosphoric acid. A thin layer of unfilled adhesive could be recognized (pointer) and some of the resin cement (RC) were found within some of the dentinal tubules (arrow).

131

Fig. 3 Representative TEM micrograph of the interface between the root dentin and ED Primer/Panavia 21. A 1 µm thick of partially demineralized hybridized dentin (H) was present, together with some hybridized smear layer remnants. This hybridized complex was separated from the unfilled primer (P) and resin cement (RC), and the space (asterisk) was infiltrated with laboratory epoxy resin. Smear plugs (SP) were present in the dentinal tubules. RD: root dentin.

Fig 4. Representative TEM micrograph of the interface between the root entin and RelyX Unicem. A 3-4 µm thick, loosely organized, highly mineralized smear layer (S) was present that was partially separated from the underlying unetched, intact root dentin (RD). The self-adhesive resin cement (RC) was predominantly attached to the smear layer and was only partially delaminated from the latter. The weak link in this resin cement was more likely to be caused by its inability to etch through a clinically-relevant, thick smear layer. Smear plugs (SP) were also evident within the dentinal tubules. Spaces between the root dentin, the smear layer, and the resin cement (asterisks) were infiltrated by the laboratory embedding epoxy resin.

132

Discussion In the light of the push-out test results, the null hypothesis that there is no

difference in the efficacy of the three resin cements for bonding of glass fiber

posts to intraradicular dentin has to be rejected. Indeed, the system Excite

DSC/Variolink II demonstrated a significantly greater bonding potential than

the other two materials (Table III).

As the three resin cement systems contain proprietary ternary catalytic

systems making them optimally compatible with acidic resin monomers,37 the

differences observed cannot be attributed to the incompatibility between

resin cements and dentin adhesives.

On the contrary, our TEM findings suggest that the differences among the

adhesive systems may be partially attributed to the ability of the dentin

adhesives or self-adhesive resin cements to etch through clinically relevant,

thick smear layers.38

Most of the in vitro studies on mild self-etching adhesive products performed

by manufacturers are conducted by polishing dentin with 600-grit silicon

carbide papers that produced thin smear layers that are less than 1 µm

thick.39,40

It has been shown that self-etch adhesives of different aggressiveness

respond variably to smear layers created by different bur types.41 In this

study, the intraradicular dentin was prepared using slow speed stainless

steel drills, creating dentin smear layers that are 3-4 µm thick (Figure 4). The

use of phosphoric acid etching completely dissolved such a thick smear

layer (Figure 2), while the use of the milder self-etching ED primer partially

dissolved the smear layer (Figure 3). Nevertheless, in both resin cement

systems, the complementary adhesives were able to etch into the underlying

dentin to create micromechanical retention with the intact bonding

substrates. Similar to our previous examination of the bonding efficacy of

RelyX Unicem on crown dentin,23 this mild self-adhesive resin cement is

ineffective in etching through clinically relevant smear layers created in both

coronal and intraradicular dentin. Although this initially anhydrous resin

cement system may bond to the smear layer via the unsubstantiated

133

mechanisms of water generation and subsequent water recycling proposed

by the manufacturer,23 the weak link in this system lies in its lack of genuine

hybridization of the intact bonding substrates. This is clearly illustrated by the

failure between the smear layer and the unetched intraradicular dentin in

Figure 4.

Although ED primer was able to etch through the smear layer and created a

thin zone of partially demineralized, hybridized dentin, this self-etching

primer is a one-step self-etch adhesive that is designed exclusively for the

resin cement. Like all one-step self-etch adhesives currently available in the

market that behave as permeable membranes after polymerization,42,43

crown dentin that was treated with the proprietary ED primer was also highly

permeable to water movement.18,22 Using silver nitrate as a tracer,

mushroom-shaped water blisters has been previously observed between

crown dentin and the ED primer.18 These water blisters may act as stress

raisers that result in premature delamination of the primer layer from the

hybridized dentin complex. As vital teeth and endodontically treated teeth do

not differ significantly in their moisture content,44 the effect of adhesive

permeability is also applicable to bonding within root canals. Although a

positive pulpal pressure is absent in endodontically-treated teeth, increase in

radicular permeability may follow reduction in root dentin thickness and

removal of sealers that penetrated the dentinal tubules during the

preparation of post-spaces for cementation of endodontic posts.45,46 By

taking impressions of intraradicular dentin from dowel spaces in human

patients that were bonded with simplified dentin adhesives, water droplets

were detected along the surface of the polymerized adhesives (Chersoni &

others, unpublished results). Such a scenario may be responsible for the

weak push-out strengths recorded for this simplified, self-etch resin cement

system.

The “thin slice” push-out test, adapted from its widespread use for testing of

ceramic matrix composites (CMCs), metal matrix composites (MMCs), and

intermetallic matrix composites (IMCs), is emerging as a practical tool for

evaluating the interfacial shear behavior of the attachment of fiber posts to

134

intact root canals. The latest studies in this field have highlighted the

important contribution of sliding friction to the interfacial strength in

composite materials.47-50 By plotting load/displacement curves of reinforcing

fibers slowly pushed-out or pulled-out of the embedding matrix, it has been

shown that the friction between the newly debonded interfaces plays a major

role in delaying the final failure of the specimen, thus significantly increasing

the load carrying capability of the composite material.47 When a compressive

load is applied on top of a fiber, friction occurs between the debonded

portion of the fiber and the facing matrix, whereas shear stress continues to

develop at the front of the propagating crack. From complete debonding to

extrusion, only friction opposes to fiber dislocation.35,48

The described progress can be assumed to occur also in the push-out test of

an endodontic post. The retentive strength of a bonded post can be

considered as the combined result of micromechanical interlocking, chemical

bonding, and sliding friction. Thus, interpreting the results derived from a

record of the maximum load during a push-out test as “bond strength”, as it

has commonly been referred to in the dental literature1,32,51 has to be viewed

upon with reservation. In this study, we prefer to address our results as

push-out strengths.

In our experimental setting, the fiber post surface was silanized in order to

enhance the post-cement bond. In addition, in a previous study where the

bond strengths of several cements at the post-cement and the cement-

dentin interfaces were assessed separately, all the tested luting materials

achieved a stronger adhesion to the post than to root dentin.1 Furthermore,

in previous microtensile bond strength tests on composite overlays luted to

coronal dentin with RelyX Unicem23 and Panavia F22, a microscopic analysis

of the fractured specimens revealed that the most frequent failure mode was

adhesive along the cement-dentin interface. Based on these premises,

failure of the bonded posts in the present study can be anticipated to occur

at the cement-dentin interface, and to be accompanied by the development

of friction between the cement-coated post, possibly pictured as a “macro-

fiber”, and its embedding matrix, the root canal.

135

As the TEM results demonstrated the existence of gaps in the interfaces of

the self-etching resin cement system ED primer/Panavia 21 (Figure 3) and

the self-adhesive cement RelyX Unicem (Figure 4), we speculate that the

push-out strengths obtained in the present study for these two resin cements

were predominantly contributed by sliding friction. Such a speculation

appears to be supported by the very low and highly inconsistent microtensile

bond strengths results obtained for the bonding of fiber posts via resin

cements inside intact dowel spaces.24,28 We are currently testing this

hypothesis by performing comparative thin-slice push-out tests for total-etch

and self-etch resin cement systems in the presence and absence of dentin

adhesive applications.

The potential clinical implication from the present study is that the resistance

to dislocation of fiber posts from root canals via the use of mild self-etch or

self-adhesive resin cement systems may have very little to do with the actual

bonding ability of these systems, and may largely be contributed by friction

within the dowel spaces. Under such a premise, it would be of clinical

significance to compare the resistance to dislocation of fiber posts that are

luted with conventional zinc phosphate cements or resin-modified glass-

ionomer cements, or posts that are constructed out of bonded amalgams,

with the timely and much advertised resin composite cement systems.

Ongoing work is also being performed in our laboratories to evaluate the

long-term aging of resin cement systems for luting of fiber posts with the use

of the thin slice push-out test.

Conclusions Interfacial strengths and ultrastructural findings concurrently demonstrated a

greater bonding potential of the total-etch resin cement investigated.

Conversely, the acidic resin monomers responsible for substrate

conditioning in Panavia 21 and RelyX Unicem appeared to be less effective

in etching through the thick smear layer created on root dentin during post

space preparation. This may have accounted for the significantly lower

retentive strength recorded by posts luted with these materials.

136

Acknowledgements

The fiber posts examined in this study were generously sponsored by Ivoclar-Vivadent, and the

resin cements by Ivoclar-Vivadent, Kuraray Medical Inc., and 3M ESPE. The TEM work was

supported by grant 20003755/90800/08004/400/01, Faculty of Dentistry, University of Hong

Kong. The authors are grateful to Anna Tay and Cris Ferrari for secretarial support.

137

References 1. Boschian Pest L, Cavalli G, Bertani P, Gagliani M (2002). Adhesive post-

endodontic restoration with fiber posts: push-out tests and SEM

observations. Dental Materials 18(8) 596-602.

2. Asmussen E, Peutzfeldt A, Heitmann T (1999). Stiffness, elastic limit,

and strength of newer types of endodontics posts. Journal of Dentistry

27(4) 275-278.

3. Albuquerque Rde C, Polleto LT, Fontana RH, Cimini CA (2003). Stress

analysis of an upper central incisor restored with different posts. Journal of

Oral Rehabilitation 30(9) 936-943.

4. Akkayan B, Gulmez T (2002). Resistance to fracture of endodontically

treated teeth restored with different post systems. Journal of Prosthetic

Dentistry 87(4) 431-437.

5. Dallari A, Rovatti L (1998). Six years of in vitro/in vivo experience with

Composipost. Compendium of Continuing Education in Dental Supplies

20 S57-S63.

6. Fredriksson M, Astback J, Pamenius M, Arvidson K (1998). A

retrospective study on 236 patients with teeth restored by carbon fiber-

reinforced epoxy resin posts. Journal of Prosthetic Dentistry 80(2) 151-

157.

7. Ferrari M, Vichi A, García-Godoy F (2000). A retrospective study of fiber-

reinforced epoxy resin posts vs cast posts and cores: a four year recall.

American Journal of Dentistry 13 B9-B14.

8. Monticelli F, Grandini S, Goracci C, Ferrari M (2003). Clinical behavior of

translucent fiber posts: a two-year prospective study. International Journal

of Prosthodontics 16(6) 593-596.

9. Nikaido T, Takano Y, Sasafuchi Y, Burrow MF, Tagami J (1999). Bond

strengths to endodontically-treated teeth. American Journal of Dentistry

12(4) 177-180.

138

10. Morris MD, Lee KW, Agee KA, Bouillaguet S, Pashley DH (2001).

Effects of sodium hypochlorite and RC-prep on bond strengths of resin

cement to endodontic surfaces. Journal of Endodontics 27(12) 753-757.

11. Ngoh EC, Pashley DH, Loushine RJ, Weller N, Kimbrough F (2001).

Effects of eugenol on resin bond strengths to root canal dentin. Journal of

Endodontics 27(6) 411-414.

12. Erdemir A, Ari H, Gungunes H, Belli S (2004). Effect of medications for

root canal treatment on bonding to root canal dentin. Journal of

Endodontics 30(2) 113-116.

13. Ferrari M, Mannocci F, Vichi A, Cagidiaco MC, Mjör IA (2000). Bonding

to root canal: structural characteristics of the substrate. American Journal

of Dentistry 13(5) 255-260.

14. Mannocci F, Sherriff M, Ferrari M, Watson TF (2001). Microtensile bond

strength and confocal microscopy of dental adhesives bonded to root

canal dentin. American Journal of Dentistry 14(4) 200-204.

15. Serafino C, Gallina G, Cumbo E, Ferrari M (2004). Surface debris of

canal walls after post space preparation in endodontically treated teeth: a

scanning electron microscopic study. Oral Surgery Oral Medicine Oral

Pathology Oral Radiology and Endodontics 97(3) 381-387.

16. Dong CC, McComb D, Anderson JD, Tam LE (2003). Effect of mode of

polymerization of bonding agent on shear bond strength of autocured resin

composite luting cements. Journal of the Canadian Dental Association

69(4) 229-234.

17. Pfeifer C, Shih D, Braga RR (2003). Compatibility of dental adhesives

and dual-cure cements. American Journal of Dentistry 16(4) 235-238.

18. Carvalho RM, Pegoraro TA, Tay FR, Pegoraro LF, Silva NR, Pashley

DH (2004). Adhesive permeability affects coupling of resin cements that

utilise self-etching primers to dentine. Journal of Dentistry 32(1) 55-65.

19. Alster D, Feilzer AJ, de Gee AJ, Davidson CL (1997). Polymerization

contraction stress in thin resin composite layers as a function of layer

thickness. Dental Materials 13(3) 146-150.

139

20. Hagge MS, Wong RD, Lindemuth JS (2002). Effect of dowel space

preparation and composite cement thickness on retention of a

prefabricated dowel. Journal of Prosthodontics 11(1) 19-24.

21. De Jager N, Pallav P, Feilzer AJ (2004). The apparent increase of the

Young's modulus in thin cement layers. Dental Materials 20(5) 457-462.

22. Mak YF, Lai SC, Cheung GS, Chan AW, Tay FR, Pashley DH (2002).

Microtensile bond testing of resin cements to dentin and an indirect resin

composite. Dental Materials 18(8) 609-621.

23. Goracci C, Ferrari M, Grandini S, Monticelli F, Tay FR (2004). Bonding

of a self-adhesive resin cement to dental hard tissues. Journal of Adhesive

Dentistry (in press).

24. Bouillaguet S, Troesch S, Wataha JC, Krejci I, Meyer JM, Pashley DH

(2003). Microtensile bond strength between adhesive cements and root

canal dentin. Dental Materials 19(3) 199-205.

25. Foxton RM, Nakajima M, Tagami J, Miura H (2003). Bonding of photo

and dual-cure adhesives to root canal dentin. Operative Dentistry 28(5)

543-551.

26. Gaston BA, West LA, Liewehr FR, Fernandes C, Pashley DH (2001).

Evaluation of regional bond strength of resin cement to endodontic

surfaces. Journal of Endodontics 27(5) 321-324.

27. Sahafi A, Peutzfeldt A, Asmussen E, Gotfredsen K (2003). Bond

strength of resin cement to surface-treated posts of titanium alloy, glass

fiber, and zirconia, and to dentin. Journal of Adhesive Dentistry 5(2) 153-

162.

28. Goracci C, Tavares AU, Fabianelli A, Monticelli F, Raffaelli O, Cardoso

PC, Tay FR, Ferrari M (2004). The adhesion between fiber posts and root

canal walls: comparison between microtensile and push-out bond strength

measurements. European Journal of Oral Sciences 2004; 112: 353-361.

29. Withers PJ, Bhadeshia HKDH (2001). Residual stress. Part I –

measurement techniques. Materials Science and Technology 17 355-365.

140

30. Berzins A, Summer DR (1999). Implant pushout and pullout tests. In: An

YH, Draughn RA, eds. Mechanical testing of bone and the bone–implant

interface. Boca Raton, FL: CRC Press LCC, 463-476.

31. Dhert WJA, Jansen JA (1999). The validity of a single pushout test. In:

An YH, Draughn RA, eds. Mechanical testing of bone and the bone–

implant interface. Boca Raton, FL: CRC Press LCC, 477-488.

32. Kurtz JS, Perdigão J, Geraldeli S, Hodges JS, Bowles WR (2003). Bond

strengths of tooth-colored posts, effect of sealer, dentin adhesive, and root

region. American Journal of Dentistry 16(Spec No) 31A-36A.

33. Prisco D, De Santis R, Mollica F, Ambrosio L, Rengo S, Nicolais L

(2003). Fiber post adhesion to resin luting cements in the restoration of

endodontically treated teeth. Operative Dentistry 28(5) 515-521.

34. Kallas M, Koss D, Hahn H, Hellmann J (1992). Interfacial stress state

present in a "thin-slice" fiber push-out test. Journal of Materials Sciences

27(18) 3821-3826.

35. Chandra N, Ananth CR (1995). Analysis of interfacial behavior in MMCs

and IMCs using thin slice push-out tests. Composites Science and

Technology 54(1) 87-100.

36. Tay FR, Moulding KM, Pashley (1999). Distribution of nanofillers from a

simplified-step adhesive in acid conditioned dentin. Journal of Adhesive

Dentistry 1(2) 103-117.

37. Suh BI, Feng L, Pashley DH, Tay FR (2003). Factors contributing to the

incompatibility between simplified-step adhesives and chemically-cured or

dual-cured composites. Part III. Effect of acidic resin monomers. Journal of

Adhesive Dentistry 5(4) 267-282.

38. Ogata M, Harada N, Yamaguchi S, Nakajima M, Tagami J (2002). Effect

of self-etching primer vs phosphoric acid etchant on bonding to bur-

prepared dentin. Operative Dentistry 27(5) 447-454.

39. Tay FR, Carvalho R, Sano H, Pashley DH (2000). Effect of smear layers

on the bonding of a self-etching primer to dentin. Journal of Adhesive

Dentistry 2(2) 99-116.

141

40. Chan KM, Tay FR, King NM, Imazato S, Pashley DH (2003). Bonding of

mild self-etching primers/adhesives to dentin with thick smear layers.

American Journal of Dentistry 16(5) 340-346.

41. Ogata M, Harada N, Yamaguchi S, Nakajima M, Pereira PN, Tagami J

(2001). Effects of different burs on dentin bond strengths of self-etching

primer bonding systems. Operative Dentistry 26(4) 375-382.

42. Tay FR, Pashley DH, Suh BI, Carvalho RM, Itthagarun A (2002). Single-

step adhesives are permeable membranes. Journal of Dentistry 30(7) 371-

382.

43. Chersoni S, Suppa P, Grandini S, Goracci C, Monticelli F, Yiu C, Huang

C, Prati C, Breschi L, Ferrari M, Pashley DH, Tay FR (2004). In vivo and in

vitro permeability of one-step self-etch adhesives Journal of Dental

Research 83(6) 459-464.

44. Papa J, Cain C, Messer HH (1994). Moisture content of vital vs

endodontically treated teeth. Endodontics and Dental Traumatology 10(2)

91-93.

45. Fogel HM, Marshall FJ, Pashley DH (1988). Effects of distance from the

pulp and thickness on the hydraulic conductance of human radicular

dentin. Journal of Dental Research 67(11) 1381-1385.

46. Guignes P, Faure J, Maurette A (1996). Relationship between

endodontic preparations and human dentin permeability measured in situ.

Journal of Endodontics 22(2) 60-67.

47. Lin G, Geubelle PH, Sottos NR (2001). Simulation of fiber debonding

with friction in a model composite pushout test. International Journal of

Solids Structures 38(46-47) 8547-8562.

48. Chandra N, Ghonem H (2001). Interfacial mechanics of push-out tests:

theory and experiments. Composites A (Composites Part A: Applied

Science and Manufacturing) 32(3-4) 578-584.

49. Chai YS, Mai Y (2001). New analysis on the fiber push-out problem with

interface roughness and thermal residual stress. Journal of Materials

Science 36 2095-2104.

142

50. Kalton AF, Howard SJ, Janczak-Rusch J, Clyne TW (1998).

Measurement of interfacial fracture energy by single fibre push-out testing

and its application to the titanium-silicon carbide system. Acta Materialia

46 3175-3189.

51. Vallittu PK, Kurunmaki H (2003). Bond strength of fibre-reinforced

composite to the metal surface. Journal of Oral Rehabilitation 30(9) 887-

892.

143

IV.2 The contribution of friction to the dislocation resistance of bonded fiber posts Goracci C, Fabianelli A, Sadek FT, Papacchini F, Tay FR, Ferrari M. Journal

of Endodontics 2004, in press.

Introduction Improvements in dentin adhesive technology in the past decade have

fostered attempts to reduce coronal leakage1 and improve retention2 by

bonding to root canals in the restoration of endodontically-treated teeth.

However, due to the highly unfavorable cavity configuration factors3

encountered within post spaces,4,5 and the high wall-to-wall contraction

experienced in thin resin films,6 bonding of posts to intraradicular dentin

presents challenges in relieving shrinkage stresses that are generated along

canal walls during the polymerization of resin cements.5,7 Recent studies

indicated that restorations bonded with fiber posts fail via the dislodging of

the bonded posts from root canals.8,9 These results were supported by the

frequent observations of premature bond failures when root sections

containing fiber posts bonded to root canals were prepared for microtensile

bond testing.10

To prevent superimposing disruptive stresses11,12 during specimen trimming,

the “thin slice” push-out test13 has been advocated as a more forgiving test

for evaluating the fixation strengths of fiber posts bonded to root canals.10 It

is known that sliding friction derived from interfacial roughness14 contributes

substantially to the results derived from push-out tests of composite

materials.15-17 The discrepancy in experiences with the microtensile and

push-out tests10 strongly suggests the dislocation resistance of bonded fiber

posts may be largely derived from sliding friction. Thus, the objective of this

study was to examine, with the use of a push-out test, the fixation strengths

of fiber posts that were cemented with either resin cements only, or in

conjunction with a self-etch and a total-etch dentin adhesive. The null

hypothesis tested was that the use of dentin adhesives produces no

144

additional improvement on the fixation of fiber posts with resin cements in

endodontically-treated teeth.

Methods Thirty-six single-rooted teeth were mechanically cleaned with a curette to

remove soft tissue remnants from the root surfaces. The crown of each tooth

was removed at 2-mm beneath the cementoenamel junction using a high

speed diamond saw (Isomet, Buehler, Lake Bluff, IL) under water cooling.

The working length was established 1-mm short of the apex. Instrumentation

of the root canals was performed with a crown-down technique, using Profile

nickel-titanium rotary instruments (Dentsply TulsaDental, OK). All canals

were prepared to ISO size 35, 0.06 taper. Each canal was irrigated in

between instrumentation with 17% EDTA and 5% sodium hypochlorite, dried

with paper points and obturated with gutta-percha and AH26 (Dentsply

DeTrey, Konstanz, Germany). Downpacking was performed using the

continuous wave warm vertical compaction technique (System B,

SybronEndo, Orange, CA), and backfilling was performed with Obtura II

(Spartan, Fenton, MO).

After 24 hr, the gutta-percha was removed from the coronal and middle

thirds of each root. At least 5 mm of intact gutta-percha and sealer was left

behind to preserve the apical seal. A dowel space was then prepared with

increasing sizes of post hole drills (FRC Postec, Ivoclar-Vivadent, Schaan,

Liechtenstein). A size #3 drill was used as the largest drill, corresponding to

the same size of a tapered FRC Postec glass fiber post (1mm in diameter

apically). Each post was silanized with Monobond-S (Ivoclar-Vivadent) prior

to cementation.

The teeth were randomly divided into two experimental groups, depending

on the type of resin cement system (self-etch vs total etch) employed for

post cementation. Each group was further divided into two subgroups (N=6),

according to whether the post spaces were treated with the corresponding

dentin adhesive (adhesive vs no adhesive).

Group I: Self-etch adhesive with self-cured resin cement

145

In subgroup IA the self-cured resin cement Panavia 21 (Kuraray

Medical Inc., Tokyo Japan) was employed without using the proprietary self-

etch adhesive. In subgroup IB, the post holes were treated with ED primer

for 60 s and dried with paper points prior to the application of Panavia 21.

Group II: Total-etch adhesive with dual-cured resin cement

In subgroup IIA, the dual-cure resin cement Variolink II (Ivoclar-

Vivadent) was applied without acid-etching and adhesive application. In

subgroup IIB, intraradicular dentin was etched with 37% phosphoric acid,

and bonded with the self-activated adhesive Excite DSC (Ivoclar-Vivadent;

Table I) prior to the application of resin cement.

Fixation Strength Evaluation

After storing in distilled water for 24 hr, each root was sectioned

transversally into 4-6 one-mm thick slices containing cross sections of the

fiber post. Seven specimens were used, resulting in 32-37 slices for each

subgroup. Fixation strength evaluation was performed by an evaluator who

was unaware of the group designations. Each slice was secured with

cyanoacrylate glue to a loading fixture. A compressive load was applied to

the slice via a 1-mm diameter cylindrical punch attached to a universal

testing machine, with the apical aspect of the slice (i.e. an inverted cone-

shaped post hole) facing the punch tip (Figure 1A).

Loading was performed at a crosshead speed of 0.5 mm/min until the post

segment was dislodged from the root slice. Interfacial fixation strength was

calculated by dividing the maximum failure load by the area of the bonded

interface. The data was analyzed using a two-way ANOVA and Tukey

multiple comparison tests with “root level” (coronal vs middle part of the root)

and “cementation procedure” (luting vs bonding) as factors, and with α=0.05.

Transmission Electron Microscopy (TEM)

The remaining two roots from each subgroup were evaluated for leakage

along the dentin-cement interface using a silver tracer penetration

technique.18 Similarly prepared slices were immersed in a 50 wt%

ammoniacal silver nitrate solution for 24 hr. The silver-impregnated slices

were fixed, dehydrated, and embedded in epoxy resin using the protocol

146

reported by Tay et al.19 Undemineralized, unstained 90-120 nm thick

sections were examined by an operator who was unaware of the group

designations, using a TEM (EM208S, Philips, Eindhoven, The Netherlands)

operating at 80 kV. Fig.1 A. A schematic of the “thin slice” push-out test employed for examination of the fixation of fiber posts to root canals. B. A load-displacement curve illustrating the progressive dislocation of a fiber through the surrounding matrix in a push-out test. Pmax represents the maximum shear stress recorded that is used in the calculation of the interfacial strength. Three zones could be recognized along the load-displacement curve. Zone I corresponds to the progressive increase in load prior to the initiation of delamination (Pi). Zone II represents a combination of partial delamination and frictional sliding. Zone III represents the onset of complete delamination in which there is a sudden drop in load (Pfr), with the resistance to further movement of the fiber contributed mainly by friction and surface roughness. (Modified from: Bechels VT, Sottos NR. Application of debond length measurements to examine the mechanics of fiber pushout. J Mech Phys Solids 1998;46:1675-1697).

Results

The fixation strengths obtained for the four subgroups are shown in the

Table.

147

Table. Push-out strengths of fiber posts coupled with the self-etch and total-etch resin cement systems to post holes created in root-treated teeth, with or without the use of proprietary dentin adhesives.

Group Subgroup Number of slices Push-out strength (Mpa)*

(A) Without

Adhesive 32 3.37 ± 2.89 A

(I) Panavia 21 (B)

Self-etch adhesive ED primer

36 5.04 ± 2.81 A

(A) Without

Adhesive 32 8.57 ± 2.50 B

(II) Variolink II

(B)

Total-etch adhesive Excite DSC

37 10.18 ± 2.68 B

*Values are means±standard deviations. Subgroups with the same letter superscripts are not statistically significant (P>0.05) Two-way ANOVA revealed that only the “cementation procedure” was a

significantly influenced the fixation strength results (P<0.05). Multiple

comparison tests further showed that for both resin cements, the fixation

strengths obtained from specimens luted with resin cement only did not differ

significantly from those that were first bonded with a dentin adhesive.

Cement-dentin interfaces in posts luted with Panavia 21 only revealed

compact smear layers with extensive silver deposits (Figure 2A). Application

of the self-etching primer did not completely dissolve the thick smear layer.

Partially dissolved smear layer particles20,21 and larger dentin chips were

dispersed within the primer-infiltrated smear layer. Silver infiltrated gaps

could be identified and no hybrid layer was evident (Figure 2B).

Extensive silver deposits were also found within gaps along the smear layer-

dentin junction and within the smear layer, in post spaces that were luted

with Variolink II only (Figure 2C). Thick smear layers that were not treated

with phosphoric acid contained large dentin chips and coarser smear layer

particles, with the Variolink II cement extending as discrete islands into the

smear layer (Figure 2D).

Due to the buffering capacity of the thick smear layers, phosphoric acid-

etching did not create uniform hybrid layers within the post space, with gaps

depicted by silver deposition along the Excite DSC-root dentin interface

148

(Figure 2E). In other better-etched areas, 4-6 µm thick hybrid layers were

observed. However, silver-filled gaps were still evident along the surface of

the hybrid layer (Figure 2F). Fig.2 TEM micrographs taken from post spaces with fiber posts bonded with resin cements with or without the adjunctive use of dentin adhesives. A. Panavia 21 (subgroup IA) without self-ecthing primer application. The thick smear layer (S) separated from the root dentin (RD) during sectioning, leaving behind an empty space (asterisk). Extensive leakage could be seen within the smear layer (pointer), between the smear layer and the dentin surface, and within the dentinal tubules (arrow). B. Panavia 21 (subgroup IB) with self-etching primer application. No hybrid layer was seen on the surface of the root dentin (D). A silver-infiltrated gap was evident (asterirsk) adjacent to the artifactual space (S). The remnant smear layer contained dentin chips (C) and smear layer particles (open arrowhead), with extensive silver deposits (pointers). Subsurface cracks created by the post hole drill were found in the root dentin (arrow), and within the dentinal tubules (arrow).

A B C. Variolink II (subgroup IIA) without total-etch adhesive application. A gap (arrows) was present between the root dentin (RD) and the smear layer. No hybrid layer was formed. The smear layer contained large dentin chips (C) and smear layer particles (open arrowhead), with extensive silver deposits (pointer). S: artifactual space. D. Variolink II (subgroup IIA), showing extension of islands of the resin cement (asterisk) into the smear layer. The latter contained dentin chips (C) and large smear layer particles (open arrowhead), with extensive silver deposits (pointer).

C D

149

E. Variolink II (subgroup IIB) with application of the total-etch adhesive. A thin hybrid layer (arrow) was occasionally observed along the root dentin surface (RD). The resin cement (RC) and adhesive (A) was separated from the dentin by a large gap (between open arrows). Silver infiltration was present in the original gap (asterisk) where dentinal chips from the smear layer were present (open arrowhead). Shrinkage during specimen processing created an artifactual gap (S) filled with laboratory epoxy resin (E). F. Variolink II (subgroup IIB), taken from the coronal part of the root canal where more efficient acid-etching resulted in the creation of a 4-6 µm thick, unstained, hybrid layer (H). Separation of the interface resulted in the presence of silver deposits (pointer) along the surface of the hybrid layer. The artifactual space caused by specimen shrinkage was filled with laboratory eposy resin (R). A: adhesive; RD: root dentin.

E F Discussion As neither the self-etch nor the total-etch adhesive produces significant

improvement on the dislocation resistance of fiber posts luted with the

respective resin cement, the null hypothesis tested in this study cannot be

rejected. The results of this study indicate that creating an adhesive

continuum or monobloc22 between the fiber posts and root dentin is not a

realistic expectation with the use of adhesive resin cements. For Panavia 21,

the inability of the mild self-etching primer23 to etch through thick smear

layers24 provided a reasonable explanation for the low fixation strengths

observed in the self-etch control and experimental subgroups. Similarly, for

Variolink II, application of the total-etch adhesive did not produce additional

improvement over the use of resin cement alone. Retention of thick smear

layers and other debris on root canal walls after acid-etching25 could have

prevented optimal adhesive infiltration. Adhesive resin cements are capable

of achieving high regional bond strengths to exposed root dentin under the

ideal conditions of optimal cleaning and maximum resin flow for shrinkage

150

stress relief.26 However, these criteria are difficult to realize when total-etch

adhesives are applied to post spaces, which may be viewed upon as deep,

narrow Class I cavity preparations3 with highly unfavorable cavity

configuration factors.5,10 Only slow setting, self-cured resin cements27 or

glass-ionomer cements28 are capable of providing the viscoelastic

parameters for bond integrity to be maintained under these extremely taxing

conditions.

The push-out test should not be misinterpreted as being more reliable than

microtensile bond tests in assessing the retention of fiber posts in root

canals.10 On the contrary, our results are in agreement with previous

engineering studies, in that a major contribution to the fixation strength is

contributed by interfacial sliding friction.15-17,29 The different stages involved

in a “thin slice” push-out test can be seen in a typical load-displacement

curve (Figure 1B) obtained when slowly pushing a fiber out of its surrounding

matrix.16 When a load is applied to the top of a fiber, shear stresses are

increasingly introduced to the top of the interfaces (Zone I). When the load

arrives at Pi, the shear stress reaches a critical value wherein delaminating

is initiated, usually resulting in a change in the slope of the load-

displacement curve (Zone II). Once delamination is initiated, the shear stress

in the delaminated zone drops, and the region of maximum shear stress

moves away from the top as the applied load continues to increase. During

the progressive delaminating phase (Zone II), frictional sliding occurs along

the delaminated upper portion of the fiber, while interfacial shear stresses

continue to be present along the propagating crack front. The Poisson’s

expansion of the fiber due to the applied stress in the upper delaminated

part increases the contact pressure, and results in increased work to

overcome friction. When the load reaches Pmax, the maximum shear stress

reaches the critical value at the bottom face. As a result, the entire length of

the fiber delaminates, causing a sudden sharp drop in the load Pfr (Zone III),

as the resistance to further movement of the fiber is mainly due to friction

and surface roughness.

151

By analogy, the same principles may be applied when pushing a fiber post

out of its surrounding matrix, the root canal. As the luting conditions in both

the self-etch and total-etch control groups were the same except for the use

of different resin cements, it is speculated there may be subtle differences in

the resin cements that resulted in an increase in the friction coefficient15 and

hence the higher fixation strength in the total-etch control group. It is not the

objective of this study to alert practitioners on the futility of using dentin

bonding systems in adhesive cementation of fiber posts. The intention,

rather, is to emphasize that similar to the fixation of endodontic posts9 and

hip arthroplasty stems30 with conventional non-bonding cements, there is a

predominant role contributed by sliding friction in the reported clinical

success of these procedures.

152

References

1. Reid LC, Kazemi RB, Meiers JC. Effect of fatigue testing on core

integrity and post microleakage of teeth restored with different post systems.

J Endod 2003;29:125-131.

2. Schwartz RS, Robbins JW. Post placement and restoration of

endodontically treated teeth: a literature review. J Endod 2004;30:289-301.

3. Davidson CL, de Gee AJ, Feilzer A. The competition between the

composite-dentin bond strength and the polymerization contraction stress. J

Dent Res 1984;63:1396-1399.

4. Morris MD, Lee KW, Agee KA, Bouillaguet S, Pashley DH. Effects of

sodium hypochlorite and RC-prep on bond strengths of resin cement to

endodontic surfaces. J Endod 2001;27:753-757.

5. Bouillaguet S, Troesch S, Wataha JC, Krejci I, Meyer JM and Pashley

DH. Microtensile bond strength between adhesive cements and root canal

dentin. Dent Mater 2003;19:99-205.

6. Feilzer AJ, De Gee AJ, Davidson CL. Increased wall-to-wall curing

contraction in thin bonded resin layers. J Dent Res 1989;68:48-50.

7. Ari H, Yasar E, Belli S. Effects of NaOCl on bond strengths of resin

cements to root canal dentin. J Endod 2003; 29:248-251.

8. Vichi A, Grandini S, Ferrari M. Comparison between two clinical

procedures for bonding fiber posts into a root canal: a microscopic

investigation. J Endod 2003; 28:355-360.

9. Garrido AD, Fonseca TS, Alfredo E, Silva-Sousa YT, Sousa-Neto MD.

Influence of ultrasound, with and without water spray cooling, on removal of

posts cemented with resin or zinc phosphate cements. J Endod

2004;30:173-176.

10. Goracci C, Tavares AU, Fabianelli A, Monticelli A, Raffaelli O, Cardoso

PEC, Tay FR, Ferrari M. The adhesion between fiber posts and root canal

walls: comparison between microtensile and push-out bond strength

measurements. Eur J Oral Sci 2004;112:353-361.

11. Withers PJ, Bhadeshia HKDH. Residual stress. Part I – measurement

techniques. Mater Sci Technol 2001;17:355-365.

153

12. Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshiyama M, Shono

Y, Fernandes CA, Tay F. The microtensile bond test: a review. J Adhes Dent

1999;1:299-309.

13. Chandra N, Ananth CR. Analysis of interfacial behavior in MMCs and

IMCs using thin slice push-out tests. Compos Sci Technol 1995; 54:87-100.

14. Li ZH, Bi XP, Lambros J, Geubelle PH. Dynamic fiber debonding and

frictional push-out in model composite systems: experimental observations.

Exp Mech 2002;42:417-425.

15. Lin G, Geubelle PH, Sottos NR. Simulation of fiber debonding with

friction in a model composite pushout test. Int J Solids Struct 2001;38:8547-

8562.

16. Chandra N, Ghonem H. Interfacial mechanics of push-out tests: theory

and experiments. Compos Part A: Appl Sci Manuf 2001;32:575-584.

17. Chai YS, Mai Y. New analysis on the fiber push-out problem with

interface roughness and thermal residual stress. J Mater Sci 2001;36:2095-

2104.

18. Tay FR, Pashley DH, Yoshiyama M. Two modes of nanoleakage

expression in single-step adhesives. J Dent Res 2002;81:472-476.

19. Tay FR, Moulding KM, Pashley DH. Distribution of nanofillers from a

simplified-step adhesive in acid conditioned dentin. J Adhes Dent

1999;1:103-117.

20. Pashley DH, Tao L, Boyd L, King GE, Horner JA. Scanning electron

microscopy of the substructure of smear layers in human dentine. Arch Oral

Biol 1998;33:265-270.

21. Tay FR, Carvalho R, Sano H, Pashley DH. Effect of smear layers on the

bonding of a self-etching primer to dentin. J Adhes Dent 2000;2:99-116.

22. Berekally T. Contemporary perspectives on post-core systems. Aust

Endod J 2003;29:120-127.

23. Carvalho RM, Pegoraro TA, Tay FR, Pegoraro LF, Silva NR, Pashley

DH. Adhesive permeability affects coupling of resin cements that utilise self-

etching primers to dentine. J Dent 2004;32:55-65.

154

24. Ogata M, Harada N, Yamaguchi S, Nakajima M, Tagami J. Effect of self-

etching primer vs phosphoric acid etchant on bonding to bur-prepared

dentin. Oper Dent 2002;27:447-454.

25. Serafino C, Gallina G, Cumbo E, Ferrari M. Surface debris of canal walls

after post space preparation in endodontically treated teeth: a scanning

electron microscopic study. Oral Surg Oral Med Oral Pathol Oral Radiol

Endod 2004;97:381-387.

26. Gaston BA, West LA, Liewehr FR, Fernandes C, Pashley DH.

Evaluation of regional bond strength of resin cement to endodontic surfaces.

J Endod 2001;27:321-324.

27. Bachicha WS, DiFiore PM, Miller DA, Lautenschlager EP, Pashley DH.

Microleakage of endodontically treated teeth restored with posts. J Endod

1998;24:703-708.

28. Dauvillier BS, Feilzer AJ, de Gee AJ, Davidson CL. Visco-elastic

parameters of dental restorative materials during setting. J Dent Res

2000;79:818-823.

29. Shirazi-Adl A, Forcione A. Finite element stress analysis of a push-

out test. Part II: Free interface with nonlinear friction properties. J Biomech

Eng 1992;114:155-161.

30. Nuňo N, Amabili M, Groppetti R, Rossi A. Static coefficient of friction

between Ti-6Al-4V and PMMA for cemented hip and knee implants. J

Biomed Mater Res 2001;59-191-200.

155

SUMMARY, GENERAL DISCUSSION, CONCLUSIONS, AND FUTURE DIRECTIONS

The microtensile test is a versatile technique for bond testing. Various

applications and several variants of the test have been proposed.

In the initial part of the project (Chapter I), some issues related to specimens

preparation and data interpretation in microtensile bond testing were

considered.

It was concluded that the shape and the thickness into which specimens

from enamel and dentin are prepared do have an influence on the measured

bond strength.

As far as the specimen shape is concerned, the trimming technique

producing hourglass-shaped specimens is adequate for ultimate tensile

strength testing of dental tissues and materials, for it ensures stress

concentration to occur over a surface whose area can be easily and

precisely calculated. However, this is essentially true only if the

MicroSpecimen Former is used, that is able to create a perfectly round

trimmed surface. As a matter of fact, the newest version of the

MicroSpecimen Former is now completely automated, avoiding stresses

introduced during hand movement of the rotation knob (Dr. Steve Armstrong,

personal communication). The same is unlikely to happen if trimming is done

free-hand with a high-speed handpiece, which also stresses the specimen in

a less controlled way. For this reason, free-hand trimming does not appear

as the method of choice when testing interfaces, particularly the tooth-

adhesive interface, and in general interfaces where relatively low levels of

bond strength are expected to occur. This statement is supported by the

research work reported in Chapter I, as well as by previous literature data.

Regarding specimen dimensions, higher bond strengths are measured by

thinner specimens. In a recent investigation by El Zohairy et al.1, this finding

is explained by the way microtensile specimens are usually fixed to the

testing jig, i.e. by their lateral sides.

156

The author of this thesis, however, is more convinced that it is the lower

density of defects, leading to a more uniform stress distribution, to be

responsible for the inverse relationship between cross-sectional area and

measured strength, as originally postulated by Griffith and later proposed by

Sano et al.2

To support this theory is also the microscopic evidence that faults do exist in

microtensile specimens at the adhesive interface, in the composite build-up

as well as within the dental substrates, with enamel being more prone to

defects (Chapter I.1.1).

Moreover, it appears that the 1x1 mm cross section represents the

acceptable compromise between desired stress uniformity and ease of

handling of the specimens. By preparing beam-shaped specimens in this

size, the bond strengths to enamel and dentin of several self-etching primers

were compared (Chapter I.2.1 and I.2.2).

With regard to the intrinsic flaws of the dental substrate or the built-up

material, they can be responsible for cohesive failures that, although far less

frequently than in conventional tensile or shear testing, can yet occur also in

microtensile. What interpretation to give to specimens that fail cohesively is

still object of discussion among researchers. The author’s opinion in this

regard is that, when the objective is to measure the interfacial adhesion,

cohesive failures should be eliminated because they bias the results. The

way cohesive failures are accounted for in the bond strength index of Reis et

al.3 appears arbitrary to the author of this thesis, who also questions the

interpretation given to premature failures in the index. As for the issue of

premature failures, the author agrees with Nikolaenko et al.4 that reporting

the number of pre-test failure per group is “an honest way to qualitatively

describe some deficiencies”, and therefore of interest for among groups

comparisons. Still it is the author’s belief that premature failures should be

excluded from statistical calculations, particularly in the case their inclusion

as “zero bond values” changes the data distribution to a non-normal one.

Concerning the statistical interpretation of bond strength data from tooth

specimens, it is agreed that the tooth-related variability should not be

157

overlooked. In this regard the Author of this thesis does not endorse the use

of ANOVA for repeated measures,5 since microtensile specimens are not

tested several times at consecutive intervals. Treating the tooth of origin as a

random factor seems more appropriate.6,7 Yet the approach preferred by the

author involves using the ANOVA test to check for significant differences in

bond strength among the teeth of each experimental group, prior to pooling

together the microtensile specimens from several teeth.8 Also a regression

analysis can be preliminarily run on microtensile specimens in order to verify

that the tooth of origin is not a significant factor in the variation of bond

strengths.

Provided that the specimen preparation procedure is properly chosen

according to the aim of the test and thoroughly carried out, microtensile

offers a dependable tool to measure the interfacial bond strength of

materials to dental and non-dental substrates. Chapter I and Chapter III

report examples of these applications of the microtensile test.

However, when microtensile is applied to measure the bond strength of

adhesive endocanalar posts, both variants of the technique appear to be

affected by severe limitations, encompassing the number of testable

specimens and data variability. Any cutting action is allegedly too aggressive

to be held by the relatively weak bonds established at the post-cement-

dentin interfaces. To this conclusion led the experimental work presented in

Chapter II. In this same research, in alternative to microtensile, the push-out

strength test is proposed as more practical and reliable for assessing the

retentive strength of a bonded post. In particular, the “thin-slice” variant of

the test, described in Chapter IV, favours the uniformity of stress distribution,

allows for discretion of regional differences in retentive conditions along the

root canal, and is useful to compare the retentive potential of various luting

agents, as done in the first study presented in Chapter IV.

However, when interpreting the results of the push-out test, one should

be aware of the contribution of sliding friction to what it is measured as

retentive strength. This contribution, according to Engineering studies

focusing on fiber-matrix push-out tests, is all but negligible. Sliding friction

158

may indeed account for most of the retentive strength of fiber posts luted

with self-etching and self-adhesive resin cements, that do not exhibit the

typical microscopic features of a reliable micromechanical bond. Friction may

also partly explain the clinical success of luting procedures using zinc

phosphate despite the lack of adhesion of this material.9 As a matter of fact,

in a recent experiment by Sadek et al.10, glass fiber posts luted with zinc

phosphate measured a push-out strength comparable to that achieved by

two resin cements used in combination with a total-etch and a self-etch

adhesive, and superior to the retentive strength yielded by a self-adhesive

resin cement.

In order to quantify the contribution of sliding friction along root canals in

the thin-slice push-out test of luted fiber posts, the second experimental work

presented in Chapter IV was carried out. This research, focusing on a total-

etch and a self-etch adhesive cement demonstrated that the contribution of

sliding friction was conspicuous for both materials and particularly

remarkable for the total-etch, in hypothetical relation with the cement surface

roughness.

This last study shed some further light on the largely used push-out test.

At the same time it opened a new perspective for the correct interpretation of

what in dental research is commonly and perhaps naively regarded as bond

strength.

Conclusions

1. The shape and the thickness of microtensile specimens from enamel

and dentin has an influence on the measured bond strength.

2. The hourglass design is appropriate for ultimate tensile strength testing

of dental tissues and materials, provided that trimming is performed in a

very controlled way, as by the MicroSpecimen Former.

3. Beam shaped specimens with a 1x1mm cross-section offer the most

desirable combination of favourable stress distribution and ease of

handling. This specimen design is adequate to measure interfacial bond

strength on enamel, coronal dentin, and dental materials.

159

4. When measuring the retentive strength of adhesive endocanalar posts,

the “thin-slice” push-out test emerges as a more straightforward and

dependable test than microtensile.

5. In the push-out test the friction developed by the post segment sliding

along the root canal gives a major contribution to the retentive strength

of the post.

Future directions

The microtensile test, at least in the way it has been performed in most of

the studies so far, provides only an assessment of a material or interface

static strength, i.e. the ability to resist a short-term steady load.11 However,

particularly in order to estimate the in-service performance of a restorative

material or a dental adhesive, also the ability of the material or the interface

to resist cyclic temperature changes and mechanical loadings should be

assessed through a fatigue test. Recently Frankenberger et al.12 have

propose a method to apply thermomechanical cycles directly to the bonded

interface on 4mm-high and 2mm-wide resin-dentin beams. The beams that

survived the cycles were then sectioned into smaller sticks, and the strength

of the loaded interface was tested in microtensile.

Moreover, for the establishment of a solid, durable joint not only tensile

strength, but also fracture toughness is a critical property of the adhesive

material.11,13 Many current adhesives exhibit relatively high stiffness and low

fracture toughness, which result in a brittle behaviour.11,14,15 It would then be

desirable to improve the ability of adhesive joints to absorb energy prior to

fracture. In order to achieve this, biomimetics has indicated the way of

aquatic adhesive organisms that owe their combination of strength and

toughness to the presence of protein-like domains able to unravel under

increasing stress. Chemists and biologists, however, have still to work on

making these proteinaceous components compatible with the extreme

environmental conditions currently involved in organic chemical syntheses.13

As the study into the push-out test highlighted, it seems wise that the dental

researchers of the future always keep an eye on the experience and the

160

advancements of materials scientists and engineers. From these disciplines

dental manufacturers and researchers may learn how to take advantage of

materials surface characteristics and frictional behaviour in order to improve

the retention and loading ability of restorations.

Another issue already addressed in engineering and which may become of

interest also in the dental materials science regards residual stress. Residual

stresses may arise in natural or artificial multiphase materials from

differences in thermal expansivity, yield stress and stiffness among different

regions.16 Residual stresses superimpose on in-service stresses and may

negatively affect the performance or the durability of a material. Conversely,

by an intelligent use of residual stress the static loading performance of

brittle materials can be improved. With this aim the effort to measure and

predict residual stresses is being made in engineering and material science.

This approach has been successfully followed in the fabrication of thermally

toughened glass and prestressed concrete.16 The relevance for dentistry of

these engineering advancements is evident if one considers that not only

dental materials, such as composite resins and fiber-reinforced composites,

but also dentin and enamel are multiphase materials with a brittle behaviour,

and which are called to perform a stress-bearing role ideally for a life-time

span.

Recently, Gao et al.17 drew attention to the significance for mechanical

properties of the elementary building blocks that exist in the nanometer

dimensions in biocomposites such as those that are present in enamel,

dentin, bone, and the nacre of abalone shell. The authors stated that the

organization of many biological hard tissues in mineral lamellae separated

by soft layers of protein matrix is intricately designed by nature for yielding a

high fracture toughness by stress redistribution and crack-stopping

mechanisms. More precisely, most of the load is carried by the mineral

platelets, whereas the protein matrix transfers the stress between platelets

as shear stress. Basically, the fracture toughness of the tissue relies heavily

on the tensile strength of the mineral component. Based on numerical

equations that refer to the Griffith criterion, Gao et al.17 came to the

161

conclusion that the nanometer size of mineral lamellae allows for strength

optimization and tolerance of flaws. At the nanometer scale the mineral

crystal approaches the strength of a perfect crystal, in other words the

theoretical strength, despite the presence of flaws. Conversely, when the

mineral dimension exceeds the nanometer scale, the material becomes

sensitive to crack-like defects, and fails by stress concentration at crack tips.

This study’s findings have opened a new path for bioengineers to follow in

their attempt to produce in laboratory new and superior nanomaterials.

Although this investigation has emphasized the importance of mechanical

strength as a driving force for nanostructural organization of biological

materials, however the authors pointed out that chemical factors also play a

crucial role in the formation and nucleation of mineral crystals.

Wherever future research may lead us to, one thing is evident - that nature is

always capable of teaching the most valuable lesson.

162

SOMMARIO, DISCUSSIONE COMPLESSIVA, CONCLUSIONI E DIREZIONI FUTURE

Il test microtensile è una tecnica versatile per la valutazione dell’adesione.

Sono state proposte varie applicazioni e diverse varianti del test.

Nella parte iniziale del progetto (Capitolo I), sono stati presi in

considerazione alcuni aspetti relativi alla preparazione dei campioni ed

all’interpretazione dei dati.

Si è giunti alla conclusione che la forma e lo spessore in cui i campioni

ottenuti da smalto e dentina vengono preparati influiscono sulla forza di

adesione che viene misurata.

Per quanto concerne la forma del campione, la tecnica “trimming”, che

produce campioni a forma di clessidra, è adatta alla misurazione della forza

tensile ultima di tessuti e materiali dentari, poiché fa sì che lo stress sia

concentrato su di una superficie la cui area può essere facilmente e

precisamente calcolata. In ogni caso, questo è sostanzialmente vero solo se

si usa il MicroSpecimen Former, che è in grado di sagomare il campione con

un ottimo controllo dei movimenti della fresa. Peraltro, la nuova versione del

MicroSpecimen Former è completamente automatizzata, evitando anche

che stress sia trasmesso al campione all’atto dello spostamento, finora

operato a mano, della manopola per la rotazione del campione

(comunicazione personale del Dr. Steve Armstrong). E’ improbabile che un

simile livello di controllo sia assicurato quando la sagomatura del campione

viene fatta a mano libera con un manipolo ad alta velocità, che sollecita il

campione in maniera meno uniforme. Per questa ragione la sagomatura del

campione a mano libera non sembra essere il metodo di scelta quando si

testano interfacce, l’interfaccia dente-adesivo in particolare, ma, più in

generale, tutte le interfacce in cui ci si aspetta che siano raggiunti livelli di

forza adesiva relativamente bassi. Questa affermazione è sostenuta dal

lavoro sperimentale riportato nel Capitolo I, così come da dati scientifici già

presentati in letteratura.

163

Riguardo alle dimensioni del campione, tendenzialmente i campioni più

sottili misurano forze di adesione più elevate. In una recente ricerca di El

Zohairy e coll.1, questo fenomeno viene messo in relazione al modo in cui i

campioni vengono fissati al dispositivo per il test, cioè dai lati del campione.

L’autrice di questa tesi, tuttavia, è più convinta che sia la più bassa densità

dei difetti intrinseci del campione, permettendo una distribuzione dello stress

maggiormente uniforme, ad essere responsabile della relazione inversa tra

area trasversale del campione e forza tensile che viene misurata, come

originariamente postulato da Griffith e più tardi riproposto da Sano e coll.2

A confermare questa teoria è anche la prova, con l’osservazione al

microscopio, che difetti strutturali sono effettivamente presenti nei campioni

per test microtensile all’interfaccia adesiva, nel build-up di composito ed

anche all’interno del substrato dentale; quanto a quest’ultimo, lo smalto

risulta più predisposto allo sviluppo di difetti strutturali (Capitolo I.1.1).

Inoltre risulta che l’area trasversale di 1x1mm rappresenti il compromesso

accettabile tra desiderabile uniformità dello stress e “maneggevolezza” dei

campioni. Campioni preparati in questa dimensione e in forma di bastoncino

sono stati utilmente testati per il confronto delle forze adesive su smalto e

dentina di vari self-ecthing primers (Capitoli I.2.1 e I.2.2).

Riguardo ai difetti intrinseci al substrato o al materiale dentale, essi possono

essere responsabili di fratture coesive dei campioni che, sebbene assai

meno frequentemente che nei test tensili e di taglio convenzionali, possono

comunque verificarsi anche nel test microtensile. Quale interpretazione dare

ai campioni che falliscono coesivamente è ancora oggetto di discussione tra

i ricercatori. A questo proposito, l’opinione dell’autrice è che, quando scopo

del test è misurare l’adesione all’interfaccia, le fratture coesive dovrebbero

esserre eliminate in quanto “confondono” i risultati del test. Il modo in cui le

fratture coesive vengono quantificate nell’indice di forza adesiva di Reis e

coll.3 appare arbitario all’autrice di questa tesi, che mette in discussione

anche l’interpretazione data nell’indice alle fratture premature. Riguardo alla

questione delle fratture premature, l’autrice, d’accordo con Nikolaenko e

coll.4, ritiene che riportare il numero di fratture occorse prima del test in ogni

164

gruppo sia “un modo onesto di descrivere quantitativamente certi difetti” e,

pertanto, un’informazione significativa ai fini del confronto tra gruppi.

Ciononostante, l’autrice crede che le fratture premature dovrebbero essere

escluse dall’analisi statistica, in particolare nel caso in cui la loro inclusione

come “valori zero” renda la distribuzione dei dati non normale.

Per quanto concerne l’interpretazione statistica dei dati di forza di adesione

di campioni ottenuti da denti, si è d’accordo sul fatto che la variabilità legata

al dente di origine non dovrebbe essere trascurata. A questo proposito,

l’autrice della tesi non sottoscrive l’uso del test statistico ANOVA per misure

ripetute5, poiché i campioni microtensili non vengono testati più volte ad

intervalli consecutivi. Considerare il dente di origine come un fattore casuale

sembra più appropriato.6,7 In ogni caso, l’approccio preferito dall’autrice

prevede l’uso del test ANOVA per verificare l’esistenza di differenze

significative nella forza di adesione tra denti di uno stesso gruppo

sperimentale, prima di raggruppare insieme i campioni microtensili

provenienti da più denti.8 Si può inoltre preliminarmente condurre un’analisi

regressiva sui campioni microtensili, allo scopo di verificare che il dente di

origine non abbia un’influenza significativa sulla variabilità esistente nelle

forze di adesione misurate dai campioni.

Se si sceglie una procedura di preparazione dei campioni in compatibilità

con lo scopo del test e questa procedura viene eseguita con attenzione, la

tecnica microtensile offre uno strumento affidabile per misurare la forza di

adesione interfacciale dei materiali a substrati dentali e non. I Capitoli I e II

riportano vari esempi di applicazione del test microtensile.

D’altra parte, quando il metodo microtensile viene applicato per misurare la

forza di adesione di perni endocanalari, entrambe le varianti della tecnica

presentano evidenti limiti, che si riflettono sul numero di campioni utili al test

ottenibili e sulla dispersione dei dati. Qualunque azione di taglio risulta

evidentemente troppo aggressiva per essere sostenuta dai legami

relativamente deboli esistenti a livello delle interfacce perno-cemento-

dentina. A questa conclusione ha condotto l’esperimento presentato nel

Capitolo II. In questa stessa ricerca, in alternativa alla tecnica microtensile,

165

viene proposto il test push-out come metodo più pratico ed affidabile per

misurare la forza ritentiva di un perno cementato. In particolare la variante “a

fetta sottile” del test, descritta nel Capitolo IV, assicura condizioni favorevoli

ad un’omogenea trasmissione dello stress, permette di individuare

differenze locali nelle condizioni di ritenzione lungo il canale radicolare, è

utile per confrontare il potenziale ritentivo di diversi materiali per

cementazione, come risulta dal primo studio presentato nel Capitolo IV.

Comunque, una corretta interpretazione dei risultati del test push-out

richiede la consapevolezza da parte del ricercatore che a ciò che si misura

come forza di ritenzione contribuisce significativamente l’attrito radente.

Secondo i risultati di studi di ingegneria su test push-out di fibre inglobate in

matrici di varia natura, il contributo dell’attrito radente è tutt’altro che

trascurabile. All’attrito radente può in effetti essere attribuita una consistente

quota della forza ritentiva di perni in fibra cementati con cementi resinosi

self-etch e self-adhesive, nel cui quadro microscopico non sono riconoscibili

i segni tipici di un affidabile legame micromeccanico. L’attrito può inoltre in

parte spiegare il successo clinico di procedure di cementazione che

utilizzano il fosfato di zinco, nonostante la mancanza di capacità adesiva di

questo materiale.9 Di fatto, in un recente esperimento di Sadek e coll.10,

perni in fibra di vetro cementati con fosfato di zinco hanno misurato una

forza al push-out paragonabile a quella di due cementi resinosi, usati in

combinazione con un adesivo self-etch o total-etch, nonché superiore alla

forza ritentiva raggiunta da un cemento self-adhesive.

Allo scopo di quantificare il contributo dell’attrito radente nel test push-out di

perni endocanalari in fibra è stato condotto il secondo lavoro sperimentale

presentato nel Capitolo IV. Questa ricerca, incentrata su cementi adesivi

total-etch e self-etch, ha dimostrato che il contributo dell’attrito è cospicuo

per entrambi i materiali e particolarmente consistente per il cemento adesivo

total-etch, in ipotetica relazione con la ruvidità di superficie del cemento.

Quest’ultimo studio ha approfondito la conoscenza di alcuni aspetti

meccanici, peraltro finora trascurati, del test push-out, che trova ampio

utilizzo nella ricerca in campo di materiali dentari. Allo stesso tempo il lavoro

166

sperimentale ha aperto una nuova prospettiva per la corretta interpretazione

di tutto ciò che in ricerca odontoiatrica viene comunemente, e forse

“ingenuamente”, considerato come forza adesiva.

Conclusioni

1. La forma e lo spessore dei campioni per test microtensile ottenuti da

smalto e dentina influiscono sulla forza di adesione che viene

misurata.

2. Il design a clessidra è adatto per il test di forza tensile ultima di

tessuti e materiali dentali, a patto che la sagomatura del campione

sia eseguita in modo altamente preciso e controllato, per esempio

utilizzando il MicroSpecimen Former.

3. I campioni a forma di bastoncino e con sezione trasversale di

1x1mm offrono la più soddisfacente combinazione di favorevole

distribuzione dello stress e “maneggevolezza”. Questo design è

indicato per misurare la forza di adesione interfacciale su smalto e

dentina coronale, nonché tra i materiali dentari.

4. Nella misurazione della forza ritentiva di perni endocanalari adesivi,

il test push-out “a fetta sottile” si propone come test più semplice ed

affidabile rispetto al microtensile.

5. Nel test push-out l’attrito sviluppato dal segmento di perno spinto

lungo il canale radicolare contribuisce significativamente alla forza

ritentiva del perno.

Direzioni future

Il test microtensile, per lo meno nel modo in cui è stato finora condotto

nella maggior parte degli studi, fornisce solo una misura della forza

statica di un materiale o un’interfaccia, definita come la capacità di

resistere ad un carico costante per breve tempo.11 Tuttavia, soprattutto

al fine di stimare la performance clinica di un materiale da restauro o di

un adesivo dentale, si dovrebbe valutare attraverso test di fatica anche

la sua capacità di reagire a periodiche variazioni di temperatura e di

resistere a carichi ciclici. Di recente Frankenberger e coll.12 hanno

167

proposto un metodo per applicare cicli termomeccanici direttamente

all’interfaccia adesiva in bastoncini di dentina e resina di 4 mm in altezza

e 2 mm in larghezza. I bastoncini che sono sopravvissuti ai cicli sono

stati quindi sezionati in sticks di dimensioni inferiori, la cui forza di

adesione interfacciale è stata misurata con il metodo microtensile.

Inoltre, a determinare la solidità e durevolezza di un’interfaccia adesiva,

non è soltanto la sua forza tensile, ma anche la sua resistenza a frattura

(fracture toughness).11,13 Molti adesivi attualmente in uso presentano una

rigidità piuttosto elevata ed una bassa resistenza alla frattura, che

risultano in un “comportamento fragile” del materiale.11,14,15 Sarebbe

pertanto desiderabile incrementare la capacità dell’interfaccia adesiva di

assorbire energia prima di giungere a frattura. Allo scopo di ottenere ciò,

la biomimetica ha studiato organismi adesivi acquatici che devono la

loro favorevole combinazione di forza e robustezza alla presenza di

domini proteici capaci di modificare la loro struttura, svolgendone le

involuzioni, all’aumentare dello stress. Chimici e biologi hanno tuttavia

ancora molto lavoro da compiere al fine di rendere queste componenti

proteinacee compatibili con le estreme condizioni ambientali in cui

attualmente si compie la sintesi di composti chimici organici.

Come evidenziato anche dallo studio sul push-out, sembrerebbe saggio

per il ricercatore dentale del futuro rivolgere puntualmente l’attenzione

all’esperienza ed ai progressi nel campo della scienza ed ingegneria dei

materiali. Da queste discipline i ricercatori e le ditte produttrici dei

materiali dentari possono apprendere come trarre vantaggio dalle

caratteristiche di superficie dei materiali e dall’attrito sviluppato, al fine di

incrementare la capacità ritentiva e di resistenza ai carichi.

Un altro aspetto già oggetto di studio in ingegneria e che potrebbe

diventare di interesse anche nella scienza dei materiali dentari riguarda

lo stress residuo. Stress residui possono originare nei materiali a fase

multipla naturali o artificiali dalle differenze in espansione termica, stress

di cedimento e rigidità tra diverse regioni del materiale.16 Gli stress

residui si sovrappongono a quelli legati alla funzione del materiale e

168

possono influire negativamente sulla sua durevolezza. Diversamente, un

uso intelligente dello stress residuo può migliorare la performance dei

materiali fragili nel carico statico. A questo scopo gli ingegneri e gli

scienziati dei materiali dentari stanno lavorando alla definizione di

metodi per la misura e la predizione degli stress residui. Questo

approccio è stato già applicato con successo nella fabbricazione del

vetro temprato a caldo e del cemento pre-caricato.16 La rilevanza in

campo odontoiatrico di queste acquisizioni dell’ingegneria è evidente se

si considera che non soltanto materiali dentari come resine composite e

compositi rinforzati da fibre, ma anche smalto e dentina sono materiali a

fase multipla, con comportamento meccanico fragile, e che sono

chiamati a svolgere una funzione di resistenza ai carichi, idealmente per

il tempo di una vita.

Recentemente Gao e coll.17 hanno richiamato l’attenzione sul significato,

ai fini delle proprietà meccaniche, delle dimensioni nanometriche delle

unità strutturali elementari di biocompositi come smalto, dentina, osso e

la madreperla di alcune conchiglie. Gli autori sostengono che

l’organizzazione di molti tessuti biologici duri in lamelle minerali separate

da strati più morbidi di matrice proteica è così prevista dalla natura allo

scopo di conferire un’elevata resistenza alla frattura attraverso

meccanismi di ridistribuzione dello stress e arresto di progressione delle

incrinature. Più precisamente, la maggior quota del carico è sostenuta

dalle lamelle minerali, mentre la matrice proteica trasferisce lo stress tra

le lamelle sotto forma di forze di taglio. In sostanza, la resistenza alla

frattura del tessuto dipende principalmente dalla forza tensile della

componente minerale. Sulla base di equazioni numeriche che fanno

riferimento al criterio di Griffith, Gao e coll.17 sono giunti alla conclusione

che la dimensione nanometrica delle lamelle minerali consente

un’ottimizzazione della loro resistenza e tolleranza dei difetti. Alla scala

nanometrica il minerale raggiunge livelli di resistenza vicini a quelli del

cristallo perfetto, in altre parole la resistenza teorica, indipendentemente

dalla presenza di difetti intrinseci. Diversamente, quando le dimensioni

169

del minerale superano la scala nanometrica, il materiale diviene

sensibile ai difetti intrinseci e cede per concentrazione dello stress in

corrispondenza di questi difetti. I risultati di questi studi hanno aperto ai

bioingegneri una nuova strada da seguire nel loro sforzo di produrre in

laboratorio nuovi nanomateriali dotati di superiori proprietà meccaniche.

Sebbene questo studio abbia enfatizzato l’importanza della resistenza

meccanica come proprietà determinante nell’organizzazione

nanostrutturale dei materiali biologici, tuttavia gli autori hanno anche

sottolineato come pure i fattori chimici giochino un ruolo cruciale nella

formazione e nucleazione dei cristalli minerali.

Ovunque la futura ricerca vorrà condurci, una cosa è certa – che è

sempre la natura a saper dare gli insegnamenti più validi.

170

RESUME’, DISCUSSION GENERALE, CONCLUSIONS ET DIRECTIONS FUTURES

Le test «microtensile» est une technique versatile qui peut être employé

pour l’évaluation de la adhésion. On a proposé de différentes applications et

des différentes variantes du test.

Dans la partie initiale du projet (Chapitre I), on a pris en considération des

aspects relatifs à la préparation des échantillons et à l’interprétation des

donnés.

La conclusion qu’on a fait est que la forme et l’épaisseur où les échantillons

obtenus par l’email et la dentine ont été préparés influence la force de

adhésion mesurée.

Pour ce qui concerne la forme de l’échantillons, la technique « trimming »,

qui produit des échantillons à forme de clepsydre, est approprié pour le

mesurage de la force « tensile » de tissus et de matériaux dentaires,

puisque cette forme concentre le stress sur une surface dont l’aire peut être

aisément et précisément calculée. Substantiellement on peut dire cela

seulement si l’on emploi le peut MicroSpecimen Former, qui peut façonner

l’échantillons avec un excellent contrôle des mouvement de la fraise. Du

reste la nouvelle version du MicroSpecimen Former est complètement

automatisé, et l’on évite aussi de transmettre du stress à l’échantillons au

moment du déplacement par la poigné pour la rotation de l’échantillon

(communication personnel du Dr Steve Armstrong). Il est improbable qu’un

niveau semblable de control soit assuré quand le façonnage de l’échantillon

est fait par un opérateur qui emploi une fraise diamanté (avec la turbine), et

tout cela donne du stress à l’échantillons d’un façon moins uniforme.

Pour cette raison le façonnage de l’échantillon avec la turbine ne semble

pas être la méthode la plus indiquée quand on teste les interfaces, et

particulièrement l’interface dente-adhesive, mais, plus en général, toutes les

interfaces où l’on attend des forces adhésives relativement basses. Cette

affirmation est soutenue par l’étude expérimental décrite dans le Chapitre I,

et aussi par des données scientifiques déjà présentes en littérature.

171

Pour ce qui concerne les dimensions de l’échantillon, fondamentalement les

échantillons les plus minces mesurent des forces de adhésion plus élevées.

Dans une recherche récente de El Zohairy et coll.1, ce phénomène est mis

en relation à la manière dans laquelle les échantillons sont fixés au dispositif

pour le test, c'est-à-dire latéralement à l’échantillon.

L’auteur de cette thèse, pourtant, est plus convaincu que la basse densité

des défaut intrinsèques de l’échantillon, en permettant une distribution du

stress plus uniforme, est responsable de la relation inverse entre l’aire

transversale de l’échantillon et la force « tensile » qui est mesuré, comme

postulé en origine par Griffith et plus tard reproposé par Sano et coll.2

En confirmation de cette théorie il y a aussi l’épreuve, avec l’observation au

microscope, que des défauts structural sont effectivement présents dans les

échantillons pour le test « microtensile » à l’interface adhésive, dans le

« build-up » de composite et aussi à l’intérieure du substrat dentaire; pour ce

qui concerne ce dernier, l’email résulte plus disposé aux développement des

défaut structuraux (Chapitre I.1.1).

De plus il est évident que l’aire transversale de 1x1 mm représente le

compromis acceptable entre l’uniformité qu’on désire du stress et la

«maniabilité» des échantillons. Des échantillons prépare dans cette

dimension et avec la forme de bâtonné ont été testé pour la comparaison

des forces adhésives sur l’email et la dentine de plusieurs (Chapitre I.2.1 et

Chapitre I.2.2).

Pour ce qui concerne les défauts intrinsèques aux substrats où au matériel

dentaire, ils peuvent provoquer des fractures cohésives des échantillons qui,

même si moins fréquemment que dans les tests tensile et de coupure

conventionnelle, peuvent de toute façon se vérifier même dans le test

microtensile. On discute encore sur l’interprétation à donner aux échantillons

qui échouent de façon cohésive. A ce propos, l’opinion de l’auteur est que,

quand le but du test est celui de mesurer l’adhésion à l’interface, on devrait

éliminer les fractures adhésives parce qu’elle « confondent » les résultats du

test. La façon dans laquelle les fractures cohésives sont quantifiées dans

l’index de force adhésive di Reis e coll.3 semble un abus à l’auteur de cette

172

thèse, qui aussi met en discussion l’interprétation donnés dans l’index aux

fractures prématurés. Pour ce qui concerne les fractures prématurées,

l’auteur, en accord avec Nikolaenko e coll.4, pense que le fait de reconduire

le nombre des fractures prématurés dans chaque groupe est « une manière

honnête de décrire quantitativement certaines défauts » et, pourtant, un

renseignement significatif pour la comparaison entre des groups.

Malgré cela, l’auteur croit qu’on doit éliminer les fractures prématurés de

l’analyse statistique, en particulier dans le cas où leur inclusion comme

« valeurs zero » rend la distribution des donnés pas normale.

Pou ce qui concerne l’interprétation statistique des données de force

d’adhésion d’échantillons obtenus avec les dents, on est d’accord sur le fait

que la variabilité liée à la dent d’origine ne doit pas être oublié. A ce propos,

l’auteur de la thèse ne suscrit pas l’usage du test statistique ANOVA pour

des mesure répétées, 5 parce que les échantillons microtensile ne sont pas

testés plusieurs fois à intervalles consécutifs. Il semble plus approprié de

considérer la dent d’origine comme un facteur casuelle .6,7 En tout cas,

l’auteur prévoit l’usage du test ANOVA pour vérifier l’existence de

différences significatives dans la force de adhésion entre les dents d’un

même group expérimental, avant de rassembler les échantillons microtensile

qui proviennent de plusieurs dents.8 Il est aussi possible au préalable

conduire une analyse régressive sur les échantillons microtensile, au but de

vérifier que la dent d’origine n’influence significativement la variabilité

existant dans les forces d’adhésion mesurées par les échantillons.

Si l’on choisit une procédure de préparation des échantillons compatible

avec le but du test, et si la procédure est correctement effectuée, la

technique microtensile offre un instrument fiable pour mesurer la force de

adhésion à l’interface des matériaux à substrat dentaire où pas. Les

chapitres I et II décrivent de différentes exemples de application du test

microtensile.

D’autre part, quand la méthode microtensile est appliqué pour mesurer la

force de adhésion de tenon à l’intérieure du canal, toutes les deux variantes

de la technique présentent des limites évidentes, qui se reflètent sur le

173

nombre de échantillons utiles au test et sur la dispersion des données.

N’importe quelle action de coupure résulte évidemment trop agressive pour

être soutenue par le lien relativement faible existent au niveau des interfaces

tenon-ciment-dentine. L’expérimentation présenté dans le chapitre II a

conduite à cette conclusion. Dans cette recherche, comme alternative au

microtensile, le test « push out » est proposé comme une méthode plus

pratique et fiable pour mesurer la force de rétention d’un tenon cimenté. En

particulier, la variante « à tranche mince » du test, décrite dans le chapitre

IV, assure des conditions favorable à une transmission homogène du stress,

permet de repérer des différences locales dans les condition de rétention

tout le long le canal radiculaire, et elles est utile pur comparer le potentiel de

rétention de différents ciments, comme il résulte de la première étude

présentée dans le Chapitre IV.

De toute façon, une correcte interprétation des résultats du test push out

demande la conscience de la part du chercheur que le frottement de

glissement contribue significativement à ce que l’on mesure comme force de

rétention. Selon les résultats des études par des ingénieurs à propos de

tests push out sur des fibres englobées dans des matrices différentes, la

contribution du frottement de glissement est très importante. Au frottement

de glissement on peut en effet attribuer un remarquable quote-part de la

force de rétention au cas de tenons de fibre cimentés avec des ciments

résineux self-etch et self-adhésive. Au niveau microscopique on ne peut pas

reconnaître les signes typiques d’un fiable lien micro-mécanique.

Le frottement peut en outre expliquer partiellement le succès clinique de

procédure de cimentation qui emploient le phosphate de zinc, aussi si ce

matériel n’a pas de capacité d’adhésion.9 Dans un récente expérimentation

de Sadek et coll.10, des tenons en fibre de verre cimentés avec du

phosphate de zinc ont mesuré une force au push out qu’on peut comparé à

celle de ciments résineux, employés en combinaison avec un adhésif self-

etch où total etch, et aussi supérieur à la force de rétention rejointe par un

ciment « self-adhésive ».

174

Au fin de quantifier la contribution du frottement de glissement dans le test

push out de tenons en fibre on a conduit la deuxième étude expérimentale

présenté dans le Chapitre IV. Cette recherche, basé sur des ciments

adhésives total-etch et self-etch, a démontré que la contribution du

frottement est considérable pour tout les deux matériaux, et particulièrement

consistant pour les ciments adhésifs total-etch, en relation hypothétique

avec la rugosité de la surface du ciment.

Ce dernier étude a approfondit la connaissance de quelques aspect

mécaniques, qui ont été jusqu’ici négligés, du test push out, qui est

beaucoup employé dans la recherche dans le champ des matériaux

dentaires. Au même temps l’étude expérimental a ouvert une nouvelle

perspective au fin de la correcte interprétation de tout ce qui, en recherche

dentaire, est communément, et peut être « naïvement » , considérée comme

une force adhésive.

Conclusions

6. La forme et l’épaisseur des échantillons pour le test microtensile

obtenu par l’email et la dentine ont une influence sur la force

d’adhésion qui est mesurée.

7. Le dessin à clepsydre est indiqué pour le test de force tensile de

tissues et de matériaux dentaire, si le façonnage de l’échantillon a

été fait en manière franchement précise et contrôlée, par exemple

en employant le MicroSpecimen Former.

8. Les échantillons à forme de bâtonné et avec une section

transversale de 1x1mm offrent la plus satisfaisante combinaison de

favorable distribution du stress et « maniabilité ». Ce dessin est

indiqué pour mesurer la force de adhésion à l’interface sur l’email et

la dentine coronale, et aussi parmi les matériaux dentaires.

9. Pendant le mesurage de la force de rétention de tenons adhésifs, le

test push out à « tranche mince » se propose comme le test le plus

simple et fiable si comparé au test microtensile.

175

10. Dans le test push out le frottement développé par le segment de

tenon poussé le long du canal de la racine donne une contribution

significative à la force de rétention du tenon.

Directions futures

Le test microtensile, au moins à la manière où on l’a jusqu’ici conduit

dans la plus grande partie des études, fournit seulement un mesurage

de la force statique d’un matériel ou d’une interface, définie comme la

capacité de résister à une charge constante pendant un temps limité.11

Toutefois, surtout au fin de estimer la performance clinique d’un matériel

à restauration où d’un adhésif dentaire on devrait évaluer à travers des

tests de fatigue même sa capacité de réagir à des variabilités cycliques

de température et de résister à des charges cycliques. Récemment

Frankenberger e coll.12 ont proposé une méthode pour appliquer des

cycles thermomécanique directement à l’interface adhésive un bâtonnet

de dentine et résine de 4 mm en hauteur et 2 mm en largeur. Les

bâtonnes qui ont survécu au cycle ont été après sectionnés en des

sticks de dimension plus petit, dont la force de adhésion interfaciale a

été mesurée avec la méthode microtensile.

De plus, pour déterminer la solidité et la durabilité d’une interface

adhésive, il n’y a pas seulement sa force tensile, mais aussi sa

résistance à la fracture (fracture toughness).9,13 Beaucoup d’adhésifs

qu’on emploi actuellement présentent une rigidité plutôt élevé et une

basse résistance à la fracture, qui résultent dans un « comportement

fragile » du materiel.11,14,15 On souhaiterait pourtant de augmenter la

capacité de l’interface adhésive de absorber de l’énergie avant de

arriver à la fracture. Au but de obtenir cela, la bio-mimétique a étudié

des organismes adhésifs de l’eau qui doivent leur favorable combinaison

de force et de vigueur à la présence de domaines protéiques capables

de modifier leur structure quand le stress augmente. Les chimiques et

les biologistes toutefois ont encore beaucoup de travail à développer au

fin de rendre ce composant protéique compatible avec les extrêmes

176

conditions ambiantes dans lesquelles actuellement on accomplit la

synthèse de composant chiques organiques.

Comme on a déjà vu à propos de l’étude de push out, il semblerait sage

pour le chercheur dentaire du future de se concentrer sur l’expérience et

les progrès dans le domaine de la science des matériaux. A travers ces

disciplines les chercheurs et les maison productrices des matériels

dentaires peuvent apprendre comment profiter des caractéristiques de

surface des matériels et du frottement de glissement développé au fin

de augmenter la capacité de rétention et de résistance au charge.

Un autre aspect qu’on a déjà étudié de la part des ingénieurs et qui

pourrait devenir intéressant aussi dans la science des matériels

dentaires concerne le stress restante. Des stress restante peuvent

donner origine dans quelques matériels naturels où artificiels des

différences en expansion thermique, des stress de fléchissement et de

la rigidité entre différentes régions du materiel.16 Les stress restants se

superposent au stress lié à la fonction du matériel et peuvent

conditionner négativement sa durabilité. Au contraire, un emploi

intelligent du stress restante peut améliorer la performance des

matériels fragiles dans la charge statique. A ce but, des études

d’ingénieurs et des savants des matériels dentaires sont en train de

définir des méthodes pour mesurer et prévoir les stress restants. Cette

méthode a déjà été appliqué avec succès dans la fabrication du verre

trempé à chaud et du ciment pre-stressé.14 Le relief dans l’odontologie

de ces connaissances des études d’ingénieurs est évident si l’on

considère que non seulement des matériels dentaires tels que des

résines composites et des composites renforcés par des fibres, mais

aussi l’email et la dentine sont des matériels à phase multiple, avec un

comportement mécanique fragile, et qui doivent résister aux charges,

idéalement pendant le temps d’une vie.

Récemment Gao e coll.17 ont souligné, au fin des propriétés

mécaniques, sur les dimensions nanométriques des unités structurales

élémentaires de bio composites tels que l’email, la dentine, l’os et la de

177

nacre de quelques coquilles. Les auteurs pensent que l’organisation de

plusieurs tissus biologiques durs en lamelles minérales séparé par des

couches plus souples de matrice protéique est ainsi construite par la

Nature au but de conférer une élevée résistance à la fracture à travers

des mécanismes de redistribution du stress et l’interruption de la

progression du crack. Plus précisément, la plus grande quote-part de la

charge est soutenue par les lamelles minérales, tandis que la matrice

protéique transfère le stress entre les lamelles sous forme de force de

coupure. En définitive, la résistance à la fracture du tissu dépend de la

force tensile de la partie minérale. Sur la base de l’équation numérique

qui se réfère au critère de Griffith, Gao e coll.17 sont arrivés à la

conclusion que la dimension nanométrique des lamelles minérales

permet une optimisation de leur résistance et de la tolérance des

défauts. Au niveau nanométrique, le matériel rejoint des niveaux de

résistance proches au cristal parfait. Au contraire, quand les dimensions

du minéral dépassent le niveau nanométrique, le matériel devient

sensible aux défauts intrinsèques et cède à cause de la concentration

du stress en correspondance de ces défauts. Les résultats de ces

études ont ouvert une nouvelle voie à suivre dans l’effort de produire en

laboratoire de nouveaux nano matériels doués de plus grandes

propriétés mécaniques. Même si cet étude ont souligné l’importance de

la résistance mécanique comme une propriété déterminante dans

l’organisation nano structurale des matériaux biologiques, toutefois les

auteurs ont aussi souligné que même les facteurs chimique jouent un

rôle important dans la formation et nucléation des cristaux minéraux.

N’importe où la recherche future veut nous conduire, la nature nous

donne chaque fois les enseignements les plus valables.

178

RESUMEN, DISCUSIÓN, CONCLUSIONES y DIRECCIONES FUTURAS

La prueba de microtension es una técnica versátil para valorar la

adhesión. Se han propuesto varias aplicaciones y diferentes variantes

de la prueba.

En la parte inicial del proyecto (Capítulo I), han sido tomados en

consideración algunos aspectos relativos a la preparación de los

especimenes y a la interpretación de los datos.

Llegamos a la conclusión de que la forma y el espesor en que los

especimenes conseguidos por esmalte y dentina son preparados

influyen en la fuerza de adhesión que es medida.

En cuánto a la forma del especimen, la técnica "trimming", que produce

especimenes en forma de reloj de arena, es apta para medir la maxima

fuerza tensil de tejidos y materiales dentarios, ya que hace que el estrés

sea concentrado sobre una superficie cuya área puede ser fácilmente

determinada.

Esto solo se logra si el equipo que hace los especimenes

(MicroSpecimen Former) es capaz de dejar una superficie adecuada. La

nueva version del Micro Specimen Former està completamente

automatizada, lo cual evita que se genere en el especimen el estrès que

se tenia al realizar movimentos manuales (comunicación personal del

Dr. Steve Armstrong).

No es posible asegurar un adecuado control del especimen cuando este

se hace a mano con una turbina. Ya que el especimen no queda

uniforme.

Por esta razón la elaboracìon del especimen a mano no parece ser el

método de elección cuando se prueban interfaces en particular para la

interfaz diente-adhesivo, pero, más en general, todas las interfaces en

que se espera que sean alcanzados niveles de fuerza de adhesion

relativamente bajos.

179

Esta afirmación es sustentada por el trabajo experimental indicado en el

Capítulo I y de egual manera por los datos científicos ya presentados en

la literatura.

Concierniente a las dimensiones del especimen, hay una tendencia en

la que los especimenes más sutiles miden fuerzas de adhesión más

elevada.

En una reciente investigación de El Zohairy y coll.1, pone en relaciòn

este fenómeno con el modo en que los especimenes se fijan al aparato

para el test, es decir de los lados del especimen.

El autor de esta tesis, sin embargo, está más convencido de que es la

más baja la densidad de los defectos intrínsecos del especimen,

permitiendo una distribución del estrés mas uniforme, la responsable de

la relación inversa entre la área transversal del especimen y la fuerza

tensil que es medida, como originariamente solicito por Griffith y más

tarde por Sano y coll.2

La confirmacion de esta teoría la encontramos en la prueba, con la

observación al microscopio, de que defectos estructurales están

efectivamente presentes en los especimenes por prueba de

microtension en la interfaz adhesiva, en el build-up de composite y

también dentro del sustrato dental; en cuánto a este último, el esmalte

resulta más predispuesto al desarrollo de defectos estructurales

(Capítulo I.1.1).

El área transversal de 1 x 1 mm ha demostrado brindar una uniformidad

en las pruebas de tensión y manejabilidad de los especimenes.

Los especimenes preparados con esta dimensión en forma de barrita,

han sido probados exitosamente al comparar las fuerzas adhesivas en

esmalte y dentina con primers autograbables.

Con respecto a los defectos intrínsecos del sustrato y/o el material

dental, estos, pueden ser responsables de las fracturas cohesivas de

los especimenes, que, aunque son menos frecuentes en las pruebas de

tensión y de corte convencional, pueden, en todo caso aparecer en la

prueba de microtensión.

180

Los investigadores aun no saben que interpretación darle a los

especímenes que fracasan cohesivamente, y aun es objeto de

discusión. Con respecto a esto, el autor sugiere que cuando se trata de

medir la adhesión de la interfaz, se eliminen las fracturas cohesivas, ya

que confunden los resultados de la prueba.

La manera en que las fracturas cohesivas son cuantificadas como

fuerza adhesiva por Reis y col.3 le parece arbitrario al autor de esta

tesis, ya que pone en discusión la interpretación que se da a las

fracturas prematuras.

Con relación a las fracturas prematuras, el autor, de acuerdo con

Nikolaenko y col.4 piensa que reportando el número de fallas que han

ocurrido antes de la prueba en cada grupo, es una manera honesta de

describir cuantitativamente algunos defectos y, por tanto, brinda una

información significativa de los objetivos de la comparación entre

grupos.

El autor piensa que las fracturas prematuras deberían ser excluidas del

análisis estadístico, sobre todo en el caso particular de que su inclusión

como "valores cero" modifique la distribución de los datos no normales.

Con lo que respecta a la interpretación estadística de los datos de

fuerza de adhesión de los especímenes obtenidos en los dientes, se

piensa que la variabilidad del origen de cada diente no debe tomarse a

la ligera.

Con respecto a esto, el autor no sugiere el empleo del test estadistico

ANOVA en valores repetidos5, ya que los especímenes de microtensión

no son evaluados varias veces en intervalos consecutivos.

Considerar el origen del diente como un factor aleatorio, parece más

apropiado.6,7 Por lo que el autor sugiere emplear el test ANOVA para

determinar las diferencias significativas en la fuerza de adhesión entre

dientes de un mismo grupo experimental, antes que agrupar a los

especimenes de microtensión de diferentes dientes.8 Además se puede

realizar un análisis previo sobre los especimenes de microtension, con

el objeto de determinar que el origen del diente no tenga una influencia

181

significativa sobre las fuerzas de adhesión medidas en los

especimenes.

Si se elige un procedimiento de preparación de los especimenes que

sea compatible con el objetivo de la prueba y este procedimiento se

hace con detenimeinto, el ensayo de microtensión ofrece un instrumento

confiable para medir la fuerza de adhesión interfacial de los materiales a

substratos dentales. Los Capítulos I e II mencionan varios ejemplos de

aplicación del test microtensile.

Cuando el test de microtensión se aplica para medir la fuerza de

adhesión en postes radiculares, las variantes en als técnicas presentan

limitaciones, las cuales se reflejan en el número de especímenes útiles

que se pueden conseguir y en la dispersión de los datos.

Cualquier acción de corte resulta demasiado agresiva para ser

soportada por las uniones débiles que existena a nivel de las interfaces

poste-cemento-dentina.

Es la comclusión del experimento presentado en el Capítulo II. En esta

misma investigación, se propone como alternativa a la microtensión, el

test push-out (desalojo) como método más practico y fiable para medir

la fuerza retentiva de un poste cementado.

La variente de la prueba descrita en al Capítulo IV, garantiza las

condiciones adecuadas para una homogénea transmisión del estrés,

permite localizar diferenciasd locales en as condiciones de retención a

lo largo del canal redicular, es útil para confrontar el potencial retentivo

de diferentes materiales cementados, tal como lo expone el primer

estudio presentado en el Capítulo IV.

De cualquier modo, una correcta interpretación de los resultados de la

prueba de desalojo, implica saber, por parte del investigador, que lo que

se mide como fuerza al desalojo, lleva implicita la fricción.

Según los resultados de estudios de Ingeniería sobre la prueba del

push-out (desalojo) de fibras englobadas en matrices de varios generos,

la contribución de la friccion rasante es completamente relevante. La

friccion rasante puede ser atribuida a una constante cantidad de fuerza

182

retentiva de postes de fibra cementados con cementos resinosos self-

etch y self-adhesive, lo cual no se puede observar a nivel microscópico,

ya que no son reconocibles las señales típicas de una confiable unión

micromecanica. Además, la fricción puede explicar, en parte el éxito

clínico en los procedimeintos de cementación con fosfato de zinc, a

pesar de la falta de capacidad adhesiva del mismo.9 De hecho, en un

reciente experimento de Sadek y coll.10, los postes de fibra de vidrio

cementados con fosfato de zinc han obtenido valores de fuerza al push-

out comparable a los cementos a base de resina, utilizados en

combinación con un adhesivo self-etch o total-etch, incluso obtienen

valores superiores a la fuerza retentiva alcanzada por un cemento self-

adhesive.

El objetivo de cuantificar la contribución de la fricción en la prueba al

desalojo de postes radiculares de fibra, se demuestra en el segundo

trabajo experimental del Capítulo IV. Esta investigacion, sobre cementos

adhesivos total-etch y self-etch, ha demostrado que la contribución de la

friccion esta presente en ambos materiales y es particularmente

consistente en el cemento adhesivo total-etch, existe una hipotética

relación con la aspereza de la superficie del cemento.

Este último estudio ha profundizado el conocimiento de algunos

aspectos mecánicos, que hasta ahora habian sido descuidados en el

test de push-out, el cual se emplea en la investigación en el campo de

los materiales dentarios. Al mismo tiempo, el trabajo experimental ha

abierto una nueva perspectiva para la correcta interpretación de todo lo

que en investigación odontológica es comúnmente, y quizás

"ingenuamente", considerado como fuerza adhesiva.

Conclusiones

1. La forma y el espesor de los especimenes para el test de

microtensión conseguidos en esmalte y dentina influyen en los valores

de fuerza de adhesión.

183

2. El diseño de reloj de arena es apto para la prueba de máxima fuerza

tensile de tejidos y materiales dentales, siempre y cuando, se siga la

metodología de una manera sumamente precisa y controlada, por

ejemplo, utilizando el MicroSpecimen Former.

3. Los especimenes en forma de barrita y con sección transversal de

1x1mm ofrecen la adecuada combinación de distribución del estrés y

"manejabilidad." Este diseño es indicado para medir la fuerza de

adhesión interfacial sobre esmalte y dentina coronal, además entre los

materiales dentarios.

4. Para determinar la fuerza retentiva de postes radiculares

cementados, la prueba push-out, se propone como lo prueba más

simple y confiable.

5. En la prueba push-out, la friccion desarrollada por el segmento de

poste empujado a lo largo del canal radicular contribuye

significativamente a la fuerza retentiva del poste.

Direcciones futuras

La prueba de microtension, por lo menos en el modo en que ha sido

conducido hasta ahora en la mayor parte de los estudios, sólo prevee

una medida de la fuerza estática de un material o una interfaz, definida

como la capacidad de resistir a una carga constante por breve tempo.11

Sin embargo, sobre todo para estimar el desempeño clínico de un

material de restauración o de un adhesivo dental, se debería valorar por

prueba de fatiga tambien su capacidad de reaccionar a variaciones

cíclicas de temperatura y de resistir a cargas cíclicas. Recientemente

Frankenberger y coll.12 han propuesto un método para aplicar

directamente ciclos termomecánicos a la interfaz adhesiva de barritas

de dentina y resina de 4 mm en altura y 2 mm en ancho. Las barritas

que han sobrevivido a los ciclos han sido por lo tanto seccionados en

barritas de dimensiones inferiores, cuya fuerza de adhesión interfacial

ha sido medida con el ensayo de microtension.

184

Además de determinar la solidez y durabilidad de una interfaz adhesiva,

no solamente su resistencia a la tracccion, también establece su

resistencia a la fractura.9,13 Muchos adhesivos actualmente en uso

presentan una rigidez bastante elevada y una baja resistencia a la

fractura, que dan en un "comportamiento frágil" del material.11,14,15 Sería

por tanto deseable incrementar la capacidad de la interfaz adhesiva de

absorber energía antes de llegar a fractura. Con el objetivo de conseguir

eso, la biomimetica ha estudiado organismos adhesivos acuáticos que

deben su favorable combinación de fuerza y robustez a la presencia de

dominós proteicos capaces de modificar su estructura al aumentar de el

estrés. Químicos y biólogos se dan a la tarea de desarrollar a para

devolver a estas partes proteinas compatibles con las extremas

condiciones ambientales en que actualmente se cumple la síntesis de

compuestos químicos organicos.

Como se ha observado por el estudio sobre el push-out, parece sabio

para el investigador dental del futuro dirigir puntualmente la atención a

la experiencia y a los progresos en el campo de la ciencia e ingeniería

de los materiales. De estas disciplinas los investigadores y las

empresas productoras de los materiales dentarios pueden aprender

cómo llevar ventaja de las características de superficie de los materiales

y de la friccion desarrollada para incrementar la capacidad retentiva y de

resistencia a las cargas.

Otro aspecto ya objeto de estudio en ingeniería y que también pudiera

volverse de interés en la ciencia de los materiales dentarios concierne el

estrés restante. El estrés restante pueden originar en los materiales a

fase múltiples, naturales o artificiales, las diferencias en expansión

térmica, estrés de hundimiento y rigidez entre muchas regiones del

material.16 Él estrés restante se agrega a aquellos allegados a la función

del material y puede influir negativamente en su durabilidad. De otra

manera, un empleo inteligente del estrés restante puede mejorar el

desempeño performance de los materiales frágiles en carga estática.

185

En este sentido los ingenieros y los científicos de los materiales

dentarios están trabajando en la definición de métodos para medir y

predecir los estrés restantes. Esta apoximaciòn ya ha sido aplicada con

éxito a la fabricación del vidrio templado con calor y del cemento

pretensado.14

La relevancia en el campo odontológico de estas adquisiciones de la

ingeniería es evidente si se considera que no solamente materiales

dentarios como resinas compuestas y composites reforzados por fibras,

sino también esmalte y dentina que son materiales que presentan

múltiples fases, con comportamiento mecánico frágil, y que son

llamados a desarrollar una función de resistencia a las cargas, ideados

para un determinado tiempo de una vida.

Recientemente Gao y coll.17 han llamado la atención sobre el objetivos

de las propiedades mecánicas y de las dimensiones nanometrichas de

las unidades estructurales elementales del biocomposite como esmalte,

dentina, hueso y el nácar de algunas conchas. Los autores opinan que

la organización de muchos tejidos biológicos duros en laminillas

minerales separadas por capas más blandas de matriz proteica esta

prevista por la naturaleza con el objetio de otorgar una elevada

resistencia a la fractura por mecanismos de re-distribución del estrés y

detención de progresión de las fracturas. Precisamente, la mayor cuota

de la carga es sustentada por las laminillas minerales, mientras que la

matriz proteica traslada el estrés entre las laminillas bajo forma de

fuerzas de corte. En resumen, la resistencia a la fractura del tejido

depende principalmente de la fuerza tensile de la parte mineral.

Sobre la base de ecuaciones numéricas que hacen referencia al criterio

de Griffith, Gao y coll.17 han llegado a la conclusión de que la dimensión

nanometrica de las laminillas minerales permitte una optimización de su

resistencia y tolerancia de los defectos.

A escala nanometrica el mineral alcanza niveles de resistencia

cercanos a los del cristal perfecto, en otras palabras la resistencia

teórica, independientemente de la presencia de defectos intrínsecos. De

186

otra manera, cuando las dimensiones del mineral superan la escalera

nanometrica, el material es sensible a los defectos intrínsecos y cede

por concentración del estrés en correspondencia de estos defectos. Los

resultados de estos estudios han abierto a los bioingenieros un nuevo

recorrido de seguir en su esfuerzo de producir en el laboratorio nuevos

nanomateriales dotados con superiores propiedades mecánicas.

Aunque este estudio han enfatizado la importancia de la resistencia

mecánica como propiedad determinante en la organización

nanoestructural de los materiales biológicos, los autores también han

subrayado como los factores químicos desempeñen un papel crucial en

la formación y nucleaciòn de los cristales minerales.

En todo sentido la futura investigacion nos conducirà, una cosa es

segura, que siempre es la naturaleza la que las enseñanzas más

valiosas.

187

References 1. El Zohairy AA, de Gee AJ, de Jager N, van Ruijven LJ, Feilzer AJ. The

influence of specimen attachment and dimension on microtensile bond

strength. J Dent Res 2004; 83: 420-424.

2. Sano H, Shono T, Sonoda H, Takatsu T, Ciucchi B, Horner JA, Pashley

DH. Relationship between surface area for adhesion and tensile bond

strength – Evaluation of a microtensile bond test. Dent Mater 1994; 10:

236-240.

3. Reis A, Loguercio AD, Azevedo CLN, Carvalho RM, Singer JM, Grande

RHM. Moisture spectrum of demineralized dentin for adhesive systems

with different solvent bases. J Adhes Dent 2003; 5: 183-192.

4. Nikolaenko SA, Lohbauer U, Roggendorf M, Petschelt A, Dasch W,

Frankenberger R. Influence of C-factor and layering technique on

microtensile bond strength to dentin. J Adhes Dent 2004; 20: 579-585.

5. Kurtz JS, Perdigão J, Geraldeli S, Hodges JS, Bowles WR. Bond

strengths of tooth-colored posts. Effect of sealer, dentin adhesive, and

root region. Amer J Dent 2003; 16: 31A-36A.

6. Loguercio AD, Uceda-Gomez N, Oliveira Carrilho MR, Reis A.

Influence of specimen size and regional variation on long-term resin-

dentin bond strength. Dent Mater 2004; in press.

7. De Munck J, Van Meerbeek B, Yoshida Y, Inoue S, Vargas M, Suzuki

K, Lambrechts P, Vanherle G. Four-year water degradation of total-etch

adhesives bonded to dentin. J Dent Res 2003; 82: 136-140.

8. Bouillaguet S, Troesch S, Wataha JC, Meyer JM, Pashley DH.

Microtensile bond strength between adhesive cements and root canal

dentin. Dent Mat 2003; 19: 199-205.

9. Attar N, Tam L, McComb D. Mechanical and physical properties of

contemporary dental luting agents. J Prosthet Dent 2003; 89: 127-134.

10. Sadek FT, Cury AH, Goracci C, Tay FR, Ferrari M. Push-out strength of

glass fiber posts luted with zinc phosphate and several resin cements.

Unpublished results.

188

11. Van Noort. An introduction to Dental Materials, 2nd edition. Mosby,

2002.

12. Frankenberger R, Pashley DH, Reich SM, Lohbauer U, Petschelt A,

Tay FR. Characterisation of resin-dentin interfaces by compressive

cyclic loading. Biomaterials, in press, available on line: September 2004.

13. Tay FR, Pashley DH. Dental adhesives of the future. J Adhes Dent

2002; 4: 91-103.

14. Armstrong SR, Keller JC, Boyer DB. Mode of failure in the dentin-

adhesive resin-resin composite bonded joint as determined by strength-

based (µTBS) and fracture-based (CNSB) mechanical testing. Dent

Mater 2001; 17: 201-210.

15. Lin CP, Douglas WH. Failure mechanisms at the human dentin-resin

interface: a fracture mechanics approach. J Biomech 2004; August:

1037-1047.

16. Withers PJ, Bhadeshia HKDH (2001). Residual stress. Part I –

measurement techniques. Materials Science and Technology 17 355-

365.

17. Gao H, Baohua J, Jäger IL, Arzt E, Fratzl P. Materials become

insensitive to flaws at nanoscale: Lessons from nature. PNAS

(Proceedings of the National Academy of Sciences of the United States

of America) 2003; 100: 5597-5600.

189

ACKNOWLEDGEMENTS This thesis is respectfully submitted to Prof. Piero Tosi, Rector of the

University of Siena, to Prof. Alberto Auteri, Dean of the Faculty of Medicine,

University of Siena, to Prof. Egidio Bertelli, vice-Dean of the Faculty of

Medicine and Director of the Department of Dental Science, and to Prof.

Marco Ferrari, Pro-Rector for International Affairs and President of Dental

School, University of Siena.

The illustrations of the microtensile specimen preparation procedures are

professional work of Paulo Santos, partially supported by NAPEM-FOUSP

(Nucleo de Apoio à Pesquisa em Materiais Dentarios-Faculdade de

Odontologia - USP, Brazil).

I gratefully acknowledge Dr. Vanda Grandini for the translation in French, Dr.

Francesca Monticelli and Ornella Raffaelli for the translation in Spanish of

the “Summary, general discussion, conclusions, and future directions”

section.

My thanks go to my classmates Dr. Andrea Fabianelli and Dr. Simone

Grandini for their cooperation and support, and to Dr. Francesca Monticelli,

the companion of many “school trips”.

I am sincerely grateful to Prof. Paulo Eduardo Capel Cardoso and Dr.

Fernanda Tranchesi Sadek, valuable co-workers and special friends.

I am thankful to the members of the Committee, Prof. Piero Balleri, Prof.

Egidio Bertelli, Prof. Carel Davidson, Prof. Michel Goldberg, Prof. Manuel

Toledano for their review of the thesis and their advice.

I wish to express admiration and gratitude to my Masters: Prof. Franklin Tay,

inexhaustible source of ideas and help, and Prof. Marco Ferrari, a shining

guiding light for many years.

I want to thank my family for bearing with me.

190

References Akkayan B, Gulmez T. Resistance to fracture of endodontically

treated teeth restored with different post systems. J Prosthet Dent 2002;

87: 431-437.

Albuquerque RC, Polleto LT, Fontana RH, Cimini CA. Stress

analysis of an upper central incisor restored with different posts. J Oral

Rehab 2003; 30: 936-943.

Alster D, Feilzer AJ, de Gee AJ, Davidson CL. Polymerization

contraction stress in thin resin composite layers as a function of layer

thickness. Dent Mater 1997; 13: 146-150.

Ari H, Yasar E, Belli S. Effects of NaOCl on bond strength of resin

cement to root canal dentin. J Endod 2003; 29: 248-251.

Armstrong SR, Boyer DB, Keller JC. Microtensile bond strength

testing and failure analysis of two dentin adhesives. Dent Mater 1998;

14: 44-50.

Armstrong SR, Keller JC, Boyer DB. Mode of failure in the dentin-

adhesive resin-resin composite bonded joint as determined by strength-

based (TBS) and fracture-based (CNSB) mechanical testing. J Dent

2001; 17: 201-210.

Armstrong SR, Keller JC, Boyer DB. The influence of water storage

and C-factor on the dentin-resin composite microtensile bond strength

and debond pathway utilizing a filled and unfilled adhesive resin. Dent

Mater 2001; 17: 268-276.

Armstrong S, Vargas MA, Fang Q, Laffoon JE. Microtensile bond

strength of a total-etch 3-step, total-etch 2-step, self-etch 2-step, and a

self-etch 1-step dentin bonding system through 15-month water storage.

J Adhes Dent 2003;5:47-56.

Asmussen E, Peutzfeldt A, Heitmann T. Stiffness, elastic limit, and

strength of newer types of endodontics posts. J Dent 1999; 27: 275-278.

Attar N, Tam L, McComb D. Mechanical and physical properties of

contemporary dental luting agents. J Prosthet Dent 2003; 89: 127-134.

191

Augulis L, Tamulevičius S, Augulos R, Bonneville J, Goudeau P,

Templier C. Electronic speckle pattern interferometry for mechanical

testing of thin films. Optics and lasers in engineering 2004; 42: 1-8.

Bachicha WS, DiFiore PM, Miller DA, Lautenschlager EP, Pashley

DH. Microleakage of endodontically treated teeth restored with posts. J

Endodon 1998; 24: 703-708.

Barghi N, Berry T, Chung K. Effects of timing and heat treatment of

silanated porcelain on the bond strength. J Oral Rehab 2000; 27: 407-

412.

Berekally T. Contemporary perspectives on post-core systems. Aust

Endod J 2003; 29:120-127.

Berzins A, Summer DR. Implant pushout and pullout tests. In: An

YH, Draughn RA, eds. Mechanical testing of bone and the bone–implant

interface. Boca Raton, FL: CRC Press LCC; 1999 p.463-476.

Bini F, Marinozzi A, Marinozzi F, Patanè F. Microtensile

measurements of single trabeculae stiffness in human femur. J Biomech

2002; 35: 1515-1519.

Boschian Pest L, Cavalli G, Bertani P, Gagliani M. Adhesive post-

endodontic restoration with fiber posts: push-out tests and SEM

observations. Dent Mater 2002; 18: 596-602.

Bouillaguet S, Ciucchi B, Jacoby T, Wataha JC and Pashley D.

Bonding characteristics to dentin walls of Class II cavities, in vitro. Dent

Mater 2001; 17: 316-321.

Bouillaguet S, Gysi P, Wataha JC, Ciucchi B, Cattani M, Godin C, et

al. Bond strength of composite to dentin using conventional, one-step,

and self-etching adhesive systems. J Dent 2001; 29: 55-61.

Bouillaguet S, Troesch S, Wataha JC, Krejci I, Meyer JM, Pashley

DH. Microtensile bond strength between adhesive cements and root

canal dentin. Dent Mat 2003; 19: 199-205.

Brackett WW, Covey DA, St Germain HA, Jr. One-year clinical

performance of a self-etching adhesive in class V resin composites

cured by two methods. Oper Dent 2002; 27: 218-222.

192

Burrow MF, Tagami J, Negishi T, Kikaido T, Hosoda H. Early tensile

bond strengths of several enamel and dentin bonding systems. J Dent

Res 1994; 73: 522-528.

Cardoso PE, Sadek FT, Goracci C, Ferrari M. Adhesion testing with

the microtensile method: effects of dental substrate and adhesive

system on bond strength measurements. J Adhes Dent 2002; 4: 291-

297.

Cardoso PEC, Braga RR and Carrilho MRO. Evaluation of micro-

tensile, shear and tensile tests determining the bond strength of three

adhesive systems. Dent Mater 1998; 394-398.

Cardoso PEC, Mallmann A, Burmann PA. Micro-tensile of self-

etching primer adhesive systems in enamel and dentin [abstract 12]. J

Dent Res 2001; 80: 61.

Carvalho RM, Pegoraro TA, Tay FR, Pegoraro LF, Silva NR,

Pashley DH. Adhesive permeability affects coupling of resin cements

that utilise self-etching primers to dentine. J Dent 2004; 32: 55-65.

Carvalho RM, Pereira JC, Yoshiyama M, Pashley DH. A review of

polymerization contraction: the influence of stress development versus

stress relief. Oper Dent 1996;21:17-24.

Carvalho RM, Santiago SL, Fernandes CAO, Suh B, Pashley DH.

Effects of prism orientation on tensile strength of enamel. J Adhes Dent

2000; 2: 251-257.

Chai YS, Mai Y. New analysis on the fiber push-out problem with

interface roughness and thermal residual stress. J Mater Sci 2001; 36:

2095-2104.

Chan KM, Tay FR, King NM, Imazato S, Pashley DH. Bonding of

mild self-etching primers/adhesives to dentin with thick smear layers. Am

J Dent 2003; 16: 340-346.

Chandra N, Ananth CR. Analysis of interfacial behavior in MMCs

and IMCs using thin slice push-out tests. Compos Sci Technol 1995; 54:

87-100.

193

Chandra N, Ghonem H. Interfacial mechanics of push-out tests:

theory and experiments. Compos Part A: Appl Sci Manuf 2001; 32: 575-

584.

Chersoni S, Suppa P, Grandini S, Goracci C, Monticelli F, Yiu C,

Huang C, Prati C, Breschi L, Ferrari M, Pashley DH, Tay FR. In vivo and

in vitro permeability of one-step self-etch adhesives. J Dent Res 2004;

83: 459-464.

Dallari A, Rovatti L. Six years of in vitro/in vivo experience with

Composipost. Compend Cont Edu Dent Suppl 1998; 20: S57-S63.

Dauvillier BS, Feilzer AJ, de Gee AJ, Davidson CL. Visco-elastic

parameters of dental restorative materials during setting. J Dent Res

2000; 79: 818-823.

Davidson CL, de Gee AJ, Feilzer A. The competition between the

composite-dentin bond strength and the polymerization contraction

stress. J Dent Res 1984; 63: 1396-1399.

De Hoff PH, Anusavice KJ, Wang Z. Three-dimensional finite

element analysis of the shear bond test. Dent Mater 1995; 11: 126-131.

De Jager N, Pallav P, Feilzer AJ. The apparent increase of the

Young's modulus in thin cement layers. Dent Mater 2004; 20: 457-462.

De Munck J, Van Meerbeek B, Yoshida Y, Inoue S, Vargas M,

Suzuki K, Lambrechts P, Vanherle G. Four-year water degradation of

total-etch adhesives bonded to dentin. J Dent Res 2003; 82: 136-140.

Debnath S, Wunder SL, MCCool JI and Baran GR. Silane treatment

effects on glass/interfacial shear strengths. Dent Mater, 2003; 19: 441-

448.

Della Bona A, Anusavice KJ, Shen C. Microtensile strength of

composite bonded to hot-pressed ceramics. J Adhes Dent 2000; 2: 305-

313.

Dhert WJA, Jansen JA. The validity of a single pushout test. In: An

YH, Draughn RA, eds. Mechanical testing of bone and the bone–implant

interface. Boca Raton, FL: CRC Press LCC, 1999 p.477-488.

194

Dong CC, McComb D, Anderson JD, Tam LE. Effect of mode of

polymerization of bonding agent on shear bond strength of autocured

resin composite luting cements. J Can Dent Ass 2003; 69: 229-234.

Drummond JL. In vitro evaluation of endodontic posts. Am J Dent

2000; 13: 5B-8B.

Erdemir A, Ari H, Gungunes H, Belli S. Effect of medications for root

canal treatment on bonding to root canal dentin. J Endod 2004; 30: 113-

116.

Erickson RL, Glasspoole, Retief DH. Influence of test parameters on

dentin bonding strength measurements [abstract 1543]. J Dent Res

1989; 68: 374.

Feilzer AJ, De Gee AJ, Davidson CL. Increased wall-to-wall curing

contraction in thin bonded resin layers. J Dent Res 1989;68:48-50.

Ferrari M, Goracci C, Monticelli F, Sadek FT, Cardoso PEC.

Adhesion testing with the microtensile method: effects of dental

substrate and adhesive system on bond strength measurements. J

Adhes Dent 2002; 4: 291-297.

Ferrari M, Goracci C, Sadek F, Cardoso PEC. Microtensile bond

strength tests: scanning electron microscopy evaluation of sample

integrity before testing. Eur J Oral Sci 2002; 110: 385-391.

Ferrari M, Grandini S, Simonetti M, Monticelli F, Goracci C. Influence

of a microbrush on bonding fiber posts into root canals under clinical

conditions. Oral Surg, Oral Med, Oral Path, Oral Rad, and Endod, 2002;

94: 627-631.

Ferrari M, Mannocci F, Vichi A, Cagidiaco MC, Mjör IA. Bonding to

root canal: structural characteristics of the substrate. Am J Dent 2000;

13: 255-260.

Ferrari M, Vichi A, García-Godoy F. A retrospective study of fiber-

reinforced epoxy resin posts vs cast posts and cores: a four year recall.

Am J Dent 2000; 13: B9-B14.

195

Ferrari M, Vichi A, Grandini S, Goracci C. Efficacy of a self-curing

adhesive-resin cement system on luting glass-fiber posts into root

canals: an SEM investigation. Int J Prosthod, 2001; 14: 543-549.

Ferrari M, Vichi A, Grandini S. Efficacy of different adhesive

techniques on bonding to root canal walls: an SEM investigation. Dent

Mater 2001; 17: 422-429.

Fogel HM, Marshall FJ, Pashley DH. Effects of distance from the

pulp and thickness on the hydraulic conductance of human radicular

dentin. J Dent Res 1988; 67: 1381-1385.

Foxton RM, Nakajima M, Tagami J, Miura H. Bonding of photo and

dual-cure adhesives to root canal dentin. Oper Dent 2003; 28: 543-551.

Frankenberger R, Pashley DH, Reich SM, Lohbauer U, Petschelt A,

Tay FR. Characterisation of resin-dentin interfaces by compressive

cyclic loading. Biomaterials, in press, available on line: September 2004.

Fredriksson M, Astback J, Pamenius M, Arvidson K. A retrospective

study on 236 patients with teeth restored by carbon fiber-reinforced

epoxy resin posts. J Prosthet Dent 1998; 80: 151-157.

Gallo JR, Miller T, Xu X, Burgess JO. In vitro evaluation of the

retention of composite fiber and stainless steel posts. J Prosthod 2002;

11: 25-29.

Gao H, Baohua J, Jäger IL, Arzt E, Fratzl P. Materials become

insensitive to flaws at nanoscale: Lessons from nature. PNAS

(Proceedings of the National Academy of Sciences of the United States

of America) 2003; 100: 5597-5600.

Garcia Varela S, Bravos Rabade L, Rivas Lombardero P, Linares

Sixto J, Gonzalez Bahillo J, Ahn Park S. In vitro study of endodontic post

cementation protocols that use resin cements. J Prosthet Dent 2003; 89:

146-153.

Gaston BA, West LA, Liewehr FR, Fernandes C, Pashley DH.

Evaluation of regional bond strength of resin cement to endodontic

surfaces. J Endod 2001; 27: 321-324.

196

Goracci C, Ferrari M, Grandini S, Monticelli F, Tay FR. Bonding of a

self-adhesive resin cement to dental hard tissues. J Adhes Dent 2004, in

press.

Goracci C, Monticelli F, Tavares AU, Sadek F, Cardoso PEC and

Ferrari M. The adhesion between fiber posts and composite resin cores:

microtensile bond strength of different combinations of materials. J Dent

Res, 2003; 82: B170, Abstract No. 1268.

Goracci C, Sadek FT, Monticelli F, Cardoso PEC, Ferrari M.

Influence of substrate, shape, and thickness on microtensile specimens’

structural integrity and their measured bond strengths. Dent Mater 2004;

20: 643-654.

Goracci C, Tavares AU, Fabianelli A, Monticelli F, Raffaelli O,

Cardoso PC, Tay FR, Ferrari M. The adhesion between fiber posts and

root canal walls: comparison between microtensile and push-out bond

strength measurements. Eur J Oral Sci 2004; 112: 353-361.

Gordan VV, Vargas MA, Cobb DS, Denehy GE. Evaluation of

adhesive systems using acidic primers. Am J Dent 1997; 10: 219-223.

Guignes P, Faure J, Maurette A. Relationship between endodontic

preparations and human dentin permeability measured in situ. J Endod

1996; 22: 60-67.

Hagge MS, Wong RD, Lindemuth JS. Effect of dowel space

preparation and composite cement thickness on retention of a

prefabricated dowel. J Prosthod 2002; 11: 19-24.

Hannig M, Reinhardt KJ, Bott B. Self-etching primer vs phosphoric

acid: an alternative concept for composite-to-enamel bonding. Oper

Dent 1999; 24: 172-180.

Hara AT, Amaral CM, Pimenta LA, Sinhoreti MA. Shear bond

strength of hydrophilic adhesive systems to enamel. Am J Dent 1999;

12: 181-184.

Hulley SB, Cummings SR. Planning the measurements: precision

and accuracy. In Hulley SB, Cummings SR. Designing clinical research.

Baltimore, Williams and Wilkins, 1988.

197

Ibarra G, Vargas MA, Armstrong SR, Cobb DS. Microtensile bond

strength of self-etching adhesives to ground and unground enamel. J

Adhes Dent 2002; 4: 115-124.

Inoue S, Vargas MA, Abe Y, Lambrechts P, Vanherle G, Sano H,

Van Meerbeek B. Microtensile bond strength of eleven contemporary

adhesives to dentin. J Adhes Dent 2001; 3: 237-245.

Ishida H. Structural gradient in the silane coupling agent layers and

its influence on the mechanical and physical properties of composites.

In: Ishida H, Kumar G, editors. Molecular characterization of composite

interfaces. New York: Plenum Press; 1985 p.25-50.

Kallas M, Koss D, Hahn H, Hellmann J. Interfacial stress state

present in a "thin-slice" fiber push-out test. J Mater Sci 1992; 27: 3821-

3826.

Kallio TT, Lastumäki TM, Vallittu PK. Bonding of a restorative and

veneering composite resin to some polymeric composites. Dent Mater,

2001; 17: 80-86.

Kalton AF, Howard SJ, Janczak-Rusch J, Clyne TW. Measurement

of interfacial fracture energy by single fibre push-out testing and its

application to the titanium-silicon carbide system. Acta Mater 1998; 46:

3175-3189.

Kim DG, Brunski JB, Nicolella DP. Microstructural-level strain-fields

during a microtensile test of bone at a bone implant interface.

Proceedings of the American Society of Mechanical engineers (ASME)

2001, Snowbird, Utah.

Kompella MK, Lambros J. Micromechanical characterization of

cellulose fibers. Polymer Testing 2002; 21: 523-530.

Kurtz JS, Perdigão J, Geraldeli S, Hodges JS, Bowles WR. Bond

strengths of tooth-colored posts. Effect of sealer, dentin adhesive, and

root region. Amer J Dent 2003; 16: 31A-36A..

Lee J, Ju J, Jang J, Kim W, Kwon D. Weld crack assessments in

API X65 pipeline: failure assessment diagrams with variations in

198

representative mechanical properties. Materials Science and

Engineering A 2004; 373: 122-130.

Li ZH, Bi XP, Lambros J, Geubelle PH. Dynamic fiber debonding

and frictional push-out in model composite systems: experimental

observations. Exp Mech 2002; 42:417-425.

Lin CP, Douglas WH. Failure mechanisms at the human dentin-resin

interface: a fracture mechanics approach. J Biomech 2004; August:

1037-1047.

Lin G, Geubelle PH, Sottos NR. Simulation of fiber debonding with

friction in a model composite pushout test. Int J Solids Struct 2001; 38:

8547-8562.

Loguercio AD, Uceda-Gomez N, Oliveira Carrilho MR, Reis A.

Influence of specimen size and regional variation on long-term resin-

dentin bond strength. Dent Mater 2004; in press.

Mak YF, Lai SC, Cheung GS, Chan AW, Tay FR, Pashley DH.

Microtensile bond testing of resin cements to dentin and an indirect resin

composite. Dent Mater 2002; 18: 609-621.

Manhart J, Chen HY, Mehl A, Weber K, Hickel R. Marginal quality

and microleakage of adhesive Class V restorations. J Dent 2001; 29:

123-130.

Mannocci F, Sherriff M, Ferrari M, Watson TF. Microtensile bond

strength and confocal microscopy of dental adhesives bonded to root

canal dentin. Am J Dent 2001; 14: 200-204.

Matinlinna JP, Lassila LVJ, Özcan M, Yli-Urpo A, Vallittu P. An

introduction to silanes and their clinical applications in dentistry. Int J

Prosthod 2004; 17: 155-164.

Mesfar W, Shirazi-Adl A, Dammak M. Modeling of biomedical

interfaces with nonlinear friction properties. Biomed Mater Eng 2003; 13:

91-101.

Mitchell CA, Orr JF, Connor KN, Magill JPG, Maguire GR.

Comparative study of four glass ionomer luting cements during post pull-

out tests. Dent Mater 1994; 10: 88-89.

199

Monticelli F, Goracci C, Ferrari M. Micromorphology of the fiber post-

resin core unit: a scanning electron microscopy evaluation. Dent Mater

2004; 20: 176-183.

Monticelli F, Grandini S, Goracci C and Ferrari M. Clinical behavior

of translucent fiber posts: A two-year prospective study. Int J Prosthod,

2003; 16: 593-596.

Morris MD, Lee KW, Agee KA, Bouillaguet S, Pashley DH. Effects of

sodium hypochlorite and RC-prep on bond strengths of resin cement to

endodontic surfaces. J Endod 2001; 27: 753-757.

Nakajima M, Sano H, Burrow MF, Tagami J, Yoshiyama M, Ebisu S,

Ciucchi B, Russel CM, Pashley DH. Tensile bond strength and SEM

evaluation of caries-affected dentin using adhesives. J Dent Res 1995;

74: 1679-1688.

Ngoh EC, Pashley DH, Loushine RJ, Weller N, Kimbrough F. Effects

of eugenol on resin bond strengths to root canal dentin. J Endod 2001;

27: 411-414.

Nikaido T, Takano Y, Sasafuchi Y, Burrow MF, Tagami J. Bond

strengths to endodontically treated teeth. Am J Dent 1999; 12: 177-180.

Nikolaenko SA, Lohbauer U, Roggendorf M, Petschelt A, Dasch W,

Frankenberger R. Influence of C-factor and layering technique on

microtensile bond strength to dentin. J Adhes Dent 2004; 20: 579-585.

Nuňo N, Amabili M, Groppetti R, Rossi A. Static coefficient of friction

between Ti-6Al-4V and PMMA for cemented hip and knee implants. J

Biomed Mater Res 2001; 59:191-200.

Ogata M, Harada N, Yamaguchi S, Nakajima M, Pereira PN, Tagami

J. Effects of different burs on dentin bond strengths of self-etching

primer bonding systems. Oper Dent 2001; 26: 375-382.

Ogata M, Harada N, Yamaguchi S, Nakajima M, Tagami J. Effect of

self-etching primer vs phosphoric acid etchant on bonding to bur-

prepared dentin. Oper Dent 2002; 27: 447-454.

Papa J, Cain C, Messer HH. Moisture content of vital vs

endodontically treated teeth. Endod Dental Traumatol 1994; 10: 91-93.

200

Pape PG, Plueddemann EP. Methods for improving the performance

of silane coupling agents. J Adhesion Sci Technol, 1991; 5: 831-842.

Pashley DH, Carvalho RM, Sano H, Nakajima M, Yoshiyama M,

Shono Y, Fernandes CA, Tay FR. The microtensile bond test: A review.

J Adhes Dent, 1999; 1: 299-309.

Pashley DH, Ciucchi B, Sano H, Yoshiyama M, Carvalho RM.

Adhesion testing of dentin bonding agents. A review. Dent Mater 1995;

11: 117-125.

Pashley DH, Tao L, Boyd L, King GE, Horner JA. Scanning electron

microscopy of the substructure of smear layers in human dentine. Arch

Oral Biol 1998; 33: 265-270.

Pashley DH, Tay FR. Aggressiveness of contemporary self-etching

adhesives. Part II: etching effects on unground enamel. Dent Mater

2001;17:430-444.

Pashley EL, Agee KA, Pashley DH, Tay FR. Effects of one versus

two applications of an unfilled, all-in-one adhesive on dentine bonding. J

Dent 2002; 30: 83-90.

Patierno JM, Rueggeberg FA, Anderson RW, Weller RN, Pashley

DH. Push-out and SEM evaluation of resin composite bonded to internal

cervical dentin. Endod Dent Traumatol 1996; 12: 227-236.

Perdigão J, Geraldeli S, Carmo ARP, Dutra HR. In vivo influence of

residual moisture on microtensile bond strengths of one-bottle

adhesives. J Esthet Rest Dent 2002; 14: 31-38.

Perdigão J, Lopes L, Lambrechts P, Leitao J, Van Meerbeek B,

Vanherle G. Effects of a self-etching primer on enamel shear bond

strengths and SEM morphology. Am J Dent 1997; 10: 141-146.

Pereira PNR, Okuda M, Nakajima M, Sano H, Tagami J, Pashley

DH. Relationship between bond strength and nanoleakage: Evaluation

of a new assessment method. Am J Dent 2001; 14: 100-104.

Pfeifer C, Shih D, Braga RR. Compatibility of dental adhesives and

dual-cure cements. Am J Dent 2003; 16: 235-238.

201

Phrukkanon S, Burrow MF, Tyas MJ. The influence of cross-

sectional shape and surface area on the microtensile bond test. Dent

Mater 1998; 14: 212-221.

Plueddemann EP. Silane coupling agents. New York: Plenum Press,

1991.

Prisco D, De Santis R, Mollica F, Ambrosio L, Rengo S, Nicolais L.

Fiber post adhesion to resin luting cements in the restoration of

endodontically-treated teeth. Oper Dent 2003; 28: 515-521.

Purton DG, Love RM. Rigidity and retention of carbon fibre versus

stainless steel root canal post. Int Endod J 1996; 29: 262-265.

Qualthrough AJ, Chandler NP, Purton DG. A comparison of the

retention of tooth-colored posts. Quint Int 2003; 34: 199-201.

Reis A, Loguercio AD, Azevedo CLN, Carvalho RM, Singer JM,

Grande RHM. Moisture spectrum of demineralized dentin for adhesive

systems with different solvent bases. J Adhes Dent 2003; 5: 183-192.

Sadek FT, Cury AH, Goracci C, Tay FR, Ferrari M. Push-out

strength of glass fiber posts luted with zinc phosphate and several resin

cements. Unpublished results.

Sahafi A, Peutzfeldt A, Asmussen E and Gotfredsen K. Bond

strength of resin cement to surface-treated posts of titanium alloy, glass

fiber, and zirconia, and to dentin. J Adhes Dent, 2003; 5: 153-162.

Sanders JE, Nicholson BS, Mitchell SB, Ledger RE. Polymer

microfiber mechanical properties: a system for assessment and

investigation of the link with fibrous capsule formation. J Biomed Mater

Res Part A 2003; 67: 1412-1416.

Sano H, Ciucchi B, Mathhews WG, Pashley DH. Tensile properties

of mineralized and demineralized human and bovine dentin. J Dent Res

1994; 73: 1205-1211.

Sano H, Shono T, Sonoda H, Takatsu T, Ciucchi B, Horner JA,

Pashley DH. Relationship between surface area for adhesion and tensile

bond strength – Evaluation of a microtensile bond test. Dent Mater 1994;

10: 236-240.

202

Schreiner RF, Chappell RP, Glaros AG, Eick JD. Microtensile testing

of dentin adhesives. Dent Mater 1998; 14: 194-201.

Schwartz RS, Robbins JW. Post placement and restoration of

endodontically treated teeth: a literature review. J Endodon

2004;30:289-301.

Sen D, Poyrazoglu E, Tuncelli B. The retentive effects of pre-

fabricated posts by luting cements. J Oral Rehabil 2004; 31: 585-589.

Serafino C, Gallina G, Cumbo E, Ferrari M. Surface debris of canal

walls after post space preparation in endodontically treated teeth: a

scanning electron microscopic study. Oral Surg Oral Med Oral Path Oral

Rad and Endod 2004; 97: 381-387.

Sharpe WN. Tensile testing at the micrometer scale: opportunities in

experimental mechanics. Experimental Mechanics 2003; 43: 228-237.

Shi XQ, Wang ZP, Pang HLJ, Zhang XR. Investigation of effect of

temperature and strain rate on mechanical properties of underfill

material by use of microtensile specimens. Polymer Testing 2002; 21:

725-733.

Shirazi-Adl A, Forcione A. Finite element stress analysis of a push-

out test. Part II: Free interface with nonlinear friction properties. J

Biomech Eng 1992;114: 155-161.

Shono Y, Ogawa T, Terashita M, Carvalho RM, Pashley DH.

Regional measurement of resin-dentin bonding as an array. J Dent Res

1999; 78: 669-705.

Shono Y, Terashita M, Pashley EL, Brewer PD, Pashley DH. Effects

of surface area on resin-enamel tensile bond strength. Dent Mater 1997;

13: 290-296.

Son D, Kim J, Lim TW, Kwon D. Evaluation of fatigue strength of

LIGA nickel film by microtensile tests. Scripta Materialia 2004; 50: 1265-

1269.

Son D, Kim J, Lim TW, Kwon D. Evaluation of fracture properties of

silicon by combining resonance frequency and microtensile methods.

Thin Solid Films 2004, in press.

203

Sorensen JA and Engelman MJ. Ferrule design and fracture

resistance of endodontically treated teeth. J Prosthet Dent, 1990; 63:

529-536.

Sudsangiam S and Van Noort R. Do dentin bond strength tests

serve a useful purpose? J Adhes Dent 1999; 1: 57-67.

Suh BI, Feng L, Pashley DH, Tay FR. Factors contributing to the

incompatibility between simplified-step adhesives and chemically-cured

or dual-cured composites. Part III. Effect of acidic resin monomers. J

Adhes Dent 2003; 5: 267-282.

Tay F, King NM, Suh BI, Pashley DH. Effect of delayed activation of

light-cured resin composites on bonding of all-in-one adhesives. J Adhes

Dent 2001; 3: 207-225.

Tay FR, Carvalho R, Sano H, Pashley DH. Effect of smear layers on

the bonding of a self-etching primer to dentin. J Adhes Dent 2000; 2: 99-

116.

Tay FR, King NM, Chan KM, Pashley DH. How can nanoleakage

occur in self-etching adhesive systems that demineralize and infiltrate

simultaneously? J Adhes Dent 2002; 4: 255-269.

Tay FR, Moulding KM, Pashley DH. Distribution of nanofillers from a

simplified-step adhesive in acid conditioned dentin. J Adhes Dent 1999;

1: 103-117.

Tay FR, Pashley DH, Suh BI, Carvalho RM, Itthagarun A. Single-

step adhesives are permeable membranes. J Dent 2002; 30: 371-382.

Tay FR, Pashley DH, Yoshiyama M. Two modes of nanoleakage

expression in single-step adhesives. J Dent Res 2002; 81: 472-476.

Tay FR, Pashley DH. Aggressiveness of contemporary self-etching

systems. I: depth of penetration beyond dentin smear layers. Dent Mat

2001; 17: 296-308.

Tay FR, Pashley DH. Dental adhesives of the future. J Adhes Dent

2002; 4: 91-103.

204

Tay FR, Smales RJ, Ngo H, Wei SHY, Pashley DH. Effect of

different conditioning protocols on adhesion of a GIC to dentin. J Adhes

Dent 2001; 3: 153-166.

Väkiparta TM, Yli-Urpo A, Vallittu PK. Flexural properties of glass

fiber reinforced composite with multiphase biopolymer matrix. J Mat Sci

Mater Med 2004, 2004; 15: 7-11.

Vallittu PK, Kurunmaki H. Bond strength of fibre-reinforced

composite to the metal surface. J Oral Rehab 2003; 30: 887-892.

Van Meerbeek B, Vargas M, Inoue S, Yoshida Y, Peumans M,

Lambrechts P, Vanherle G. Adhesives and cements to preserve

preservation dentistry. Oper Dent 2001; (suppl 6): 119-144.

Van Noort R, Cardew GE, Howard IC and Noroozi S. The effect of

local interfacial geometry on the measurement of the tensile bond

strength to dentin. J Dent Res 1991; 70: 889-93.

Van Noort R, Noroozi S, Howard IC, Cardew G. A critique of bond

strength measurements. J Dent 1989; 17: 61-67.

Van Noort. An introduction to Dental Materials, 2nd edition. Mosby,

2002.

Verdonschot N, Huiskes R. Subsidence of THA stems due to acrylic

cement creep is extremely sensitive to interface friction. J Biomech

1996; 29:1569-1575.

Withers PJ, Bhadeshia HKDH. Residual stress. Part I –

measurement techniques. Mater Sci Tech 2001; 17: 355-365.

Wu MK, Pehlivan Y, Kontakiotis EG, Wesselink PR. Microleakage

along apical root fillings and cemented posts. J Prosthet Dent

1998;79:264-269.

Yoshiyama M, Carvalho RM, Sano H, Horner JA, Brewer PD,

Pashley DH. Regional bond strength of resin to human root dentin. J

Dent Res 1996; 24: 435-442.

Yoshiyama M, Sano H, Ebisu S, Tagami J, Ciucchi B, Carvalho RM,

Johnson MH, Pashley DH. Regional strengths of bonding agents to

cervical sclerotic dentin. J Dent Res 1996; 75: 1404-1413.

205

Zhang Y, Agee K, Nör J, Sachar B, Russel C, Pashley DH. Effect of

acid etching on the tensile properties of demineralized dentin matrix.

Dent Mater 1998; 14: 222-228.

206

CURRICULUM VITAE Date of birth: June 22nd, 1971

Place of birth: Orbetello (Grosseto), Italy

Civil status: Unmarried

Citizenship: Italian

Research activity 1999: Master of Science in Dental Research at Tufts University School of

Dental Medicine, Boston, U.S.A

2000: One-year scholarship for research activity at the School of Dentistry of

the University of Siena, Siena, Italy.

Professional positions: Institutional

2002- Professor of Basic Principles of Dentistry, School of Dentistry,

University of Siena, Italy

Private

Office: 4 via S. Martino, Orbetello (GR) 58015, Italy

Telephone and fax: +39(0564)867071

E-mail: cecilia.goracci@tin.it

Professional Organizations membership 2004- Academy of Dental Materials

2001 and 2003 International Association for Dental Research

2003 Società Italiana di Odontoiatria Conservatrice (Italian Society of

Restorative Dentistry)

1996-1998 and 2000- American Association of Orthodontists.

1993- Società Italiana di Ortodonzia (Italian Society of Orthodontics)

207

Publications Goracci C, Tavares AU, Fabianelli A, Monticelli A, Raffaelli O, Cardoso PEC, Tay FR, Ferrari M. The adhesion between fiber posts and root canal walls: comparison between microtensile and push-out bond strength measurements. Eur J Oral Sci 2004; 112: 353-361. Goracci C, Raffaelli O, Monticelli F, Balleri B, Bertelli E, Ferrari M. The adhesion between fiber posts and composite resin cores: microtensile bond strength with and without post silanization. Dental Materials 2004, in press. Goracci C, Sadek FT, Fabianelli A, Tay FR, Ferrari M. Evaluation of the adhesion of fiber posts to intraradicular dentin. Accepted for publication in Operative Dentistry (2004). Goracci C, Sadek FT, Monticelli F, Cardoso PEC, Ferrari M. Microtensile bond strength to ground enamel and dentin of simplified ashesives. Journal of Adhesive Dentistry 2004; 6. Goracci C, Sadek FT, Monticelli F, Cardoso PEC, Ferrari M. Influence of substrate, shape, and thickness on microtensile specimens’ structural integrity and their measured bond strengths. Dental Materials 2004; 20 (7): 643-654. Goracci C, Bertelli E, Ferrari M. Bonding to worn or fractured incisal edges: shear bond strength of new adhesive systems. Quintessence International 2004; 35: 1-7. Goracci C, Gheewalla R, Kugel G, Ferrari M. Orthodontic-restorative treatment of chipped or worn incisors. American Journal of Dentistry 2001; 14: 50-55. Ferrari M, Mason PN, Goracci C, Tay FR. Collagen degradation in endodontically treated teeth after clinical function. Journal of Dental Research 2004, in press. Papacchini F, Goracci C, Sadek FT, Monticelli F, García-Godoy F, Ferrari M. Microtensile bond strength to enamel by glass-ionomers, resin-modified glass-ionomers, and resin composites used as pit and fissure sealants. Journal of Dentistry 2004, in press. Fabianelli A, Goracci C, Monticelli F, Grandini S, Ferrari M. In vitro evaluation of wall-to-wall adaptation of a self-adhesive resin cement used for luting gold and porcelain inlays. Journal of Adhesive Dentistry 2004, in press.

208

Grandini S, Goracci C, Monticelli F, Tay FR, Ferrari M. Fatigue resistance and structural integrity of fiber posts: three-bending test and SEM evaluation. Dental Materials 2004, in press. Chersoni S, Suppa P, Grandini S, Goracci C, Monticelli F, Yiu C, Huang C, Prati C, Breschi L, Ferrari M, Pashley DH, Tay FR. In vivo and in vitro permeability of one-step self-etch adhesives. Journal of Dental Research 2004; 83: 459-464. Chieffi N, Goracci C, Simonetti M, Monticelli F, Ferrari M. The effect of pre-sealing tooth preparations with dentin adhesives on permanent crown cementation: a pilot study. Journal of Adhesion Dentistry 2004, in press. Gotti G, Goracci, Garcìa-Godoy F, Ferrari M. Microscopic evaluation of the bonding mechanism of an adhesive material to primary teeth. Journal of Dentistry for Children 2004; 71:54-60. Monticelli F, Goracci C, Ferrari M. Micromorphology of the fiber post-resin core unit: a scanning electron microscopy evaluation. Dental Materials 2004; 20: 176-183. Monticelli F, Grandini S, Goracci C, Ferrari M. Clinical behavior of translucent fiber posts and luting and restorative materials: a 2-year report. International Journal of Prosthodontics 2003; 16: 593-596. Ferrari M, Goracci C, Sadek F, Cardoso PEC. Microtensile bond strength tests: SEM evaluation of samples integrity before testing. European Journal of Oral Sciences 2002; 110: 385-391. Ferrari M, Goracci C, Monticelli F, Sadek FT, Cardoso PEC. Adhesion testing with the microtensile method: effects of dental substrate and adhesive system on bond strength measurements. Journal of Adhesive Dentistry 2002; 4: 291-297. Fabianelli A, Goracci C, Ferrari M. Sealing ability of packable resin composites in Class II restorations. Journal of Adhesive Dentistry 2003; 5: 217-223. Grandini S, Sapio S, Goracci C, Monticelli F, Ferrari M. SEM evaluation of the cement layer thickness after the luting procedure of two different fiber posts.

209

Endodontics and Dental Traumatology, 2004, in press. Gesi A, Magnolfi A, Goracci C, Ferrari M. Comparison of two techniques for fiber posts removal. Journal of Endodontics 2003; 29: 580-582. Ferrari M, Grandini S, Simonetti M, Monticelli F, Goracci C. Influence of a microbrush on bonding fiber posts into root canals under clinical conditions. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontics 2002; 94: 627-631. Ferrari M, Vichi A, Grandini S, Goracci C. Efficacy of a self-curing adhesive-resin cement system on luting glass-fiber posts into root canals: an SEM investigation. International Journal of Prosthodontics 2001; 14: 543-549. Abstracts Goracci C, Sadek FT, Grandini S, Vichi A, Borracchini A, Tay FR, Ferrari M. The adhesion to root canal dentin of fiber post luting agents: push-out bond strength measurements and transmission electron microscope evaluation. CED IADR Meeting Istanbul 2004 Goracci C, Monticelli F, Tavares AU, Sadek FT, Cardoso PEC, Ferrari M. The adhesion between fiber posts and composite resin cores: microtensile bond strength of different combinations of adhesives. Journal of Dental Research, 2003; 82: Abstract 1268. Sadek FT, Goracci C, Monticelli F, Ferrari M, Cardoso PEC. Influence of substrate, shape, and thickness on microtensile specimens’ structural integrity and their measured bond strength. Journal of Dental Research, 2003; 82: Abstract 1436. Monticelli F, Grandini S, Goracci C, Ferrari M, Tay FR. Transmission electronic microscopic evaluation of a self-adhesive material luted to different dental substrates. Journal of Dental Research, 2003; 82: Abstract1439. Grandini S, Borracchini A, Goracci C, Monticelli F, Ferrari M. SEM study to compare the luting procedures of two different fiber posts. Journal of Dental Research, 2003; 82: Abstract 1442. Fabianelli A, Grandini S, Goracci C, Ferrari M. One-year clinical trial of Gradia Direct Class II restorations. Journal of Dental Research, 2003; 82: Abstract 1472. Monticelli F, Goracci C, Grandini S, Bertelli E, Balleri P, Ferrari M. Scanning electron microscopic evaluation of fiber post-resin core units.

210

Journal of Dental Research, 2003; 82: Abstract 1953. Grandini S, Goracci C, Monticelli F, Tay FR, Ferrari M. Fatigue resistance of different kinds of fiber posts. Journal of Dental Research, 2003; 82: Abstract 2935. Fuentes MV, Monticelli F, Goracci C, Toledano M, Ferrari M. Microtensile bond strength of different self-etch adhesives to sound human dentin. Journal of Dental Research, 2003; 82: Abstract 0351. Grandini S, Balleri P, Goracci C, Monticelli F, Bertelli E, Ferrari M. Scanning Electron Microscopic Evaluation of two different techniques for luting glass fiber posts. Meeting of the European Societies of Restorative Dentistry (Conseuro), Monaco (Germany), June 5-8, 2003 Monticelli F, Goracci C, Balleri P, Grandini S, Ferrari M. Clinical behavior of translucent fiber posts and luting restorative materials: a 2-year report. Meeting of the European Societies of Restorative Dentistry (Conseuro), Monaco (Germany), June 5-8, 2003 Grandini S, Ferrari M, Goracci G, Bertelli E. Quantitative evaluation of dentin morphology in root canals after shaping and irrigation of the endodontic space. Journal of Dental Research 81 (special issue B) 2002, #52, pag B238